1.1. Contact organisation
Statistics Sweden.
1.2. Contact organisation unit
Economic Statistics and analysis
Innovation, Business sector production and Research
1.3. Contact name
Confidential because of GDPR
1.4. Contact person function
Confidential because of GDPR
1.5. Contact mail address
Statistics Sweden
Att. Andres Quinones
ESA/NUP/INF
Solna strandväg 86, Solna
SWEDEN
1.6. Contact email address
Confidential because of GDPR
1.7. Contact phone number
Confidential because of GDPR
1.8. Contact fax number
Confidential because of GDPR
2.1. Metadata last certified
25 September 2025
2.2. Metadata last posted
25 September 2025
2.3. Metadata last update
25 September 2025
3.1. Data description
Statistics on Government Budget Allocations for R&D (GBARD) measure government support to research and development (R&D) activities, and thereby provide information about the priority governments give to different public R&D funding activities. This type of funder-based approach for reporting R&D involves identifying all the budget items that may support R&D activities and measuring or estimating their R&D content.
Main concepts and definitions used for the production of R&D statistics are given by the OECD (2015), Frascati Manual 2015: Guidelines for Collecting and Reporting Data on Research and Experimental Development, The Measurement of Scientific, Technological and Innovation Activities (FM 2015, Chapter 12), which is the internationally recognised standard methodology for collecting R&D statistics and by Eurostat’s European Business Statistics Methodological Manual on R&D Statistics (EBS Methodological Manual on R&D Statistics).
Since the beginning of 2021, the collection of R&D statistics is based on Commission Implementing Regulation (EU) No 2020/1197 of 30 July 2020.
The Regulation sets the framework for the collection of R&D statistics and specifies the main variables of interest and their breakdowns at predefined level of detail (Commission Implementing Regulation (EU) 2020/1197 of 30 July 2020 laying down technical specifications and arrangements pursuant to Regulation (EU) 2019/2152 of the European Parliament and of the Council on European business statistics repealing 10 legal acts in the field of business statistics (europa.eu)).
Statistics on science, technology and innovation were collected until the end of 2020 based on Commission Implementing Regulation (EU) No 2012/995 concerning the production and development of Community statistics on science and technology.
3.2. Classification system
Distribution by socioeconomic objectives (SEO) is based on the Nomenclature for the Analysis and Comparisons of Scientific Programmes and Budgets (NABS) at one digit level.
3.2.1. National classification
| National nomenclature of SEO used | NABS 2007 from the reference year 2021. NORDFORSK classification for previous years. |
|---|---|
| Correspondence table with NABS | Not applicable. |
3.2.2. NABS classification
| Deviations from NABS | No deviations. |
|---|---|
| Problems in identifying / separating NABS chapters and sub chapters | No specific problems known. |
| Ability to distribute Non-oriented research and General University Funds (GUF) by fields of R&D | Non-oriented research and GUF is broken down by fields of research and development (FORD). NABS12, or GUF, includes only R&D funding to HE institutions. Non-oriented research (NABS13) does not only include R&D-funding to public or private research institutions (NACE Rev.2 Division 72) but also some public R&D funding to public instituions whose primary activity is not NACE Rev.2 Division 72. |
3.3. Coverage - sector
See below.
3.3.1. General coverage
| Definition of R&D | The definition of R&D which is used is in line with the Frascati Manual (FM) definition. |
|---|---|
| Coverage of R&D or S&T in general | GBARD statistics cover only R&D. |
| Fields of R&D (FORD) covered | All fields of R&D covered. No deviations. |
| Socioeconomic objective (SEO by NABS) | All SEO covered. No deviations. |
3.3.2. Definition and coverage of government
GBARD statistics are assumed to report detailed data on all the government's budget items that may support R&D activities and to measure or estimate their R&D content. For the purposes of GBARD, the Government sector comprises (a) the central (federal) government, (b) regional (state) government and (c) local (municipal) government subsectors (FM2015, Chapter 12).
| Levels of government | Definition | Included / Not included | Comments |
|---|---|---|---|
| Central (federal) government | Central government consists of all administrative departments of the state and other central agencies whose competence extends normally over the whole economic territory, except for the administration of social security funds. This includes: |
Included | No survey is used to collect GBARD data. For National contributions for transcoordinated R&D, government research funding agencies (including other government agencies who fund R&D to the rest of the world) are surveyed. |
| Regional (state) government | Not applicable for Sweden. | Not included. | This level of government does not exist in Sweden. |
| Local (municipal) government | Local government consists primary local government units (municipalities, federations of local government authorities - mostly consisting of municipalities, or NPIs controlled by these) and secondary local government units (regions, federations of local government authorities - mostly consisting of regions, or NPIs controlled by these) | Not included |
3.4. Statistical concepts and definitions
Not requested.
3.5. Statistical unit
Budget item/appropriation.
3.6. Statistical population
See below.
3.6.1. National target population
The target population is the population for which inferences are made. The frame (or frames, as sometimes several frames are used) is a device that permits access to population units. The frame population is the set of population units, which can be accessed through the frame and the survey data really refer to this population.
| Definition of the national target population | All budget items in the central goverment budget, including final budget appropriations, that contain an element of R&D. |
|---|---|
| Estimation of the target population size | 130. |
3.7. Reference area
Not requested.
3.8. Coverage - Time
Not requested. See point 5.
3.9. Base period
Not requested. The base year for the unit Purchasing Power Standard (PPS) and PPS per inhabitant at constant prices is currently 2005. All calculations of non-basic unit (national currencies) are done by Eurostat.
Not requested.
a) Calendar year: 2023
b) Fiscal year: -
Start month: 1 January 2023
End month: 31 December 2023
6.1. Institutional Mandate - legal acts and other agreements
See below.
6.1.1. European legislation
Since the beginning of 2021, GBARD statistics are based on Commission Implementing Regulation (EU) No 2020/1197 of 30 July 2020 laying down technical specifications and arrangements pursuant to Regulation (EU) 2019/2152 of the European Parliament and of the Council on European business statistics repealing 10 legal acts in the field of business statistics. The Regulation sets the framework for the collection of R&D statistics and specifies the main variables of interest and their breakdowns at predefined level of detail. GBARD statistics were based until the end of 2020 on the Commission Implementing Regulation (EU) No 2012/995 concerning the production and development of Community statistics on science and technology.
6.1.2. National legislation
No existence of R&D specific legislation at the national level.
All statistical data collection and production of official statistics is regulated by the Official Statistics Act (2001:99) and the Official Statistics Ordinance (2001:100).
Annexes:
Official statistics act (Swedish only)
Official Statistics Ordinance (Swedish only)
6.1.3. Standards and manuals
- Frascati Manual 2015, Guidelines for Collecting and Reporting Data on Research and Experimental Development.
- EBS Methodological Manual on R&D Statistics.
6.2. Institutional Mandate - data sharing
Not requested.
7.1. Confidentiality - policy
Confidentiality, being one of the process quality components, concerns the privacy of data providers (households, enterprises, administrations and other respondents), the confidentiality of the information they provide and the extent of its use for statistical purposes.
A property of data indicating the extent to which their unauthorised disclosure could be prejudicial or harmful to the interest of the source or other relevant parties.
a) Confidentiality protection required by law: According to national law, Public Access to Information and Secrecy Act (2009:400), data may only be published in a way that no conclusions on individual units can be drawn. In practice, this does not apply to GBARD since it is based on raw data from government agencies (and therefore not protected by confidentiality according to national law) and other adminstrative data.
b) Confidentiality commitments of survey staff: The major policy in place to ensure confidentiality and prevent unauthorised disclosure of data that identify a person or economic entity is the Public Access to Information and Secrecy Act (2009:400). There are also specific conditions concerning the confidentiality of official statistics in the Official Statistics Act (2001:99).
7.2. Confidentiality - data treatment
GBARD data are publicly accessible and no confidentiality safeguards are necessary.
8.1. Release calendar
The release calendar is publicly available at Statistics Sweden's website.
8.2. Release calendar access
For Eurostat this is: Release calendar - Eurostat (europa.eu)
For Statistics Sweden this is: Release calendar - Statistics Sweden
8.3. Release policy - user access
Statistics Sweden's release policy states that all statistics must be made available to all users equally and at the same time. Statistics are always released at 8.00 am on weekdays. Users are also informed of the availability of new statistics by news releases on Statistics Sweden's website. It is possible for users to subscribe to get e-mail notifications when new statistics within a certain subject area are released. Statistics are released by being made available in the statistical database. The release policy is available on Statistics Sweden's website.
The frequency of R&D data dissemination at Eurostat level is yearly for provisional and final data.
National data is disseminated yearly.
10.1. Dissemination format - News release
See below.
10.1.1. Availability of the releases
| Availability (Y/N)1 | Content, format, links, ... | |
|---|---|---|
| Regular releases | Y | The statistical database is updated and a news release is published on Statistics Sweden's website in March and December. Link news release preliminary GBARD data 2023: Government budget allocations for R&D 2023 Link news release final GBARD data 2023: Government budget allocations for R&D 2023 |
| Ad-hoc releases | N |
1) Y - Yes, N – No
10.2. Dissemination format - Publications
See below.
10.2.1. Availability of means of dissemination
| Means of dissemination | Availability (Y/N)1 | Content, format, links, ... |
|---|---|---|
| General publication/article (paper, online) |
N | |
| Specific paper publication (paper, online) |
N |
1) Y – Yes, N - No
10.3. Dissemination format - online database
An online statistical database is available on Statistics Sweden’s website: Statistical database - Select table
10.3.1. Data tables - consultations
Not requested.
10.4. Dissemination format - microdata access
See below.
10.4.1. Provisions affecting the access
| Access rights to the information | Tables are freely available to everyone. Microdata is also freely available if users are intrested in it. Tables requiring additional processing are available for a fee. |
|---|---|
| Access cost policy | Statistics Sweden applies the principle of full cost coverage, i.e. the charge covers the actual cost of processing and producing the microdata or tables requested. |
| Micro-data anonymisation rules | Not applicable. |
10.5. Dissemination format - other
See below.
10.5.1. Metadata - consultations
Not requested.
10.5.2. Availability of other dissemination means
| Dissemination means | Availability (Y/N)1 | Micro-data / Aggregate figures | Comments |
|---|---|---|---|
| Internet: main results available on the national statistical authority’s website | Y | Only aggregate figures are available on Statistics Sweden’s website. | Data are available in the online statistical database on Statistic Sweden’s website. Main resultas available on the webiste, Government budget allocations for R&D (scb.se) |
| Data prepared for individual ad hoc requests | Y | Both microdata and aggregate figures. | Swedish National Financial Management Authority use this data as input in Central Government Accounting (NA). |
| Other |
1) Y – Yes, N - No
10.6. Documentation on methodology
The main documentation on GBARD methodology is titled Statistikens framställning - Statliga anslag till forskning och utveckling (translates to Statistical production) which is updated when new statistics are published. This documentation is only available in Swedish.
GBARD 2023 methodolgy: GBARD 2023 - Methodology report.
10.6.1. Metadata completeness - rate
Not requested.
10.7. Quality management - documentation
The main documentation on quality management is titled Kvalitetsdeklaration (translates to Quality report) and is updated when new statistcs are published. This documentation is only available in Swedish.
GBARD 2023 quality report, see this website: uf0306_kd_2023_v2.
10.7.1. Documentation and users’ requests
| Type(s) of data accompanying information available (metadata, graphs, etc.) | Statistical data is always accompanied by a quality report and a methodology report. These reports are available online on Statistics Sweden's website and follows a common standard for all official statistics in Sweden. Statistical database tables also contain footnotes in case there is important information about the data that users need to be aware of when using the data, for example explanations to changes due to a break in the time series. |
|---|---|
| Request on further clarification | Users could ask questions about reasons behind changes in the figures over time, whether it is due to a time-series break or other reasons. |
| Measure to increase clarity | Statistics Sweden works continuosly with improving the clarity in the documentation. |
| Impression of users on the clarity of the accompanying information to the data | The overall impression of users is that clarity is good |
11.1. Quality assurance
The quality management process at Statistics Sweden is described in a Quality policy. There is also a handbook on quality in official statistics which provides guidance concerning quality management and definitions and guidance on the quality criteria. The following quality criteria for official statistics are regulated by the Official Statistics Act (2001:99) and are the same as are reported in this document:
- Relevance
- Accuracy
- Timeliness
- Punctuality
- Availability and clarity
- Comparability
- Coherence
The framework for quality assurance set out in the Quality policy is a cyclic process with four steps. First is understanding legal requirements and user needs. Second is ensured processes. The third step is evaluation and analysis followed by improvement and development as the fourth step. The first step requires a good dialog with users of the statistics. One forum for such dialog is the User Council for R&D statistics. The second step is based on standardised, efficient, and secure processes which are ensured partly by automatization and digitalisation, partly by following the standardised methods, tools and processes set up for statistical production and found in Statistikproduktionsstödet (translates to the Statistical Production Guide). The third step means that the production processes continuously need to be evaluated. One way in which this done is by a yearly survey to all producers of official statics in which they evaluate the quality of the statistics produced or published during the year. Based on the results of the evaluations, decisions are made concerning which improvement and development activities are to be prioritised over the coming period, constituting the fourth and final step before the process begins again at the first step.
11.2. Quality management - assessment
The overall quality of the national GBARD methodology is considered high. It is based on international standards, including the Frascati Manual and Eurostat’s European Business Statistics Methodological Manual on R&D Statistics (EBS Methodological Manual). The method uses accessible administrative data sources, such as budget appropriations and data from national R&D surveys. These sources form the basis for estimating the share of public funding allocated to research and development.
GBARD statistics are compiled by applying coefficients—estimated using R&D survey data and budget outcomes—to initial and final budget appropriations. Initial appropriations from the national budget are used for preliminary data, while final appropriations—based on supplementary budgets—are used for final data. The coefficients are updated every two years and are never older than three years. The most recent version is always used when compiling GBARD figures. The national budget bill is also reviewed to identify additional R&D-appropriations, or other R&D-funds, not captured by the coefficients.
The GBARD methodology ensures full coverage of the target population by relying on comprehensive administrative data. Since GBARD compilation does not depend on direct survey responses, response rate issues are limited. However, the underlying R&D surveys used for coefficient estimation are subject to standard follow-up procedures to reduce non-response and improve the accuracy.
GBARD statistics are produced within Statistics Sweden’s statistical production support (SPS), a process-oriented quality framework aligned with the European Statistics Code of Practice. SPS includes standardized production processes, metadata management, documentation requirements, validation routines, and internal quality policies. The system supports regular self-assessments, compliance monitoring, and continuous improvement. Quality reports are published for all official statistics, including GBARD.
Strengths of the GBARD methodology include high international comparability, transparent documentation, full coverage of the target population, and strong alignment with international standards. Limitations include some uncertainty due to model assumptions and the use of coefficients based on earlier reference years (in case of preliminary data). Classification by socio-economic objective may also be affected by interpretation differences among reporting units. Trade-offs exist between timeliness and accuracy, particularly in preliminary statistics where coefficients are based on older data.
12.1. Relevance - User Needs
See below.
12.1.1. Needs at national level
| Users’ class1 | Description of users | Users’ needs |
|---|---|---|
| 1 | The European commission | Data used for deteemining European research policy. Comparability over time is one of the most important requirements |
| 1 | OECD | Data used for international comparisons. Comparability over time is one of the most important requirements |
| 1 | Ministries (in particual the ministry of education), research councils | Data used for national research policies. The Ministry of Education in particular also require a high degree of timeliness as the statistics are used when formulating the state budget and other policy indicators |
| 4 | Researchers and research institutes. | Specific data for further in-depth analyses of the national situation of R&D. |
1) Users' class codification
1- Institutions:
- European level: Commission (DGs, Secretariat General), Council, European Parliament, ECB, other European agencies etc.
- in Member States, at the national or regional level: Ministries of Economy or Finance, other ministries (for sectoral comparisons), National Statistical Institutes and other statistical agencies (norms, training, etc.), and
- International organisations: OECD, UN, IMF, ILO, etc.
2- Social actors: Employers’ associations, trade unions, lobbies, among others, at the European, national or regional level.
3- Media: International or regional media – specialized or for the general public – interested both in figures and analyses or comments. The media are the main channels of statistics to the general public.
4- Researchers and students (Researchers and students need statistics, analyses, ad hoc services, access to specific data.)
5- Enterprises or businesses (Either for their own market analysis, their marketing strategy (large enterprises) or because they offer consultancy services)
6- Other (User class defined for national purposes, different from the previous classes.)
12.2. Relevance - User Satisfaction
To evaluate if users' needs have been satisfied, the best way is to use user satisfaction surveys.
12.2.1. National Surveys and feedback
| Conduction of a user satisfaction survey or any other type of monitoring user satisfaction | No user satisfaction survey has been conducted. Statistics Sweden regularly arranges meetings with our primary users to take into account their suggestions for improvements. |
|---|---|
| User satisfaction survey specific for GBARD statistics | No specific user satisfaction survey for GBARD statistics has been conducted. There is, however, a specific user council for R&D statistics which also include GBARD statistics. |
| Short description of the feedback received | Overall user satisfaction is considered high. |
12.3. Completeness
See below.
12.3.1. Data completeness - rate
Mandatory data cells: 19/19
Optional data cells: 12/14
12.3.2. Completeness - overview
Completeness is assessed via comparison of the data delivered against the requirements of Commission Implementing Regulation (EU) No 2020/1197.
| 5 (Very Good) |
4 (Good) |
3 (Satisfactory) |
2 (Poor) |
1 (Very poor) |
Reasons for missing cells | |
|---|---|---|---|---|---|---|
| Provisional budget statistics1 | x | |||||
| Obligatory final budget statistics1 | x | |||||
| Optional final budget statistics2 | x | Breakdown of GBARD by funding mode is not available at the moment, and therefore not provided since it is optional. |
1) Criteria: Obligatory data (provisional budget and final budget). Only 'Very Good' = 100% and 'Very Poor' <100% apply.
2) Criteria: Optional data (final budget). 'Very Good' = 100%; 'Good' = >75%;'Satisfactory' 50 to 75%%; 'Poor' 25 to 50%; 'Very Poor' 0 to 25%.
12.3.3. Data availability
See below.
12.3.3.1. Data availability – Provisional data
| Availability1 | Frequency of data collection | Gap years – years with missing data | Time of compilation (T+x)2 | Comments | |
|---|---|---|---|---|---|
| Total GBARD | Y-1980 | Annual | T-8 | ||
| NABS Chapter level | Y-1981 | Annual | T-8 | ||
| NABS Sub-chapter level | Y-2004 | Annual | T-8 | ||
| Special categories - Biotech | Y-1999 | 1999 and 2001 | |||
| Special categories - Nanotech | N | ||||
| Special categories - Security | N |
1) Availability of the data: N: No, data are not available, Y: Yes, data are available + start year.
2) Time of compilation: T is assumed to represent the end of reference period, x expresses the number of months after (positive) or before (negative) T when data is compiled
12.3.3.2. Data availability – Final data
| Availability1 | Frequency of data collection | Gap years – years with missing data | Time of compilation (T+x)2 | Comments | |
|---|---|---|---|---|---|
| Total GBARD | Y-1980 | Annual | T+12 | Until the reference year 2022, final GBARD data are assumed not to differ substantially from preliminary figures. Therefore, final data provisional data did not differ. |
|
| NABS Chapter level | Y-1981 | Annual | T+12 | ||
| NABS Sub-chapter level | Y-2004 | Annual | T+12 | ||
| Special categories - Biotech | |||||
| Special categories - Nanotech | |||||
| Special categories - Security |
1) Availability of the data: N: No, data are not available, Y: Yes, data are available + start year.
2) Time of compilation: T is assumed to represent the end of reference period, x expresses the number of months after (positive) or before (negative) T when data is compiled
12.3.3.3. Data availability – Other special categories
| Special categories | Stage1 | Availability1 | Frequency of data colletion | Gap years – years with missing data | Time of compilation (T+x)3 | Comments |
|---|---|---|---|---|---|---|
| GBARD by funding ministry | P, F | Y-2010 | Annual | T-8 | For Final GBARD the time of compilation is T+12 | |
1) Stage: P - provisional, F - final.
2) Availability of the data: No, data are not available, Y: Yes, data are available + start year.
3) Time of compilation: T is assumed to represent the end of reference period, x expresses the number of months after (positive) or before (negative) T when data is compiled
13.1. Accuracy - overall
Accuracy in the statistical sense denotes the closeness of computations or estimates to the exact or true values. Statistics are not equal with the true values because of variability (the statistics change from implementation to implementation of the survey due to random effects) and bias (the average of the possible values of the statistics from implementation to implementation is not equal to the true value due to systematic effects).
Several types of statistical errors occur during the survey process. The following typology of errors has been adopted:
- Sampling errors. These only affect sample surveys. They are due to the fact that only a subset of the population, usually randomly selected, is enumerated.
- Non-sampling errors. Non-sampling errors affect sample surveys and complete enumerations alike and comprise:
-
- Coverage errors,
- Measurement errors,
- Non response errors and
- Processing errors.
Model assumption errors should be treated under the heading of the respective error they are trying to reduce.
13.1.1. Accuracy - Overall by 'Types of Error'
| Sampling errors | Non-sampling errors1) | Model-assumption Errors1) | Perceived direction of the error2) | |||
|---|---|---|---|---|---|---|
| Coverage errors | Measurement errors | Processing errors | Non response errors | |||
| - | 4 | 3 | 2 | - | 1 | +/- |
- Ranking of the type(s) of errors that result in over/under-estimation, from the most important source of error (1) to the least important source of error (5) In the event that errors of a particular type do not exist, is used the sign ‘:‘.
- The perceived direction of the ‘overall’ error using the signs “+” for over estimation, “-” for under estimation and “+/-” when assumption of the direction of the error cannot be made for GBARD.
13.1.2. Assessment of the accuracy
| Indicators | 5 (Very Good)1 | 4 (Good)2 | 3 (Satisfactory)3 | 2 (Poor)4 | 1 (Very poor)5 |
|---|---|---|---|---|---|
| GBARD | X | ||||
| National public funding to transnationally coordinated R & D | X |
1) High level of coverage (At least all national or federal ministries and the ministries and agencies responsible for R&D funding at state or regional level). High rate of response (>90%) in data collection. All figures broken down by NABS.
2) If at least one out of the three criteria described above would not be fully met.
3) In the event that the rate of response would be lower than 80% even by meeting the two remaining criteria.
4) In the event that the average rate of response would be lower than 70% and at least one of the two remaining criteria would not be met.
5) If all the three criteria described above are not met.
13.2. Sampling error
Not requested.
13.2.1. Sampling error - indicators
Not requested.
13.3. Non-sampling error
Non-sampling errors occur in all phases of a survey. They add to the sampling errors (if present) and contribute to decreasing overall accuracy. It is important to assess their relative weight in the total error and devote appropriate resources for their control and assessment.
13.3.1. Coverage error
Coverage errors are due to divergences between the target population and the frame population. The frame population is the set of target population members that has a chance to be selected into the survey sample. It is a listing of all items in the population from which the sample is drawn that contains contact details as well as sufficient information to perform stratification and sampling.
a) Description/assessment of coverage errors: The frame population is defined as all appropriations listed in the adopted national state (central government) budget, including final appropriations from supplementary budgets. By including all relevant budget items—whether directly or indirectly R&D-related—the frame ensures complete coverage of the central government target population, without any gaps that could affect the reliability of the statistics.
Appropriations from local government are not included in the GBARD frame, as their contribution to R&D funding is considered negligible and relevant data are not available through standard budgetary procedures. This exclusion is in line with FM2015 §12.5 and does not introduce significant bias.
In the context of appropriations related to transnationally coordinated R&D, the frame includes government research funders (agencies) that are expected to allocate resources to projects involving international collaboration. Since these funds originate from the state budget (internal funds), the frame fully captures the relevant units.
No over-coverage is present, as only institutions (government agencies, reseach institutions) with budgetary appropriations are included. Potential under-coverage from excluded local government units is assessed to be minimal and not statistically significant.
b) Measures taken to reduce their effect: To reduce the effect of coverage errors, the following measures are taken:
- The statistical frame is built directly from official budget documents to ensure completeness.
- All relevant government research funders are included based on their likelihood to allocate resources to transnationally coordinated R&D.
- Administrative sources are used to validate and supplement the data.
- The exclusion of local government units is reviewed periodically to ensure continued relevance.
13.3.1.1. Over-coverage - rate
None.
13.3.1.2. Common units - proportion
Not requested.
13.3.2. Measurement error
Measurement errors occur during data collection and generate bias by recording values different than the true ones. The survey questionnaire used for data collection may have led to the recording of wrong values.
a) Description/assessment of measurement errors: Measurement errors in the estimation of GBARD primarily stem from the derivation of coefficients based on administrative data and R&D survey data from the government and higher education sectors. These coefficients are sensitive to the completeness, accuracy, and response rates of the underlying survey data. High levels of unit- or item-non-response can reduce reliability, although such issues are assessed to be limited.
Administrative sources used in the GBARD compilation consist of open data on appropriations and actual expenditures, published by the Ministry of Finance and the Swedish National Financial Management Authority (ESV). While minor inaccuracies or omissions may occur, there are no known indications of such errors, and their potential impact on GBARD estimates is assessed as negligible.
To address incomplete coverage of R&D appropriations, supplementary analysis of budget bills and budget documents is conducted. This process may introduce interpretation-related uncertainties, particularly when identifying the primary objectives of the R&D funds. Nonetheless, these risks are judged to have minimal influence on the accuracy of the statistics.
Double-counting of public R&D funding is not considered a risk in the GBARD compilation. Each appropriation is counted only once, and the method avoids duplication.
Model-based assumptions are used to estimate both the share of each budget appropriation allocated to R&D (R&D coefficients) and the distribution of those R&D funds across socio-economic objectives (SEO coefficients). For appropriations to government agencies, R&D coefficients are derived from the government sector R&D survey, where central government agencies report their R&D expenditures by budget item. It is assumed that the proportion of R&D spending linked to a specific appropriation in the latest available survey also applies to the reference year. SEO coefficients are based on how the agencies have classified their R&D expenditures by socio-economic objective. In some cases, agencies may have misunderstood the classification guidelines and reported based on the content of the research rather than its intended purpose. Since SEO coefficients are used as a proxy to allocate GBARD by socio-economic objective, this may lead to a mismatch between the actual policy intent of the budget appropriation and its statistical classification. However, such discrepancies are considered minor and have a negligible impact on the overall reliability of the GBARD statistics.
For appropriations to higher education institutions, such as general university funds (GUF), the share allocated to research is assumed to follow the proportions (R&D-coeffcients) estimated in the higher education R&D survey. The distribution of these R&D funds, GBARD for the general advancement of knowldge, across subcategories is assumed to reflect how institutions have reported their GUF-funded research by fields of R&D. These assumptions introduce some measurement error, particularly when survey data do not reflect current budget priorities or when the purpose of funding is ambiguous. The overall impact on the quality and reliability of GBARD statistics is assessed as low.
Measurement errors may also arise during direct data collection for transnationally coordinated R&D, which uses electronic questionnaires. Despite built-in validation and concept clarification, misinterpretation or difficulty in providing precise figures can occur. These errors are considered limited in their effect.
b) Measures taken to reduce their effect:
To reduce the impact of measurement errors, the following measures are applied:
- R&D and SEO coefficients are regularly reviewed and updated based on the latest available data from the official R&D surveys for government agencies and higher education institutions.
- For transnationally coordinated R&D, electronic questionnaires are designed with built-in validation checks and clearly defined concepts to guide respondents and reduce the risk of incorrect entries. Definitions and examples are embedded directly in the questionnaire to reduce ambiguity and improve the accuracy of responses.
13.3.3. Non response error
Non response errors: occur when a survey failed to collect data on all survey variables from all the population units designated for data collection in a sample or complete enumeration.
a) Problems in obtaining data from targeted information providers: Not applicable for GBARD. The survey used to collect data on national contributions to transnationally coordinated R&D achieved a full response rate, with all targeted providers submitting complete information.
b) Measures taken to reduce their effect: Not applicable.
c) Effect of non-response errors on the produced statistics: Not applicable.
13.3.3.1. Unit non-response - rate
Not requested.
13.3.3.2. Item non-response - rate
Not requested.
13.3.4. Processing error
Between data collection and the beginning of statistical analysis, data must undergo a certain processing: coding, data entry, data editing, imputation, etc. Errors introduced at these stages are called processing errors. Data editing identifies inconsistencies or errors in the data.
a) Data processing and editing processes: Data from multiple sources—including administrative budget data and survey data—are integrated through a structured and largely automated workflow. This includes calculations and plausibility checks at the appropriation level (i.e. per budget item or appropriation number), as well as the application of R&D and SEO coefficients to estimate GBARD allocations. In some cases, coefficients that significantly affect the results are manually reviewed and, if necessary, adjusted before being processed.
b) Description of errors: Processing errors can occur when data from different sources are merged, for example due to formatting issues or mismatches between datasets. There is also a small risk of mistakes during manual preparations or reviews. These types of errors are considered unlikely and have a negligible effect on the quality of the GBARD statistics.
c) Measures taken to reduce their effect: The processing workflow is designed to minimize human error through automation and built-in validation steps. Manual inputs are limited, and plausibility checks are applied to detect inconsistencies before final compilation.
13.3.5. Model assumption error
Model assumption errors occur when the assumptions made for the estimation of parameters, models, the testing of statistical hypotheses, etc., are violated. As a result, the quality of the resulting statistics is affected (e.g. degrees of confidence might be inflated).
Description/assessment: Not applicable. Se 13.3.2 Measurement error.
14.1. Timeliness
Timeliness and punctuality refer to time and dates, but in a different manner: the timeliness of statistics reflects the length of time between their availability and the event or phenomenon they describe. Punctuality refers to the time lag between the release date of the data and the target date on which they should have been delivered, with reference to dates announced in the official release calendar.
14.1.1. Time lag - first result
Date of first release of national data: 30 March 2023
14.1.2. Time lag - final result
Date of first release of national data: 5 December 2024
14.2. Punctuality
Punctuality refers to the time lag between the release date of data and the target date on which they were scheduled for release as announced officially.
14.2.1. Punctuality - delivery and publication
Punctuality of time schedule of data release = (Actual date of the data release) - (Scheduled date of the data release)
14.2.1.1. Deadline and date of data transmission
| Transmission of provisional data | Transmission of final data | |
|---|---|---|
| Legally defined deadline of data transmission (T+_ months) | 6 | 12 |
| Actual date of transmission of the data (T+x months) | T-8 | T+12 |
| Delay (days) | 0 | 0 |
| Reasoning for delay |
15.1. Comparability - geographical
See below.
15.1.1. Asymmetry for mirror flow statistics - coefficient
Not requested.
15.1.2. Survey Concepts Issues
The following table lists a number of key survey concepts and conceptual issues; it gives reference to the Commission Regulation No 2020/1197, Frascati manual and the EBS Methodological Manual on R&D Statistics paragraphs with recommendations about these concepts / issues.
| Concept / Issue | Reference to recommendations | Deviation from recommendations | National definition / Treatment / Deviations from recommendations |
|---|---|---|---|
| Research and development | FM2015 Chapter 2 (mainly paragraphs 2.3 and 2.4). | No | The definition of R&D follows the Frascati Manual. No deviations. |
| Coverage of levels of government | FM2015, §12.5 to 12.9 | No | GBARD statistics in Sweden cover central government budget funds in accordance with the Frascati Manual (§12.5–12.9). Local government budget funds are not included, as their contribution is considered insignificant and data are not available. Regional government is not applicable in the Swedish context. No deviations. |
| Socioeconomic objectives coverage and breakdown | Reg. 2020/1197: Annex 1, Table 20 | No | GBARD is broken down by socioeconomic objectives according to NABS. No deviations. |
| Reference period | Reg. 2020/1197: Annex 1, Table 20 | No | The reference period is the calendar year, in line with the regulation. No deviations. |
15.1.3. Deviations from recommendations
GBARD encompass all spending allocations met from sources of government revenue foreseen within the budget, such as taxation. Spending allocations by extra-budgetary government entities are within the scope only to the extent that their funds are allocated through the budgetary process (FM2015 §12.9). The following table lists a number of key methodological issues, which may affect the international comparability of national GBARD statistics.
| Methodological issues | Reference to recommendations | Deviation from recommendations | National definition / Treatment / Deviations from recommendations |
|---|---|---|---|
| Definition of GBARD | FM § 12.9 | No | GBARD includes central government budget appropriations for R&D, funded by taxation and other public revenue. |
| Stages of data collection | FM2015 §12.41 | No | Budgetary data are collected from administrative budget sources at initial and final budget appropriation levels (stages 4 and 5), both of which are in line with Frascati’s recommended stages for GBARD. |
| Gross / net approach, net principle | FM2015 §12.20 and 12.21 | No | Net principle is applied. Direct EU funding is excluded as it does not pass through the budget. |
| EU/other funds | Eurostat's EBS Methodological Manual on R&D Statistics | No | Only national co-financing to EU programmes is included. Direct EU contributions (e.g. Horizon Europe) are excluded, in line with EBS Manual’s guidance that GBARD should reflect national budgetary effort. |
| Types of expenditure | FM2015 §12.15 to 12.18 | No | All types of expenditure are included. GBARD covers all government-financed R&D, regardless of performing sector, including abroad. Only budgeted appropriations from central government are counted, including transfers via intermediary agencies. Double counting is avoided since only budgeted appropriations are included. |
| Current and capital expenditure | FM §12.15 | No | Both current and capital expenditure are included in GBARD. This reflects the structure of Swedish budget appropriations, which do not distinguish between the two. |
| Extra budgetary funds | FM §12.8, 12.20, 12.38 | No | GBARD only covers funds that are part of the government budget. Agencies' own revenues (if applicable) or off-budget funds are excluded. |
| Loans | FM §12.31, 12.32, 12.34 | No | Loans are generally not a type of R&D support in the budget, and therefore not likely to be included in GBARD totals. No distinction is made between repayable and non-repayable loans. |
| Indirect funding, tax rebates, etc. | FM §12.31 - 12.38 | No | Indirect funding such as tax incentives is not included since it is not budgeted expenditures (appropriations), and therefore not separately reported. |
| Treatment of multi-annual projects | FM2015 §12.44 | No | Only budgeted funds for the reference year are included. Multi-annual projects follow annual allocations. |
| Treatment of GBARD going to R&D abroad | FM2015 §12.19 | No | GBARD includes funding to the rest of the world, contributions to international R&D programmes and organisations (e.g. Horizon Europe, ESA, OECD). |
| Criterion for distribution by socioeconomic objective | FM2015 §12.50 to 12.71 | No | GBARD is distributed according to the primary purpose of the funding, in line with Frascati’s recommendation to classify by objective rather than content or performer. Classification follows NABS 2007. |
| Method of identification of primary objective | Eurostat's EBS Methodological Manual on R&D Statistics, topic 2, statement B.6 | Yes | The primary objective is identified using coefficients from GOV and HES surveys (Topic 2, statement B.2). Where coefficients are not available, expert judgement and budget documentation are used to determine the funder’s intended purpose. |
15.2. Comparability - over time
See below.
15.2.1. Length of comparable time series
See below.
15.2.2. Breaks in time series
| Length of comparable time series | Break years1 | Nature of the breaks | |
|---|---|---|---|
| Provisional data | GBARD: 1998-2022 | 2023 | 2023: Starting with the reference year 2023, a new method for GBARD has been implemented, creating a break in the time series. Under the new approach, total GBARD is no longer obtained through direct data collection. Instead, it is estimated by applying model‑based coefficients to budget appropriations. |
| Final data | GBARD: 1998-2022 National public funding to transnationally coordinated R&D: 2012-2020, 2021-2022 |
2021, 2023 | 2021: For variable National public funding to transnationally coordinated R&D - The Swedish National Space Agency’s R&D funding to ESA programmes is now reported under national contributions to Europe-wide transnational public R&D programmes, reflecting a shift from its previous classification under bilateral or multilateral public R&D contributions. 2023: For variable GBARD - Starting with the reference year 2023, a new method for GBARD has been implemented, creating a break in the time series. Under the new approach, total GBARD is no longer obtained through direct data collection. Instead, it is estimated by applying model‑based coefficients to budget appropriations. 2023: For variable GBARD - A break in the GBARD time series also occurs due to a reclassification of certain government budget funds. These funds, initially received by universities with medical faculties, were previously reported as government budget allocations for the general advancement of knowledge financed by public general university funds (GUF) - NABS12. As the funds are ultimately transferred to the regions and used for the performance or funding of clinical research within regional health care, they are now reported under the socioeconomic objective Health (NABS07). 2023: For variable National public funding to transnationally coordinated R&D - The increase in reported contributions to bilateral or multilateral public R&D programmes is due to the Swedish Research Council’s new reporting of funding to European Research Infrastructure Consortiums (ERIC), notably to the European Spallation Source. As no such contributions were reported in previous years, this marks a break in the time series.
|
1) Breaks years are years for which data are not fully comparable to the previous period.
15.3. Coherence - cross domain
In Sweden, GBARD is based on central government budget appropriations, while government-financed GERD reflects actual R&D expenditures reported by performers. The same HES-based coefficients are used to estimate the R&D share of General University Funds (GUF) in both GBARD and HERD, making this component broadly comparable.
GBARD includes payments to foreign performers and international organisations, while GERD only covers R&D performed by resident units. GERD includes R&D financed by all levels of government (central and local government), whereas GBARD excludes local government. In some cases, appropriations may not be fully realised, or the timing of R&D activities may differ from the year in which funds are allocated, which can affect comparability.
15.3.1. Coherence - sub annual and annual statistics
Not requested.
15.3.2. Coherence - National Accounts
Not requested.
15.4. Coherence - internal
This part compares GBARD statistics from the provisional and final budget for the reference year.
15.4.1. Comparison between provisional and final data according to NABS 2007
| R&D allocations in the provisional budget delivered at T+6 | R&D allocations in the final budget delivered at T+12 | Difference (of final data) | |
|---|---|---|---|
| Exploration and exploitation of the Earth | 278 167 | 544 649 | 266 482 |
| Environment | 1 346 690 | 1 114 291 | -232 398 |
| Exploration and exploitation of space | 950 261 | 1 166 731 | 216 470 |
| Transport, telecommunication and other infrastructures | 2 893 818 | 2 031 082 | -862 736 |
| Energy | 1 021 768 | 2 648 359 | 1 626 591 |
| Industrial production and technology | 1 986 642 | 834 920 | -1 151 721 |
| Health | 1 413 833 | 3 269 534 | 1 855 701 |
| Agriculture | 801 020 | 566 945 | -234 075 |
| Education | 316 983 | 382 691 | 65 709 |
| Culture, recreation, religion and mass media | 84 643 | 119 191 | 34 548 |
| Political and social systems, structures and processes | 847 668 | 575 251 | -272 417 |
| General advancement of knowledge: R&D financed from General University Funds (GUF) | 21 477 232 | 19 360 726 | -2 116 506 |
| General advancement of knowledge: R&D financed from other sources than GUF | 11 522 165 | 10 843 932 | -678 233 |
| Defence | 1 877 064 | 1 403 056 | -474 008 |
| TOTAL GBARD | 46 817 952 | 44 861 359 | -1 956 593 |
The assessment of costs associated with a statistical product is a rather complicated task since there must exist a mechanism for appointing portions of shared costs (for instance shared IT resources and dissemination channels) and overheads (office space, utility bills etc). The assessment must become detailed and clear enough so that international comparisons among agencies of different structures are feasible.
16.1. Costs summary
| Costs for the statistical authority (in national currency) | % sub-contracted1) | |
|---|---|---|
| Staff costs | Not available. | |
| Data collection costs | Not applicable. | |
| Other costs | Not available. | |
| Total costs | Not available. | |
| Comments on costs | ||
1) The shares of the figures given in the first column that are accounted for by payments to private firms or other Government agencies.
16.2. Components of burden and description of how these estimates were reached
| Value | Computation method | |
|---|---|---|
| Number of Respondents (R) | Not applicable. | |
| Average Time required to complete the questionnaire in hours (T)1 | Not applicable. | |
| Average hourly cost (in national currency) of a respondent (C) | Not applicable. | |
| Total cost | Not applicable. |
1) T = the time required to provide the information, including time spent assembling information prior to completing a form or taking part in interview and the time taken up by any subsequent contacts after receipt of the questionnaire (‘Re-contact time’)
17.1. Data revision - policy
Statistics Sweden's revision policy covers three types of revisions: planned and reoccuring revisions, revisions due to conceptual and/or methodological changes and corrections.
- Planned and reoccuring revisions - In order to accommodate user timeliness needs, Statistics Sweden publisch preliminary figures. These figures are then revised once or several times before final data are released. In case of planned revisions, users will be informed of the number of revisions as well as the revision dates.
- Revisions due to conceptual and/or methodological changes - Methodological changes can have systematic effects on the statistics. Concepts, definitions or classifications can be changed in order to better capture the target variables. In case of such changes, and if deemed necessary and possible, revisions of earlier final data can be made in order to produce comparable time series. Users will be informed of revisions of this kind in advance, with an explanation of why the revision is necessary.
- Corrections - In case of errors in published data, corrections can be made. When an error has been identified, the need for correction is evaluated without delay based on the magnitude of the error and the importance of the statistics. Corrections are always published in a clear and easily accessible manner, with information on why the correction is necessary.
17.2. Data revision - practice
Not requested.
17.2.1. Data revision - average size
Not requested.
18.1. Source data
a) Provisional data:
- Initial budget appropriations – Provided by the Ministry of Finance.
- R&D-coefficients and SEO coefficients, per budget item, derived from the most recent Government R&D Survey (GOV) and Higher Education R&D Survey (HES).
- Budget Bill (main budget document) with budget proposals for the reference year – Used together with coefficients to identify additional R&D‑related budget items and determine their primary objectives.
b) Final data:
- Final budget appropriations from supplementary budgets – Obtained through open data published by the Swedish National Financial Management Authority.
- Budget documents detailing proposals for supplementary budgets.
- Microdata from the latest Government R&D Survey (GOV) – Used to derive R&D-coefficients and SEO coefficients for appropriations to central government agencies (from central government).
- Appropriation outcomes (expenditures) - Obtained through open data published by the Swedish National Financial Management Authority. Used to derive R&D-coefficients and SEO coefficients for appropriations to central government agencies.
- Microdata from latest Higher Education R&D Survey (HES) – Used to derive SEO coefficients to distribute R&D funds in GUF (NABS12) by subcategories.
- For National public funding to transnationally coordinated R&D: Direct data collection.
c) General University Funds (GUF): GUF are directly identifiable in the budget appropriations through their respective budget items.
18.2. Frequency of data collection
See 12.3.3.
18.3. Data collection
See below.
18.3.1. Data collection overview
| Provisional data | Final data | Comments | |
|---|---|---|---|
| Data collection method | Provisional GBARD estimates are based on initial budget appropriations per budget item, provided by the Ministry of Finance for the reference year. These appropriations are combined with R&D and SEO coefficients derived from the most recent Government R&D Survey (GOV) and Higher Education R&D Survey (HES) to produce aggregate estimates of total GBARD and its distribution across socioeconomic objectives. The Budget Bill is also used to identify additional R&D-relevant appropriations and determine their primary objectives. See “Final data” section for methodological details |
Data for GBARD are collected from administrative sources and survey microdata. Final budget appropriations and expenditure outcomes per budget item are retrieved from open data published by the Swedish National Financial Management Authority. Expenditure outcomes are used in combination with microdata from the Government R&D Survey (GOV) to estimate R&D coefficients, which are then applied to appropriations to government agencies to calculate total GBARD. SEO coefficients—also derived from GOV microdata—are applied to GBARD estimates at the appropriation level to allocate funding by socioeconomic objective. For general university funds (GUF), institution-specific R&D coefficients—previously established through the time-use survey component of the Higher Education R&D Survey (HES)—are applied to each university’s GUF appropriation to estimate GBARD. Additional microdata from HES are used to derive allocation coefficients for GUF-funded R&D, classified under NABS 12, in order to distribute it across its six subcategories (NABS 121–126). Data on national public funding to transnationally coordinated R&D are collected directly from relevant government agencies via electronic questionnaires. |
|
| Stage of data collection | Initial budget appropriations—referring to the amounts approved by the legislature for the upcoming fiscal year, including any amendments introduced during the parliamentary debate—are used in the compilation of provisional GBARD (Stage IV). | Final budget appropriations—reflecting the figures voted by parliament for the fiscal year, including any additional allocations adopted through supplementary budgets during the year—are used in the compilation of final GBARD (Stage V). | |
| Reporting units | For National public funding to transnationally coordinated R&D: Institutional unitis in the form of Central government agencies. | Those included are entities known or assumed to fund R&D projects abroad, including contributions to transnationally coordinated R&D programmes. | |
| Basic variable | R&D appropriations based on initital budget appropriations. | R&D appropriations based on final budget appropriations. | |
| Time of data collection (T+x)1) | T-12 | T+10 | Provisional data: Initial budget appropriations are typically available in December of the year preceding the reference year. The relevant coefficients are already available, as they are updated biennially in connection with the compilation of final GBARD data for odd reference years. Final data: Final budget appropriations are available at T+3 but are retrieved at T+10. This timing corresponds with the derivation of coefficients used to compile final GBARD data for odd reference years, following the publication of final R&D statistics. For even reference years, final GBARD data are also compiled at T+10, although technically possible at T+4. This later timing is chosen to align with the data collection on transnationally coordinated public R&D, which takes place at T+10—when the list of Europe-wide programmes active during the reference year becomes available. |
| Problems in the translation of budget items | No problems. | ||
1) Time of data collection (T+x): T is assumed to represent the end of reference period. x expresses the number of months after (positive) or before (negative) T when data is collected.
18.3.2. General University Funds (GUF)
The R&D share of General University Funds (GUF) is estimated using institution-specific coefficients. These coefficients—previously estimated through the time-use survey component of the Higher Education R&D Survey (HES)—are applied to each institution’s GUF appropriation to calculate its R&D content. The coefficients remain valid until updated through a new time-use survey. No direct data collection is carried out for GUF in the GBARD context
18.3.3. Distribution by socioeconomic objectives (SEO)
| Level of distribution of budgetary items – institution or programme/project | Institutional level. GBARD estimates are based on budget appropriations (central government) to central government agencies and higher education institutions. Distribution and classification are carried out at the level of institutional funding, not individual programmes or projects. |
|---|---|
| Criterion of distribution – purpose or content | Distribution by socioeconomic objective is based on the purpose of the underlying budget appropriations, in line with Frascati Manual §12.50–12.52. GBARD is allocated to SEO categories according to the intended purpose of each appropriation, using coefficients derived from the Government R&D Survey (GOV) and the Higher Education R&D Survey (HES). Content-based classification is not applied. |
| Method of identification of primary objectives | In line with Frascati Manual §12.53, the allocation of GBARD to socioeconomic objectives is carried out at the institutional level, reflecting the funder’s intended purpose. Due to practical constraints, programme- or project-level information is not used. Instead, coefficients derived from the Government R&D Survey (GOV) and the Higher Education R&D Survey (HES) are applied to estimate the distribution of GBARD across SEO categories. These coefficients serve as proxies for the primary objectives of the underlying appropriations |
| Difficulties of distribution | One key difficulty in distributing GBARD across socioeconomic objectives is that it is not practically feasible to identify the specific purpose of each individual research project or programme. Instead, classification is done at the institutional level using coefficients derived from the Government R&D Survey (GOV) and the Higher Education R&D Survey (HES). Since these coefficients are based on historical data, they assume that funding patterns remain stable over time, which may overlook recent changes in policy direction. As a result, the method may not fully reflect the diversity of R&D activities within each appropriation |
18.3.4. Questionnaire and other documents
| Annex | Name of the file |
|---|---|
| GBARD national questionnaire and explanatory notes in English: | UF0306 Transnationell samordnad FoU 2023 |
| GBARD national questionnaire and explanatory notes in the national language: | UF0306_StaF_2023 |
| Other relevant documentation of national methodology in English: | UF0306_KD_2023 |
| Other relevant documentation of national methodology in the national language: |
Annexes:
Questionnaire National public funding to transnationally coordinated R&D (Swedish only)
GBARD 2023 methodology report
GBARD 2023 quality report
18.4. Data validation
Validation procedures are applied throughout the production of GBARD statistics to ensure accuracy and consistency. Source data are primarily based on administrative registers, including the national budget and data from the Swedish National Financial Management Authority (ESV). These are cross-checked against the Budget Bill and agency instructions to ensure full coverage of relevant appropriations.
For the model-based part of the statistics, coefficients derived from the Government R&D Survey (GOV) and the Higher Education R&D Survey (HES) are reviewed for plausibility. Appropriations with high R&D coefficients (≥0.75) undergo targeted microdata validation, including manual checks and calibration using textual analysis of budget documents. Logical consistency is verified, ensuring that the sum of distributed values matches total R&D appropriations.
Macro-level validation includes comparisons with previous years, identification of outliers, and investigation of unexpected shifts. If large deviations are found, the underlying appropriations or institutions are examined and adjusted if necessary. For the survey-based part (transnationally coordinated R&D), incoming responses are checked for unexplained deviations and followed up directly with respondents.
Before publication, all tables and figures are reviewed in SCB’s publishing system to ensure completeness, clarity, and consistency across outputs. Validation results are documented and used to refine estimation procedures and improve future data collection.
18.5. Data compilation
See below.
18.5.1. Imputation - rate
Not applicable for GBARD. No imputations performed for the survey that collect funding to transnationally coordinated R&D.
18.5.2. Data compilation methods
See below.
18.5.2.1. Identifying R&D
| Method(s) of separating R&D from non-R&D | To separate R&D from non-R&D within central government budget appropriations, a coefficient-based method is applied in line with Frascati Manual §12.13. For each budget item, an R&D coefficient is used to estimate the share attributable to R&D. These coefficients are derived from microdata collected through the Government R&D Survey (GOV). The values vary depending on the institution and budget item, ranging from near-zero to 1.0. Appropriations with high R&D shares (typically ≥0.75) may be examined more closely, but manual review and text analysis of budget documents—such as the Budget Bill and agency instructions—are used more broadly as a complement to the survey-based method to support interpretation of the appropriation’s purpose |
|---|---|
| Description of the use of the coefficient (if applicable) | R&D coefficients are applied to budget appropriations at the item level to estimate the share allocated to R&D. For each item, GBARD is calculated by multiplying the appropriation amount by the corresponding coefficient. Additional coefficients are used to distribute GBARD across socioeconomic objectives (SEO), based on reported R&D expenditure in the Government R&D Survey (GOV). |
| Coefficient estimation method | R&D coefficients are estimated using survey data from central government agencies and higher education institutions, combined with open budget data. For central government appropriations, the coefficients are calculated by matching reported R&D expenditure from the Government R&D Survey (GOV) with expenditure outcomes per budget item, published by the national financial management authority. Agencies report extramural and intramural R&D expenditure funded by, and originating from, the central government budget, broken down by individual budget item. These data are used to estimate the share of each appropriation allocated to R&D. Socioeconomic objective (SEO) distribution (extramural & intramural R&D expenditures) is also reported directly in GOV, classified according to NABS 2007, and used to derive SEO coefficients. |
| Frequency of updating of coefficients | Coefficients are updated every two years, following the biennial cycle of the GOV survey. |
18.5.2.2. General University Funds (GUF)
| Method(s) of separating R&D from non-R&D | To separate R&D from non-R&D within General University Funds (GUF), institution-specific R&D coefficients are applied. These coefficients are retrieved from previously estimated values based on a time-use survey conducted within the Higher Education R&D Survey (HES). They are used to estimate the share of GUF appropriations attributable to R&D. |
|---|---|
| Description of the use of the coefficient (if applicable) | For General University Funds (GUF) only, R&D coefficients are applied to the appropriation to estimate the share allocated to R&D. GBARD is calculated by multiplying the GUF amount by the corresponding institution-specific coefficient. The estimated R&D total is then distributed across socioeconomic objectives (NABS 121–126) using SEO coefficients, which are based on HES microdata and reflect the distribution of R&D funded by GUF by field of research (FORD) |
| Coefficient estimation method | For General University Funds (GUF), estimated R&D expenditure is distributed across socioeconomic objectives (SEO) under NABS12 using coefficients derived from HES microdata. These coefficients are based on a variable that measures R&D expenditure funded by GUF, broken down by field of research (FORD). Since all R&D in GUF is classified under NABS 12, the coefficients are used to allocate this total across its six subcategories (NABS 121–126). |
| Frequency of updating of coefficients | Coefficients are updated every two years, following the biennial cycle of the HES survey. |
18.5.2.3. Other issues
| Treatment of multi-annual programmes | Multi-annual programmes are not reported in a single year, they are allocated to the years they are budgeted. |
|---|---|
| Possibility to classify budgetary items by COFOG functions | Possible. |
| Possibility to classify budgetary items by other nomenclatures e.g. NACE | A distribution of GBARD by sector of performance receiving the funds might be possible to do but it is uncertain if the quality would be good enough. |
| Method of estimation of future budgets | No future budgets are estimated. |
18.6. Adjustment
Not requested.
18.6.1. Seasonal adjustment
Not requested.
Statistics on Government Budget Allocations for R&D (GBARD) measure government support to research and development (R&D) activities, and thereby provide information about the priority governments give to different public R&D funding activities. This type of funder-based approach for reporting R&D involves identifying all the budget items that may support R&D activities and measuring or estimating their R&D content.
Main concepts and definitions used for the production of R&D statistics are given by the OECD (2015), Frascati Manual 2015: Guidelines for Collecting and Reporting Data on Research and Experimental Development, The Measurement of Scientific, Technological and Innovation Activities (FM 2015, Chapter 12), which is the internationally recognised standard methodology for collecting R&D statistics and by Eurostat’s European Business Statistics Methodological Manual on R&D Statistics (EBS Methodological Manual on R&D Statistics).
Since the beginning of 2021, the collection of R&D statistics is based on Commission Implementing Regulation (EU) No 2020/1197 of 30 July 2020.
The Regulation sets the framework for the collection of R&D statistics and specifies the main variables of interest and their breakdowns at predefined level of detail (Commission Implementing Regulation (EU) 2020/1197 of 30 July 2020 laying down technical specifications and arrangements pursuant to Regulation (EU) 2019/2152 of the European Parliament and of the Council on European business statistics repealing 10 legal acts in the field of business statistics (europa.eu)).
Statistics on science, technology and innovation were collected until the end of 2020 based on Commission Implementing Regulation (EU) No 2012/995 concerning the production and development of Community statistics on science and technology.
25 September 2025
Not requested.
Budget item/appropriation.
See below.
Not requested.
a) Calendar year: 2023
b) Fiscal year: -
Start month: 1 January 2023
End month: 31 December 2023
Accuracy in the statistical sense denotes the closeness of computations or estimates to the exact or true values. Statistics are not equal with the true values because of variability (the statistics change from implementation to implementation of the survey due to random effects) and bias (the average of the possible values of the statistics from implementation to implementation is not equal to the true value due to systematic effects).
Several types of statistical errors occur during the survey process. The following typology of errors has been adopted:
- Sampling errors. These only affect sample surveys. They are due to the fact that only a subset of the population, usually randomly selected, is enumerated.
- Non-sampling errors. Non-sampling errors affect sample surveys and complete enumerations alike and comprise:
-
- Coverage errors,
- Measurement errors,
- Non response errors and
- Processing errors.
Model assumption errors should be treated under the heading of the respective error they are trying to reduce.
Not requested.
See below.
a) Provisional data:
- Initial budget appropriations – Provided by the Ministry of Finance.
- R&D-coefficients and SEO coefficients, per budget item, derived from the most recent Government R&D Survey (GOV) and Higher Education R&D Survey (HES).
- Budget Bill (main budget document) with budget proposals for the reference year – Used together with coefficients to identify additional R&D‑related budget items and determine their primary objectives.
b) Final data:
- Final budget appropriations from supplementary budgets – Obtained through open data published by the Swedish National Financial Management Authority.
- Budget documents detailing proposals for supplementary budgets.
- Microdata from the latest Government R&D Survey (GOV) – Used to derive R&D-coefficients and SEO coefficients for appropriations to central government agencies (from central government).
- Appropriation outcomes (expenditures) - Obtained through open data published by the Swedish National Financial Management Authority. Used to derive R&D-coefficients and SEO coefficients for appropriations to central government agencies.
- Microdata from latest Higher Education R&D Survey (HES) – Used to derive SEO coefficients to distribute R&D funds in GUF (NABS12) by subcategories.
- For National public funding to transnationally coordinated R&D: Direct data collection.
c) General University Funds (GUF): GUF are directly identifiable in the budget appropriations through their respective budget items.
The frequency of R&D data dissemination at Eurostat level is yearly for provisional and final data.
National data is disseminated yearly.
Timeliness and punctuality refer to time and dates, but in a different manner: the timeliness of statistics reflects the length of time between their availability and the event or phenomenon they describe. Punctuality refers to the time lag between the release date of the data and the target date on which they should have been delivered, with reference to dates announced in the official release calendar.
See below.
See below.
