RDSurvey@oecd.org

Joint OECD-Eurostat international data collection on resources devoted to RD

Sources and Methods Databases

The R&D data used in this publication have been collected and presented in line with the standard OECD methodology for R&D statistics as laid out in the OECD “Frascati Manual” (see http//oe.cd/Frascati). The 2002 edition of the manual has now been superseded by the 2015 edition. The revised guidelines and definitions are in the course of being implemented and are not expected to change the main indicators significantly although some terminology changes will occur. This edition of RDS has been compiled in accordance with the 2015 Frascati Manual, some countries may still be following the 2002 edition for certain series or datapoints.

2005 onward.

Data are provided in million national currency (for the euro zone, pre-EMU euro or EUR), million current PPP USD and million constant USD (2015 prices and PPPs).

This table presents research and development (R&D) expenditure statistics performed in the business enterprise sector by industry according to the International Standard Industrial Classification (ISIC) revision 4.

Data in ISIC rev. 4 are presented from 2005, first year back to which some countries converted their data from ISIC rev.3.1 to the current classification. Some countries do not yet appear in this dataset, but will be included gradually as data are submitted and processed in the new classification.

This breakdown between industries is, in principle, made at the enterprise level, although some countries are able to break down R&D data for multi product enterprises between their main lines of business. National statistical regulations prevent publication of results where there are very few firms in the given category, hence the many gaps in the tables. Depending on the country, R&D institutes serving enterprises are either classified with the industry concerned, or grouped under “Research and Development” (ISIC rev.4, Division 72). When these R&D institutes are classified with the industry served, the evaluation of R&D in these industries is more accurate and more comparable between countries for the industries concerned. This results, however, in an underestimation of the percentage of BERD performed by the service sector as compared with other countries.

The Frascati Manual recommendation concerning data on R&D by industry is to report BERD on an enterprise basis (see FM2015 section 7.49). When this is interpreted strictly, all the BERD of a diversified enterprise will be allocated to the industrial class of its principal activity. In circumstances where a few large firms dominate R&D spending in several areas, this can and does lead to underestimates of R&D associated with the secondary activities of the firms. Overall, R&D is therefore overestimated for some industries and underestimated for others.

However, not all countries follow a strict enterprise basis for allocating R&D expenditures to industrial classes. Some countries make a disaggregation of the R&D of their largest, diversified firms into a number of different activities. In other countries, the enterprise approach has been abandoned and data are reported on an industry served basis. This is the reason why two classification criteria for BERD by industry are included in this view of “BERD by industry” (see the variable CLASSIFICATION CRITERIA: Main activity or industry served) depending on which approach is more closely followed by each country (only a few countries currently collect these data both ways and are therefore included according to both criteria).