Frascati Manual

Definition of R&D

“Research and experimental development (R&D) comprise 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.”

Frascati Manual 2015, §1.32

Note the words “culture and society”: the definition covers research in the social sciences, humanities and the arts, not only laboratories and engineering, although artistic performance itself is normally excluded (§2.67).

Five criteria

An activity counts as R&D only if it meets all five core criteria jointly (§1.33, §2.13). It must be:

1. Novel: aimed at new findings.
2. Creative: based on original, not obvious, concepts and hypotheses.
3. Uncertain: about the final outcome.
4. Systematic: planned and budgeted.
5. Transferable and/or reproducible: leading to results that could be possibly reproduced.

Criteria named in §1.33; explanations follow the headings of §§2.14–2.20.

To apply them to a real project, see the article Is it R&D? A five-question test.

The Oslo Manual uses this definition, with the same five criteria, when it counts R&D among a firm’s innovation activities.

Three types of R&D

Basic research

Experimental or theoretical work undertaken primarily to acquire new knowledge of the underlying foundations of phenomena and observable facts, without any particular application or use in view.

Applied research

Original investigation undertaken in order to acquire new knowledge, directed primarily towards a specific, practical aim or objective.

Experimental development

Systematic work, drawing on knowledge gained from research and practical experience and producing additional knowledge, directed to producing new products or processes or to improving existing ones.

Definitions from §2.25, §2.29 and §2.32 (summarised in §1.35). The order is not a sequence: the manual describes the relationship between the three types as dynamic rather than linear, with feedback between them (§2.38).

What is not R&D

Many activities sit close to R&D, share its people and institutions, and should still, as far as possible, be excluded when R&D is measured (§2.111). The manual lists them in section 2.8:

Scientific and technical information services

Collecting, coding, classifying, translating, analysing or disseminating information, for example in libraries, patent services or advisory services or scientific conferences. The exception is work done solely or primarily to support R&D: writing up the original report of R&D findings is part of R&D (§2.112).

Testing and standardisation

Maintaining national standards, calibrating secondary standards and routine testing and analysis of materials, products, soils or air (§2.113).

Feasibility studies

Investigating a proposed engineering project with existing techniques before deciding to go ahead is not R&D. A feasibility study for a research project is (§2.114).

Specialised health care

Routine investigation and normal application of specialised medical knowledge is usually not R&D, although there may be an R&D element, for example in university hospitals (§2.115).

Policy-related studies

Close support for policy and legislation, such as policy advice, legal advice, public relations or administrative support, is excluded. Research that gives decision makers thorough knowledge of social, economic or natural phenomena is included (§§2.116–2.118).

Programmatic evaluations

Ex-ante and ex-post appraisals may in some cases meet the R&D criteria, but not always. The manual warns of “a significant risk that some types of socio-economic consultancy … are inaccurately presented as R&D” (§2.119).

Purely R&D-financing activities

Raising, managing and distributing R&D grants by ministries, agencies, foundations or charities (§2.121).

Indirect supporting activities

Transport, storage, cleaning, repair, maintenance, security and central administration. They are left out of R&D personnel, but their cost enters R&D expenditure as overheads (§2.122).

Artistic performance is also normally excluded from R&D (§2.67).

Is software development R&D?

Sometimes. Software development is part of many projects that contain no R&D at all. It counts as R&D only if it leads to an advance in the area of computer software, and using software for a new application does not by itself count as an advance (§2.70).

In 2025 the OECD added guidance on software. Its main message: a software project counts as R&D only if it meets all five criteria, and particular attention should go to novelty, creativity and uncertainty, “which many software development projects do not necessarily satisfy in full or in part”. Respondents are encouraged to keep records that separate eligible R&D from the rest of a software project. The guidance is meant mainly for training statisticians and for their contact with respondents, not as material for comprehensive inclusion in surveys; questions on software R&D need not feature in them regularly.

Who performs R&D: five sectors

R&D statistics group institutions into sectors, following the System of National Accounts (chapter 3):

• Business enterprise
• Government
• Higher education: a sector unique to the Frascati Manual, because of its importance for policy. It brings together institutions that can belong to any national-accounts sector.
• Private non-profit
• Rest of the world: used only to show who funds R&D (formerly called “Abroad”).

Counting the money: GERD and related indicators

The basic unit is intramural R&D expenditure: all current expenditure plus gross fixed capital expenditure for R&D performed within a unit during a given period, whatever the source of funds (§4.10).

“GERD is total intramural expenditure on R&D performed in the national territory during a specific reference period.”

Frascati Manual 2015, §4.7

GERD, gross domestic expenditure on R&D, is a country's principal R&D indicator. It includes R&D carried out in the country with foreign money, but not R&D that the country funds abroad (§4.8). Divided by GDP, it gives R&D intensity, widely used to compare countries (§4.162); how to read it is explained in the article R&D spending as a share of GDP. Its parts by performing sector have their own names:

BERD

Business enterprise expenditure on R&D

GOVERD

Government expenditure on R&D

HERD

Higher education expenditure on R&D

PNPERD

Private non-profit expenditure on R&D

Two indicators look at government support from the other side. GBARD, government budget allocations for R&D, shows what governments plan to spend on R&D from their budgets (chapter 12). GTARD, government tax relief for R&D, measures the cost of R&D tax incentives separately (chapter 13).

Counting the people: headcount and FTE

R&D personnel fall into three groups:

Researchers

“Professionals engaged in the conception or creation of new knowledge” who conduct research and develop concepts, theories, models, techniques, instrumentation, software or operational methods (§5.35).

Technicians and equivalent staff

People whose main tasks require technical knowledge and experience and who carry out scientific and technical tasks in R&D, normally under the supervision of researchers (§5.40).

Other supporting staff

Skilled and unskilled craftsmen and administrative, secretarial and clerical staff working on R&D projects or directly associated with them (§5.43).

Each group is counted twice: as a headcount (number of people) and in full-time equivalents (FTE), the ratio of working hours actually spent on R&D to the total number of hours conventionally worked in the same period (§§5.46, 5.49). For example, a lecturer who spends 40% of working time on research counts as one person in headcount and 0.4 in FTE. FTE gives a more accurate estimate of the human resources devoted to R&D (§5.60).

R&D personnel are only part of the people with scientific and technical skills. The wider group, human resources in science and technology (HRST), is defined in the Canberra Manual; see how the two differ.

Fields of R&D

R&D is also classified by field, using the Fields of Research and Development classification (FORD, Table 2.2). It has six broad fields, each divided into more detailed ones:

1. Natural sciences
2. Engineering and technology
3. Medical and health sciences
4. Agricultural and veterinary sciences
5. Social sciences
6. Humanities and the arts

How Poland applies the manual

Statistics Poland (GUS) runs the R&D survey on the basis of the Frascati Manual 2015, which keeps Polish data comparable with other countries. The survey is compulsory, covers all eligible units and is carried out every year. Since 2009 the work has been led by the Statistical Office in Szczecin, and since reporting year 2021 all R&D data, from companies, universities and public bodies alike, are collected on a single form, PNT-01.

The obligation to report covers, among others, ministries, the National Centre for Research and Development (NCBR), the National Science Centre (NCN), universities, units whose main activity is research and development, and any other unit that performs R&D alongside its main business, whatever its size. To find such units, GUS also uses administrative lists, including those of taxpayers who claimed the R&D tax relief or the IP Box.

Poland in figures

According to GUS (released on 22 October 2026), in 2025:

• 7,018 entities were involved in R&D, 3.5% fewer than a year earlier, and GERD reached PLN 58.8 billion, 14.2% more than in 2024 (current prices). R&D intensity rose to 1.50% of GDP, from 1.41%. GERD per inhabitant was PLN 1,571.
• Business enterprises performed the most R&D: BERD was PLN 38.2 billion, or 64.9% of GERD. The higher education sector accounted for 32.8% of GERD, government for 2.2% and the private non-profit sector for 0.1%.
• Business funded 58.8% of GERD and government 32.7%.
• R&D personnel, including external personnel, numbered 329.1 thousand people, equal to 206.7 thousand FTE. Internal staff made up 82.2% of headcount and 83.9% of FTE. Researchers made up 69.6% of internal personnel in headcount and 73.5% in FTE.

Breakdowns by type of R&D, field and region come from the 2024 edition:

• In 2024 experimental development accounted for 59.0% of GERD (PLN 30.4 billion), basic research for PLN 15.0 billion and applied research for PLN 6.1 billion.
• By field: engineering and technology 44.2%, natural sciences 27.3%, medical and health sciences 14.2%, social sciences 6.5%, humanities and the arts 4.0%, agricultural and veterinary sciences 3.8%.
• The Mazowieckie voivodeship accounted for 36.1% of intramural R&D expenditure, Małopolskie for 14.4%.

Is a new edition coming?

Not as of June 2025. The manual is looked after by an OECD group of national experts on science and technology indicators (NESTI), which adds guidance alongside the 2015 text instead of rewriting it. The software guidance was published in June 2025 as an OECD Technical Paper of complementary guidance to the 2015 edition. In it, NESTI states that “for the time being” it does not envisage a wholesale revision of the manual.

Polish edition

Statistics Poland (GUS) published the Polish edition, Podręcznik Frascati 2015 (ISBN 978-83-88718-97-7), in 2018. The edition states that it is a translation published in agreement with the OECD, not an official OECD translation, and that in case of any discrepancy only the original text is valid.