Source: German Centre for Higher Education Research and Science Studies, calculations; OECD, Main Science snd Technology Indicators March 2026

Notes:
¹⁾ Some figures are revised, provisional, estimated, differently defined or only partly comparable with figures from previous years (see original edition of "Main Science and Technology Indicators März 2026").
²⁾ PNP: private non-profit institutions.
³⁾ Until 1990, only West German Länder. Starting with the 2014 reference year, the government survey collects data on the breakdown of R&D personnel by occupation, whereas previously this was estimated based on data by qualification level. Increases in the higher education sector from 2016 onward are also due to methodological changes resulting from an adjustment of the R&D coefficients. Starting in 2023, transition to the statistical enterprise as the unit of presentation in the business enterprise sector.
⁴⁾ Since 1993 (1992 for the business enterprise sector), the data have been based on comprehensive surveys rather than on a combination of household data and survey results.
⁵⁾ The government and higher education sector cover all fields of the natural sciences as well as social sciences and humanities, while the business enterprise sector covers only the natural sciences. Since there are few organizations in the PNP sector, no R&D survey was conducted in this sector, and no data is available. Since 2009, data on researchers have been collected in accordance with the definition of "researcher" in the Frascati Manual. Previously, this applied only to independent research institutions, while data for the other sectors were collected in accordance with the UNESCO concept of "scientists and engineers". In 2009, the scope of the survey was expanded in the business enterprise and government sectors. Prior to 2000, all personnel data in the business enterprise sector refer exclusively to large and medium-sized enterprises. Since 2000, however, the survey has covered almost all economic sectors and all enterprises above a certain threshold. In 2000 and 2004, a complete census of all enterprises was conducted, while in the intervening years, data for small enterprises were estimated. As part of the reform of the science and technology system, some government institutions were converted into enterprises, and their R&D data have been recorded in the business enterprise sector since 2000.
⁶⁾ As of 2002, the business enterprise survey specifically requests data on researchers, technicians and other personnel. Earlier data for R&D personnel by occupation are based on qualification. Break in time series in 2007 due to a methodological adjustment in data collection and a change in the administrative structures of certain institutes, which were classified under the government sector until 2006 and under the higher education sector from 2007 onwards.
⁷⁾ As of 1997, the higher education sector includes central university hospitals. From 1997 and the implementation of ISCED 97, "researchers" also includes holders of engineering degrees and graduates of vocational polytechnics, degrees which are now classified in first stage tertiary education (ISCED 5A). As of 2004, R&D personnel data are available according to occupation. Previous breakdown was by formal qualification. In 2011, a new methodology for calculating the time spent on R&D by personnel in the higher education sector was implemented. As a consequence, R&D personnel (measured in FTE) in the higher education sector decreased.
⁸⁾ Since 2014, R&D personnel at university hospitals have been better accounted for, resulting in a break in the time series for the higher education sector. The National Center for Scientific Research (CNRS) is classified under the higher education sector, whereas such institutions in other countries are classified under the government sector. The methodology for surveying public administrations was changed in 2010: The method for recording resources spent on R&D in ministries and certain public institutions was adjusted, leading to a more accurate recording of their activities. This has resulted in a decline in full-time equivalents. Since 2006, SMEs have been better accounted for, as there is no longer a minimum threshold in the business enterprise sector for inclusion in the survey. Starting in 2001, the scope of the business sector was expanded. The data reported by the Ministry of Defense was also expanded to include research work that had not been classified as R&D in previous years. In 2000, several methodological changes that improved the quality of public sector data resulted in a break in the time series for that year. In 1997, the method for recording R&D personnel in government agencies was changed. In 1991, France Télécom and GIAT Industries were transferred from the government sector to the business enterprise sector.
⁹⁾ Before 1996, higher education expenditure and personnel data in FTE are OECD estimates derived from official headcount-based data. Beginning with the survey for OECD data 2002, the supplied coefficients were applied to doctoral level students as well as teachers when calculating FTE for the higher education sector, resulting in a break in series in that year. In 2008, 2013, and 2018, the FTE coefficients for researchers in the higher education sector were revised, producing an increase in personnel for this sector and the national total in 2013 and 2018, and a decrease in 2008.
¹⁰⁾ As of 2007, the former "post-secondary colleges for teacher training" ("Pädagogische Akademien") have become "Universities of Education" and are, consequently surveyed as units of the higher education sector (until 2006, these units were covered in the government sector). Since 2009, a large unit previously omitted has been included as an R&D performer in the PNP sector. Three large units were reclassified in the government sector in 2016 (previously included in the business enterprise and the higher education sectors).
¹¹⁾ Starting in 2023, there is a break in the time series due to a methodological change in data collection, such as a switch to using the statistical enterprise as the reporting unit in the business enterprise sector. The police organization was restructured in 2015, which affected the scope of coverage for R&D personnel in the government sector. Starting in 2013, a portion of technical and other staff was reclassified as researchers; therefore, the data is only partially comparable with previous years. In 2011, the coverage of the government sector was improved. Since 2007, researchers in the business enterprise, government, and PNP sectors have been classified by occupation; until 2006, the data refer to university graduates rather than researchers. R&D data were underestimated until 2005: R&D in the government sector included only central government units, and enterprises with 10 to 49 employees were excluded from the coverage. In addition, prior to 1993, surveys in the business enterprise, government, and PNP sectors did not include R&D activities in the social sciences and humanities. Starting in 2005, the survey method in the higher education sector was changed. Starting in 2005, the government sector excludes R&D personnel from district councils.
¹²⁾ The business enterprise sector comprises private enterprises only. Public enterprises are included in the government sector. Since 2000, privatized telecommunications companies (such as Swisscom) have no longer been part of the government sector.
¹³⁾ Until 2006, excluding the social sciences and humanities.
¹⁴⁾ The United Kingdom has significantly revised its data on R&D personnel in the business enterprise sector for the years starting in 2018. These changes reflect an upward adjustment to account for underreported R&D-performing enterprises. The revised estimates for R&D personnel have not yet been included in the OECD database. Starting in 2005, the number of researchers in full-time equivalents (FTE) includes a more accurate measurement of graduate students. In 2001, the government research agency "Defense Evaluation and Research Agency" (DERA) was dissolved, and two new organizations were established; of these, only about one-quarter remained in the government sector, while the remaining three-quarters were transferred to the business enterprise sector. The reclassification of institutions explains the increase in R&D personnel in the government sector between 1991 and 1992, as well as the decline in the following year. Since no official data are available for the higher education sector, the OECD Secretariat has made estimates of the total number of researchers starting in 1999 and of the total number of R&D personnel starting in 1994.
¹⁵⁾ Since no official data is available for the various employment sectors, the total number of researchers is an OECD estimate through 2021. Starting in 2022, it is based on official personnel data available for all sectors. Eight government-funded R&D centers, accounting for one-quarter of all government-funded R&D centers, were not included because they were unable to provide data on their R&D staff. The methodology for counting researchers was changed starting in 1985: In the government, higher education, and PNP sectors, the data refer to employed researchers and engineers with doctoral degrees who list R&D or the management of R&D as their primary activity, and in the higher education sector, to the number of full-time equivalents of doctoral students with research fellowships who are estimated to spend an average of 50% of their time on R&D activities. Since 1985, military personnel have no longer been counted as researchers in the government sector.

m = data not available.

Last update: April 15, 2026