Frascati Manual
The Frascati Manual is the OECD's internationally recognized methodology for collecting, reporting, and measuring statistics on research and experimental development (R&D), ensuring comparability across countries and sectors to support evidence-based policymaking.[1] First published in 1963 following expert meetings in Frascati, Italy—hence its name—it has evolved through multiple editions to address changes in economic contexts, globalization, and innovation practices, with the seventh edition released in 2015 as the current standard.[2][3] At its core, the manual defines R&D as creative and systematic work undertaken in the hope of increasing the stock of knowledge—from fundamental principles to specific applications—and includes its three main activities: basic research (aimed at advancing knowledge without immediate practical goals), applied research (directed toward practical problems or specific aims), and experimental development (systematic efforts to produce new or improved products, processes, or services).[3] These activities must meet criteria of novelty, creative uncertainty, a systematic approach, and transferable results to qualify as R&D, distinguishing them from routine testing, data collection, or market research.[3] The manual structures R&D measurement across five institutional sectors: the business enterprise sector (market-oriented producers), government sector (public non-market entities), higher education sector (universities and research institutes), private non-profit sector (independent non-governmental organizations), and the rest of the world (for international funding flows).[3] It provides detailed guidelines for quantifying R&D inputs, including expenditures (intramural and extramural, at purchasers' prices excluding VAT and depreciation, aggregated as gross domestic expenditure on R&D or GERD) and personnel (researchers, technicians, and support staff, measured in full-time equivalents and headcounts).[3] Funding sources—such as internal funds, government budgets (GBARD), or tax incentives (GTARD)—are tracked by sector and socioeconomic objectives, with recommendations for data collection via surveys, administrative records, and quality controls like imputation and estimation.[3] Sector-specific chapters address unique challenges, such as multinational enterprises in business R&D, general university funds in higher education, or cross-border globalization effects, while annexes offer glossaries, historical context, and classifications like the Fields of Research and Development (FORD).[3] Ongoing OECD efforts continue to refine the manual to incorporate emerging trends like software development and mission-oriented innovation, maintaining its role as a foundational tool for global R&D statistics that underpin indicators like GERD-to-GDP ratios and inform international comparisons.[4][5]Overview
Purpose and Scope
The Frascati Manual serves as the internationally recognized guideline issued by the Organisation for Economic Co-operation and Development (OECD) to standardize the measurement of resources devoted to research and experimental development (R&D). Its primary purpose is to provide a comparable methodology for collecting, reporting, and analyzing R&D statistics, enabling evidence-based policymaking and economic analysis across countries. By establishing a common language and consistent practices, the manual ensures that R&D data—focusing on financial expenditures and human resources—can be reliably compared globally, treating R&D as a capital investment in alignment with the System of National Accounts (SNA) 2008. The scope of the Frascati Manual encompasses definitions, survey methodologies, and reporting frameworks for R&D activities within scientific, technological, and related innovation contexts. It applies universally across all institutional sectors, including the business enterprise, government, higher education, private non-profit, and rest-of-the-world sectors, while addressing funding flows, performance metrics, trade in R&D services, and fields of research. Key aggregates covered include gross domestic expenditure on R&D (GERD), government budget appropriations or outlays for R&D (GBARD), and government total R&D (GTARD), with an emphasis on intramural expenditures, personnel, and current costs to account for globalization and multinational enterprises. Targeted primarily at national statistical offices, policymakers, researchers, and experts involved in R&D surveys, the manual promotes enhanced timeliness, policy relevance, and alignment with socioeconomic objectives in data collection. As a core component of the OECD's "Frascati Family" of guidelines—endorsed by the Committees for Scientific and Technological Policy (CSTP) and Statistics and Statistical Policy (CSSP)—it focuses exclusively on R&D measurement, distinct from related efforts like innovation statistics in the Oslo Manual, while integrating with broader OECD frameworks such as the SNA and Government Finance Statistics (GFS).Importance in Global R&D Statistics
The Frascati Manual serves as the foundational framework for compiling internationally comparable statistics on research and experimental development (R&D) expenditures and personnel, enabling cross-country analyses that reveal disparities and trends in innovation efforts worldwide. By standardizing definitions and measurement methodologies, it supports key global benchmarks such as Gross Domestic Expenditure on R&D (GERD), which aggregates total intramural R&D spending across institutional sectors like business enterprise, government, higher education, and private non-profits within a nation's borders. This allows policymakers and researchers to track resource allocation and performance, for instance, comparing how countries prioritize R&D funding sources and performers to identify leaders in technological advancement. The manual significantly influences public policy by providing reliable data for evaluating national innovation capacity, guiding fund allocation, and monitoring progress toward international objectives, including United Nations Sustainable Development Goal (SDG) 9.5, which targets enhanced scientific research and upgrading technological capabilities. Under SDG indicator 9.5.1, R&D expenditure as a proportion of GDP is measured using Frascati definitions to assess global efforts in building resilient infrastructure and fostering innovation. Governments and organizations like the OECD rely on these metrics to design incentives, such as tax credits or grants, ensuring that R&D investments align with broader economic and societal goals.[6] In economic analysis, the Frascati Manual quantifies the contributions of R&D to knowledge-based economies, with R&D intensity—defined as GERD relative to gross domestic product (GDP)—emerging as a critical indicator of a country's commitment to long-term growth and competitiveness. For example, this ratio helps contextualize how R&D drives productivity and innovation, integrating with national accounts systems to treat R&D as a capital asset rather than mere expense. Such insights underscore R&D's role in sustainable development and global value chains. Standardization through the manual minimizes discrepancies in national reporting practices, promoting data harmonization across diverse economic contexts and facilitating collaborative efforts by bodies like the OECD and UNESCO. UNESCO's Institute for Statistics adopts Frascati guidelines to ensure consistent global R&D data collection, reducing errors from varying interpretations and enabling aggregated analyses for over 150 countries. This uniformity enhances the credibility of international databases, supporting evidence-based decisions on global challenges like climate change and health.[7]Core Concepts
Definition of Research and Experimental Development
The Frascati Manual provides the internationally recognized definition of research and experimental development (R&D) as "creative and systematic work undertaken in order to increase the stock of knowledge – including knowledge of humankind, culture and society – and to devise new applications of available knowledge." This definition emphasizes the dual objectives of R&D: expanding fundamental knowledge across scientific, social, and humanistic domains while applying that knowledge innovatively to practical ends. To qualify as R&D, activities must meet five key criteria: novelty, where the work aims at new findings (either in the theoretical or practical sense); creativity, drawing on original concepts and hypotheses; uncertainty, involving unpredictable outcomes or resource requirements; systematic organization, with planned investigations, budgets, and documented records; and transferability or reproducibility, allowing results to be reproduced or shared for further use. These criteria ensure that only innovative efforts with potential for broader impact are classified as R&D, distinguishing them from standard operational tasks. The manual explicitly distinguishes R&D from non-R&D activities, such as routine data collection, standard quality control, or market research, which lack innovative elements or do not advance knowledge in a novel way. For instance, routine testing or maintenance in industrial settings does not qualify, whereas targeted experiments to improve processes under uncertain conditions do. Similarly, educational activities or policy implementation using established methods fall outside R&D, as they do not involve creative uncertainty or systematic novelty. The 2015 edition of the Frascati Manual refined this definition without altering its core wording, but it expanded guidance on boundaries to better incorporate social sciences, humanities, and arts, reflecting their growing role in knowledge production and aligning with updates in the System of National Accounts (SNA 2008) that treat R&D as capital formation. This evolution underscores the manual's adaptability to interdisciplinary advancements while maintaining rigorous criteria for global statistical consistency.Types of Research Activities
The Frascati Manual delineates research and development (R&D) into three primary types of activities: basic research, applied research, and experimental development. These categories form the core framework for classifying R&D efforts in statistical reporting, emphasizing the intent, objectives, and outputs of the work. All activities must satisfy five fundamental criteria: novelty, creative advance, uncertainty, systematic methodology, and the potential for transferability or reproducibility of results. Basic research entails experimental or theoretical work undertaken primarily to acquire new knowledge regarding the underlying foundations of phenomena and observable facts, without any particular application or use in view. It focuses on expanding the general body of knowledge, often through fundamental inquiries into natural or social processes, and is typically conducted in academic or research institutions where results are disseminated via publications. For instance, investigations into the fundamental properties of materials, such as studying polymerization reactions or modeling crystal absorption in physics, exemplify basic research by prioritizing theoretical understanding over immediate practical utility. Applied research involves original investigations aimed at acquiring new knowledge directed primarily toward a specific practical aim or objective. It builds upon existing knowledge, including findings from basic research, to address particular problems or needs, with an emphasis on generating actionable insights for potential implementation. Examples include efforts to develop new vaccines targeting known diseases or optimizing production processes for polymers to enhance efficiency in industrial applications. Unlike basic research, applied research is oriented toward medium- to short-term outcomes that can inform subsequent development. Experimental development comprises systematic work that draws on knowledge from research and practical experience to produce new or improved materials, products, devices, processes, or services, or to enhance existing ones. This type emphasizes the creation and testing of prototypes or pilot models to verify feasibility, often involving iterative design and evaluation under real-world conditions. Representative cases include prototyping innovative software tools to resolve technological uncertainties or building and testing new medical devices, such as radiation detectors, to ensure they meet performance standards before broader production. Experimental development is distinct in its focus on tangible outputs and is common in business enterprises. The boundaries between these types are not always rigid, as projects may evolve or incorporate elements from multiple categories, with classification guided by the primary intent and expected outcomes at the project level. For example, basic research may transition into applied research when initial findings are explored for broader future applications, while applied research can overlap with experimental development during early prototyping phases where knowledge acquisition supports product creation. The Manual advises assessing the dominant purpose—whether knowledge expansion without application (basic), problem-solving with practical goals (applied), or production-oriented innovation (experimental)—to distinguish activities, acknowledging that overlaps are common in interdisciplinary or iterative R&D endeavors. The 2015 edition of the Frascati Manual introduced clarifications to address evolving R&D practices, particularly expanding the scope of experimental development to explicitly include software creation and artistic endeavors that meet the core criteria. Software development qualifies as experimental development when it involves advancing scientific or technological knowledge, such as developing new algorithms or operating systems to overcome uncertainties, but excludes routine programming or maintenance tasks. Similarly, artistic research is incorporated under experimental development if it systematically generates novel knowledge, for example through experimental projects that produce new pedagogical materials in the arts, while pure artistic expression without research intent remains outside R&D boundaries. These updates reflect greater inclusion of social sciences, humanities, and digital innovations in global R&D measurement.Classifications of Fields and Sectors
The Frascati Manual provides standardized frameworks for classifying research and experimental development (R&D) activities by fields and institutional sectors, enabling consistent international comparisons of R&D data. The Fields of Research and Development (FORD) classification organizes R&D into a hierarchical structure with six broad fields at the first level, further subdivided into more detailed categories at the second and third levels (e.g., two-digit and three-digit codes). This system ensures that R&D expenditures, personnel, and outputs can be allocated based on the primary objectives or domains of the work, with flexibility for multi-disciplinary projects.[8] The six broad FORD fields are as follows:| Code | Broad Field | Description |
|---|---|---|
| 1 | Natural Sciences | Encompasses fundamental sciences including physical, chemical, earth, space, and life sciences, focusing on understanding natural phenomena. |
| 2 | Engineering and Technology | Covers applied engineering disciplines, technological innovations, and development of tools and processes. |
| 3 | Medical and Health Sciences | Includes basic and clinical research on human health, disease prevention, and medical technologies. |
| 4 | Agricultural Sciences | Addresses agriculture, forestry, fisheries, veterinary medicine, and related food production systems. |
| 5 | Social Sciences | Studies human behavior, societies, economies, and institutions, including psychology, economics, and sociology. |
| 6 | Humanities and the Arts | Focuses on cultural, historical, linguistic, and artistic inquiries, such as history, philosophy, and literature. |
Historical Development
Origins in the 1960s
In the aftermath of World War II, governments worldwide recognized the critical role of science and technology in economic reconstruction and growth, leading to a surge in public investments in research and development (R&D). This post-war emphasis on science policy highlighted the need for reliable, comparable statistics to inform decision-making and track progress, as pioneering efforts in countries like the United States, Japan, Canada, the United Kingdom, the Netherlands, and France revealed significant inconsistencies in R&D data collection methods. To address these challenges, the Organisation for Economic Co-operation and Development (OECD), established in 1961 to promote economic growth and international cooperation, initiated efforts to standardize R&D measurement. In February 1962, the OECD formed an ad hoc group of national experts on science and technology, later evolving into the National Experts on Science and Technology Indicators (NESTI) group, which convened to develop a unified framework for R&D surveys.[3][9][10] The pivotal event in the manual's origins occurred from June 17 to 21, 1963, when the OECD convened a working meeting of national experts at the Villa Falconieri in Frascati, Italy—a location chosen for its proximity to Rome and facilities suitable for international collaboration. Hosted by the OECD's Directorate for Scientific Affairs, the gathering brought together statisticians and policymakers from member countries to review and refine a draft document prepared earlier that year. Participants proposed and unanimously endorsed standard practices for conducting R&D surveys, emphasizing the importance of harmonized methodologies to enable cross-national comparisons and stimulate systematic data gathering on scientific activities. This meeting marked the first concerted international effort to define R&D in a way that distinguished it from other forms of innovation or routine work, laying the groundwork for consistent global reporting.[9][3] The outcome of the Frascati meeting was the publication of the first edition of the manual in 1963, officially titled Proposed Standard Practice for Surveys of Research and Experimental Development (OECD DAS/PD/62.47). This inaugural document focused on providing basic definitions of R&D—such as creative work undertaken on a systematic basis to increase the stock of knowledge—and simple classifications of sectors like higher education, government, and industry to resolve discrepancies in how nations measured R&D expenditures and personnel. By prioritizing these foundational elements, the manual addressed the theoretical and practical challenges in scoping R&D activities, ensuring that statistics captured only those efforts aimed at new knowledge or applications, rather than broader technological advancements.[11][9] Contributions to the first edition came primarily from statisticians and experts representing OECD member countries, who offered critical feedback on the draft and shaped key sections through collaborative revisions during and after the Frascati meeting. Notable among them was Christopher Freeman, a consultant appointed by the OECD in 1962 to author the initial draft based on prior work from the European Productivity Agency, and Dr. J. C. Gerritsen, who conducted comparative studies on R&D measurement practices in the UK, France, the US, and Canada. These inputs ensured the manual's practicality for national statistical offices, establishing it as a tool for all OECD members to adopt uniform survey objectives and categories.[9][3]Editions and Key Revisions
The Frascati Manual has evolved through seven editions since its inception, with each iteration refining methodologies to address emerging challenges in measuring research and experimental development (R&D). The first edition, published in 1963, established the basic framework for standardizing R&D surveys, focusing on definitions and data collection practices for government and other sectors.[3] The second edition in 1970 refined these definitions, aligned them with United Nations standards such as the System of National Accounts (SNA) and International Standard Industrial Classification (ISIC), and improved international comparability through enhanced sectoral classifications.[3] The third edition of 1976 introduced the Government Budget Appropriations or Outlays for R&D (GBAORD) indicator, expanded coverage to social sciences and humanities, and updated classifications to better support functional and sectoral breakdowns.[3] Subsequent editions built on this foundation amid growing economic and technological shifts. The fourth edition, released in 1980 (with minor updates in 1981), provided enhanced guidance on measuring human and financial resources, including institutional sector breakdowns, to improve data quality and comparability across OECD member countries.[3] The fifth edition in 1993 (revised in 1994) emphasized globalization's impact on R&D, introduced more detailed institutional sectors, and aligned further with evolving SNA frameworks to capture funding flows and socioeconomic objectives.[3] The sixth edition of 2002 marked a significant shift by treating R&D as a capital investment under SNA 1993, adding emphasis on intangible assets, service sector R&D, tax relief incentives, and globalization effects such as offshoring; it also refined classifications for fields of research and institutional sectors.[3] The seventh edition, published in 2015, incorporated SNA 2008 updates by formalizing R&D as capital formation, with detailed guidance on capital expenditure measurement, including clarification of R&D activities in services and multinational enterprises.[3] It renamed GBAORD to Government Budget Allocations for R&D (GBARD) for precision, expanded sectoral coverage (e.g., business, higher education, and private non-profits), and addressed globalization challenges like offshoring and funding flows, driven by user feedback from OECD members and advances in digital technologies.[3] These revisions reflect broader drivers, including economic globalization, the rise of knowledge-based economies, and the need for better integration with innovation metrics.[4] Since 2015, the OECD's National Experts on Science and Technology Indicators (NESTI) group has focused on implementation support rather than a full new edition, issuing supplementary guidance on topics like software development and R&D measurement in emerging areas, with no eighth edition released by 2025 despite ongoing discussions for future revisions.[12][13] This work is motivated by technological advances, such as digital transformation, and feedback from national statistical offices to ensure the manual remains relevant.[4]Applications and Implementation
Data Collection Methodologies
The Frascati Manual outlines standardized methodologies for collecting R&D data to ensure international comparability and reliability, emphasizing performer-based surveys that capture intramural activities within statistical units such as enterprises, institutions, and organizations. These methodologies prioritize the identification of R&D through clear definitions and guidelines, focusing on core metrics that reflect the scale and nature of R&D efforts across sectors like business enterprise, government, higher education, and private non-profit. Data are broken down by these sectors and fields of research and development (FORD) to provide granular insights into R&D distribution.[3] Central to these methodologies are the core metrics of intramural R&D expenditure and R&D personnel, along with data on funding sources. Intramural R&D expenditure encompasses current costs—such as labor, materials, and other operational expenses—and capital expenditures, including investments in land, buildings, equipment, and software, all valued at purchasers' prices excluding deductible VAT. This metric excludes depreciation, amortization, and acquired R&D, ensuring it reflects only internally performed activities regardless of funding origin. R&D personnel are measured in both headcounts and full-time equivalents (FTE), categorizing individuals into researchers (those leading or contributing to R&D), technicians (applying scientific knowledge), and other support staff, while excluding indirect support roles. Funding sources are tracked to identify intramural versus extramural flows, including government budgets, business enterprise funding, and international contributions, preventing double-counting in performer-reported data.[3]| Metric | Key Components | Measurement Approach |
|---|---|---|
| Intramural R&D Expenditure | Current (labor, other costs); Capital (equipment, software) | Purchasers' prices; excludes VAT, depreciation |
| R&D Personnel | Researchers, technicians, support; by function and demographics | Headcounts and FTE; includes external personnel |
| Funding Sources | Intramural/extramural; by sector (e.g., government, business) | Performer-based reporting to avoid double-counting |