29 September 2010

This collection provides users with data concerning R&D expenditure and R&D personnel broken down by following institutional sectors: business enterprise (BES), government (GOV), higher education (HES), private non-profit (PNP) and total of sectors.

All data are broken down by the above mentioned sectors of performance. The R&D expenditure is further broken down by source of funds, by type of costs, by type of economic activity (NACE), by size class, by type of R&D, by fields of science, by socio-economic objectives and by regions (NUTS 2).

Besides R&D expenditures in basic unit National currency (MIO_NAC) the following units are available: Euro (MIO_EUR), Euro per inhabitant (EUR_HAB) Purchasing Power Standard (MIO_PPS), Purchasing Power Standard at 2000 prices (MIO_PPS_KP00), Purchasing Power Standard per inhabitant at constant 2000 prices (PPS_KP00_HAB), Percentage of GDP (PC_GDP).

R&D personnel data is available in full-time equivalent (FTE), in head count (HC), as a % of employment and as a % of labour force. The data is further broken down by occupation, by qualification, by gender, by size class, by citizenship, by age groups, by fields of science, by economic activity (NACE) and by regions (NUTS 2).

More details on breakdowns, derivations and different units available are found in the Eurostat indicators on R&D (See annex at the bottom of the page).

R&D data are available for following countries and country groups:

- All EU Member States, plus Candidate Countries, EFTA Countries, the Russian Federation, China, Japan, the United States and South Korea.
- Country groups: EU-27, EU-15 and EA-16.

R&D data are compiled in accordance to the guidelines laid down in the Proposed standard practice for surveys of research and experimental development - Frascati Manual (FM), OECD, 2002 (See annex at the bottom of the page).

R&D statistics are compiled in line with international statistical classifications such as ISIC, NACE, FOS, ISCED and NABS and are to some extent adapted to the system of national accounts (SNA) according to the FM recommendations (See annex at the bottom of the page).

In addition to the FM recommendations (See annex at the bottom of the page), regional breakdown for EU Member States, Candidate Countries and EFTA countries are complied following

• the Regional Dimension of R&D and Innovation Statistics - Regional Manual, European Commission, 2003 (See annex at the bottom of the page).
• and the Nomenclature of Territorial Units for Statistics, NUTS, Eurostat.

The regional breakdown of the EU Member States is based on the Nomenclature of Territorial Units for Statistics (NUTS).

R&D statistics are compiled for four institutional sectors of performance: business enterprise (BES), government (GOV), higher education (HES), private non-profit (PNP). These sectors are defined largely based on the SNA 93 (CEC et al., 1994), with the difference that higher education has been established as a separate sector and households have, by convention, been merged with the private non-profit (PNP) sector.

"Research and experimental development (R&D) comprise creative work undertaken on a systematic basis in order to increase the stock of knowledge, including knowledge of man, culture and society and the use of this stock of knowledge to devise new applications." (§ 57, Frascati Manual, OECD 2002 - See annex at the bottom of the page).

Eurostat statistics on R&D expenditure and personnel are compiled using the guidelines laid out in the Frascati Manual - (See annex at the bottom of the page), the "Proposed standard practice for surveys of research and experimental development", which in its latest version was published in 2002, by the OECD.

Eurostat online tables and data contain R&D expenditure and personnel broken down by different classifications.

The main breakdown of R&D statistics is by four institutional sectors of performance. These four sectors of performance are the business enterprise sector (BES), the government sector (GOV), the higher education sector (HES) and the private non-profit sector (PNP).

Gross domestic expenditure on R&D (GERD) is consequently composed of: Business enterprise expenditure on R&D (BERD), Higher Education expenditure on R&D (HERD), Government expenditure on R&D (GOVERD) and Private Non-profit expenditure on R&D (PNPRD).

Additional breakdowns of R&D expenditure available are: by source of funds, by fields of science, by type of costs, by economic activity (NACE), by size class, by type of R&D, by socio-economic objectives and by regions (NUTS 2).

Basic data on R&D personnel are expressed as head count (HC) and full time equivalents (FTE). Total national R&D personnel is obtained by summing up the R&D personnel employed in the four different sectors of performance.

Additional breakdowns of R&D personnel available are: by occupation, by qualification, by gender, by size class, by citizenship, by age groups, by fields of science, type of economic activity (NACE) and by regions (NUTS 2).

More details on breakdowns, derivations and different units available are found in the Eurostat indicators on R&D (See annex at the bottom of the page).

Main methodological references and definitions are available through methodological references (See annex at the bottom of the page).

Statistical units used in the national R&D surveys vary according to country and the sector surveyed.

In the business enterprise sector (BES) the statistical unit often consists of the legal business entity. Sometimes, if R&D activity can be subdivided, the statistical unit consists of the economic activity unit.

The statistical unit in the government sector (GOV) should according to FM recommendations (See annex at the bottom of the page) be similar to the legal entities in the BES. Examples of such statistical units are departments and establishments of the government, as well as other public bodies, institutes and non-profit organisations which are financed by central or local governments.

Legal entity is also the recommended statistical unit for the Private Non-Profit (PNP) sector, which includes non-market private institutions, households and private individuals. If such an entity has significant R&D activity in more than one field of science an attempt is made to split the entity into smaller units according to the relevant major fields.

Since a legal entity type of unit in the higher education sector (HES) often is involved in more than one major field of science a smaller statistical unit is normally appropriate. The FM (See annex at the bottom of the page) (§ 225, 2002-version) recommends as a statistical unit "the smallest homogenous unit predominantly involved in one of the six fields and for which a complete (or almost complete) set of factor input data can be obtained". Departments, faculties, colleges, university hospitals or clinics, research institutes and centres, are examples of statistical units chosen for this sector.

The population surveyed varies according to the institutional sectors. In the business enterprises sector (BES), countries usually survey, by census or by sample, businesses that are known to perform R&D or to be potential R&D performers. In the other sectors, the government sector (GOV) and the higher education sector (HES), all entities are generally surveyed.

R&D statistics are currently available for EU Member States and Candidate Countries, EFTA Countries, the Russian Federation, China, Japan, the United States and South Korea. Regional R&D statistics is normally available for EU Member States, Candidate and EFTA countries. Besides national and regional statistics Eurostat calculates and disseminates aggregates at the EU-and Euro-area-levels (EU-27, EU-15 and EA-16).

Data are normally available from 1980 onwards, but data availability might be lower for certain countries and indicators.

The base year for the unit Purchasing Power Standard (PPS) and PPS per inhabitant at constant prices is currently 2000. All calculations of non-basic unit (national currencies) are done by Eurostat.

R&D expenditure is available in the following units: National currency (MIO_NAC), Euro (MIO_EUR), Euro per inhabitant (EUR_HAB) Purchasing Power Standard (MIO_PPS), Purchasing Power Standard at 2000 prices (MIO_PPS_KP00), Purchasing Power Standard per inhabitant at constant 2000 prices (PPS_KP00_HAB), Percentage of GDP (PC_GDP).

R&D personnel data is available in full-time equivalent (FTE), in head count (HC), as a % of employment and as a % of labour force.

Reference period is the calendar year.

Until 2003 R&D data were collected under gentleman's agreement. From the reference year 2003 onwards the data collection is based on the Commission Regulation No 753/2004 on statistics on science and technology (OJ L 118, page 23 from 23 April 2004), as amended by the Commission Regulation (EC) No 973/2007 (OJL 216, page 10 from 21 August 2007) as regards the implementation of the statistical classification of the economic activities NACE Revision 2 - (See annex at the bottom of the page).

From December 2005 onwards R&D data are collected in co-operation with OECD using a common core questionnaire and two separate modules which cover each organisation specific statistical needs.

Regulation (EC) No 223/2009 on European statistics (recital 24 and Article 20(4)) of 11 March 2009 (OJ L 87, p. 164), stipulates the need to establish common principles and guidelines ensuring the confidentiality of data used for the production of European statistics and the access to those confidential data with due account for technical developments and the requirements of users in a democratic society.

Primary and secondary confidentiality of R&D data are flagged by the MSs. Confidential R&D data provided to Eurostat are treated according to the policy rules (see point 7.1).

The release calendar is publicly available and published at Eurostat website.

Eurostat website: http://epp.eurostat.ec.europa.eu

In line with the Community legal framework and the European Statistics Code of Practice Eurostat disseminates European statistics on Eurostat's website (see item 10 - 'Dissemination format') respecting professional independence and in an objective, professional and transparent manner in which all users are treated equitably. The detailed arrangements are governed by the Eurostat protocol on impartial access to Eurostat data for users.

Preliminary R&D data are disseminated in mid-November (T+11) and final R&D data - in October (T+22). Beside these major updates, possible minor data updates and revisions are announced in the release calendar.

News releases on-line

Statistics in Focus, Pocketbook 'Science, technology and innovation in Europe', Statistical book 'Science, technology and innovation in Europe'.

Please consult free data on-line.

None.

See also Eurostat's STI section website.

The national survey methods applied are available at Eurostat and can be found in R&D and GBAORD Statistics metadata collection tool (See annex at the bottom of the page).

The quality of R&D data mostly depends on the comparability of national survey methods (See annex at the bottom of the page).

In addition, the following paragraphs provide information on the main quality indicators of R&D statistics for the reference year 2007.

Eurostat must ensure that the statistical practices used to compile national R&D data are in compliance with Frascati Manual recommendations. Quality evaluation of R&D statistics is carried out based on the information provided in the national quality reports sent by the countries in addition to the regular metadata provision.

The overall quality of R&D statistics may be assessed through a comparison between the three main sectors of performance (BES, GOV and HES). The BES shows higher non-response rates and has some coverage concerns. The rate of available statistics in BES is smaller than in the other sectors (but the number of required variables is much higher). The coherence between preliminary and final results is better for BES and HES than in the GOV sector.

High degree of harmonization in concepts and methods has been achieved through the adoption of the Frascati Manual and only very few countries deviate from FM recommendations in a very small number of issues. Overall quality of R&D statistics is good and relatively improved compared to the previous quality evaluation done for reference year 2005.

R&D expenditure as a percentage of GDP (R&D intensity) has been one of the structural indicators used to underpin the Commission's analysis in the Annual Progress Report to the European Council on the progress made towards the Lisbon objectives. The 3% R&D intensity goal was retained in the new Europe 2020 strategy among the five headline targets established for EU-27 to be achieved by 2020.

Dedicated user satisfaction surveys have been carried out only in few countries. The assessment of relevance is therefore based mostly on ad-hoc feedback received from users. As such, users are rather satisfied with the available statistics in all sectors and no complaints have been recorded.

In 2007 data availability in BES is 88.1% , in GOV -92% and in HES - 90% on average. The amount of R&D statistics transmitted by the countries to Eurostat has much been improved compared to the 2005 data collection for all sectors. It should be noted that regarding BES the amount of required statistics is substantially larger for this sector. 'Missing' variables are not the same in all countries and smallest availability rates occur mainly in small countries.

The methods used for each institutional sector reflect good survey practice. All countries report that they make great efforts to prevent the appearance of errors in the data and that they carry out rigorous data validation to detect errors. In that direction electronic questionnaires have being introduced by most countries and respondents are encouraged to prefer this mode compared to the postal mode. Moreover, missing and erroneous data are re-checked in the majority of cases by re-contacting the respondents. Therefore the overall accuracy of R&D statistics is very good.

Many MSs carried out census survey in BES. Coefficients of variations (on two key variables R&D expenditure and R&D personnel in FTE) reported by some of the other countries which rely more on sampling vary significantly between MSs.

Coverage errors are of different nature for BES and the rest of R&D sectors. While in BES the quality of the frame used is of concern for the other sectors the possibility to include, exclude or double count activities are the main contributors to coverage error. However, countries report that this type of error is either negligible or very low.

While no quantitative information is available on measurement errors a substantial effort is reportedly taken by countries into minimizing them in all sectors. This includes setting high standards on interviewers' qualifications and training. It also entails drafting comprehensive and detailed notes accompanying the questionnaire, testing them before the actual data collection and providing extensive support to respondents especially when they have no previous experience with the R&D survey.

A comparison of response rates between Business, Government and Higher education sectors indicates that the response rate is both smaller and more variable in businesses. A median response rate reported for BES, GOV and HES is 86%, 98% and 96.5% respectively.

Based on the Commission Regulation 753/2004 final R&D data are provided to Eurostat 18 months after the end of the calendar year of the reference period (T + 18 months) with release after T + 22 months.

Preliminary data on R&D expenditure and total R&D personnel and researchers in FTE are provided to Eurostat 10 months after the end of the calendar year of the reference period (T + 10 months) with release after T + 11 months.

The vast majority of countries are very punctual in delivery of R&D data to Eurostat.

Although the R&D expenditures and R&D personnel data are collected by national surveys, which follow the guidelines and definitions outlined in the Frascati manual and the Regional Manual, data are not completely comparable. Differences include different survey methods and peculiarities of national R&D systems. For more information please refer to the RD comparability (See annex at the bottom of the page).

Breaks are common in R&D statistics as the nature of R&D is changing, the statistical domain is still developing fast and countries have to adapt to evolving standards. Since 1997 the number of reported breaks has been on average 2 for every country or less and it seems stable.

Comparisons between R&D statistics and relevant statistics from other domains is relevant mainly for the BE sector, for which R&D statistics are also collected through other national surveys. In most of the cases the differences are quite small. Where they are larger they are attributed to methodological differences between the R&D survey and the survey that produced the other relevant statistics.

The coherence between preliminary and final results is better for BES and HES than in the GOV sector. The average percentage difference between preliminary and final R&D data in GOV is 2.5% while in BES and HES this average is 1.9% and 1.1% respectively.

Systematic information is not available.

Sometimes data revisions are carried out by Member States and other countries when returning the completed questionnaires to Eurostat. The questionnaire requests that Member States and other countries provide information on any methodological changes affecting the data.

Data revisions are sometimes carried out by countries when returning the completed questionnaires to Eurostat.

The questionnaire requests that countries provide information on any methodological changes affecting the data.

The new EU Member States (enlargements 2004 and 2007) experienced significant changes in their R&D statistics last decade, due to the changes in economic conditions and by the introduction of new methodology compatible with the EU. Most new Member States introduced the Frascati methodology (revised in 2002) between 1994 and 1996, enabling internationally comparable analyses from the mid-1990s onwards.

Basic compilations of national R&D statistics are provided to Eurostat by EU Member States, Candidate Countries, EFTA and other European Countries. National Statistical Offices, Research Councils, and Ministries are often responsible R&D data compilations at national level. For China, Japan, the United States and South Korea the data source is the OECD's Main Science and Technology Indicators (MSTI).

The national compilations are provided in different breakdowns and in the basic units national currencies (NAC), for R&D expenditure, and number of full-time equivalents (FTE) and head count (HC), for R&D personnel.

Derived indicators and aggregates are calculated by Eurostat on the base of the data collected, sometimes together with Eurostat estimations and internal Eurostat data sources.

The data are normally collected by the national statistical institutes through sample or census surveys, from administrative registers or through a combination of sources.

There are two cycles of annual data collection: in October (T+10) for preliminary R&D data and in June (T+18) for final R&D data. Between these regular cycles MSs can also provide major updates and/or revisions of R&D data.

Data are collected from Members States, Candidate and other European countries via questionnaires (in paper or electronic format) or through administrative registers. Data for China, Japan, the United States and South Korea are provided by the OECD.

The basic national compilations of national R&D statistics are collected by Eurostat using harmonized electronic R&D data templates: the common OECD/Eurostat core questionnaire and the Eurostat module. From December 2005, all collection of national R&D statistics compilations is made in close co-operation with the OECD.

The data is collected in basic unit national currency for expenditures and in full-time equivalent (FTE) and in head count (HC) for R&D personnel. Other units, derived indicators and aggregates are based on data collected, sometimes together with relevant reference data from Eurostat data bases.

Basic R&D statistics provided by the national statistical institutes are checked for consistency and reasonableness, and compared with previously provided data before being imported in the production data base at Eurostat. Major breaks in series are normally flagged by the National Statistical Institutes before sending in the data to Eurostat. Suspected errors are reported to the National Statistical Institutes which corrects or explains the anomaly. Processed data are compared with other data sources such as the OECD before being disseminated.

Data provided to the National Statistical Institutes by respondents at the statistical unit level are normally derived from accounting or estimations made at the entity. If data is missing estimates could be made by national statistical institute according to methods deemed appropriate.

Geographical aggregates (e.g. EU27, EU15, EA16) are calculated by Eurostat as the sum of the national data expressed in a common unit. If national data is missing, estimates are made by Eurostat before summing. Estimates are normally not disseminated at national level but only used for calculation of the aggregates. Eurostat estimation techniques includes inter- and extrapolations or other simple ad-hoc methods as described below.

The data/ Eurostat estimates for R&D personnel in full-time equivalent (FTE) serve as a basis for the head count (HC) estimates, or vice versa. An FTE/HC ratio based on available data by country and sector for FTE and HC personnel. The HC and FTE estimates are then calculated by country and sector given the FTE/HC ratio and the available data in either unit. If no personnel data is available the estimation method interpolates, or extrapolates, a fixed ratio between HC, or as a second alternative FTE, and total national employment.

The calculation of the aggregates on expenditure is made from national data and Eurostat estimates in the unit Euro (from 1/1/1999) and ECU (up to 31/12/1998). Eurostat estimates are made at the level of sector of performance, interpolating or extrapolating the ratio between total R&D expenditure and GDP.

European aggregates should be seen as estimates and can sometimes deviate from what is achieved when summing up the national data. This can be due to dissemination of national data outside the normal data treatment cycles. It can also be due to possible inconsistencies in some countries data e.g. revisions of totals but not breakdowns. Within the European aggregates consistency are however always assured in such a way that breakdowns are proportionally adjusted to sum up to the total.

Geographical consistency: Please note that, while the European aggregates usually coincide with the sum of Member States figures, they are updated only at fixed intervals (normally twice a year), which means there may be a difference between the European aggregate and the appropriate sum of national data between these regular updates due to updates/revision of R&D data at country level.