Reference metadata describe statistical concepts and methodologies used for the collection and generation of data. They provide information on data quality and, since they are strongly content-oriented, assist users in interpreting the data. Reference metadata, unlike structural metadata, can be decoupled from the data.
The (rail_ec) group of tables provides data about railway enterprises by type of enterprise; and about investment and maintenance expenditures on railway vehicles or rail infrastructure.
There are no railways in Cyprus, Malta and Iceland. There is only a minor railway network in Liechtenstein, but the railway integrated company is Austrian.
The Common Questionnaire is not supported by a legal act, but is based on a goodwill agreement with the participating countries; its completeness may therefore vary from country to country. The Glossary for transport statistics was published since 1994 (1st edition) with the purpose of assisting countries during this data collection
Data is generally provided by the national statistical institutes or responsible transport administrations.
Eurobase tables based on the Common Questionnaire cover annual data from the EU Member States, EFTA countries, candidate countries and potential candidates to the EU. Data from other participating countries are available through the ITF and the UNECE websites. In total, comparable transport data collected through the Common Questionnaire are available for close to 50 countries worldwide.
In particular, “RAIL” for this purpose corresponds to activities “49.11 Passenger heavy rail transport” and “49.2: Freight rail transport” of NACE Rev. 2.1. (not “49.12 Other passenger rail transport”), as well as a part of “52.21 Service activities incidental to land transportation”.
3.3. Coverage - sector
The scope of the Common Questionnaire in this group of indicators is limited since 2013 to the number of railway enterprises by type and to investment and maintenance expenditures (on rolling stock and/or on railway infrastructure). Employment data are now to be found in “Structural business statistics (sbs)”.
3.4. Statistical concepts and definitions
The definitions used in the Common Questionnaire are specified in the Glossary for transport statistics. The Glossary is an official publication in EN, FR and DE. The UNECE produces also a version in Russian.
As for rail enterprises and economic performances, the definitions are listed between entries A.III-01 and A.III-17 in the Glossary.
3.5. Statistical unit
The statistical units observed are the national railway enterprises.
3.6. Statistical population
The railway enterprises and their economic performances are observed on an annual basis.
Only aggregated national data are reported by the participating countries; no micro data or confidential data are transmitted.
3.7. Reference area
Member States of the European Union; EFTA states; European Union Candidate countries and potential candidate(s). Other countries are covered by the ITF and UNECE.
3.8. Coverage - Time
The data are available as from:
1970: ‘Railway enterprises by type of enterprise’
1990: ‘Expenditure on rolling stock and railway infrastructure’.
3.9. Base period
Not applicable.
The ‘Railway enterprises by type of enterprise’ are measured in number of enterprises.
The ‘Expenditure on rolling stock and railway infrastructure’ are measured in million Euros.
Data are collected on a yearly basis.
6.1. Institutional Mandate - legal acts and other agreements
The Common Questionnaire is not supported by any legal acts; it is a voluntary data collection, based on a goodwill agreement with the participating countries.
6.2. Institutional Mandate - data sharing
Data are shared with the United Nations Economic Commission for Europe (UNECE) and the International Transport Forum (ITF) at the OECD.
7.1. Confidentiality - policy
The Common Questionnaire is a voluntary data collection of aggregated national data. No confidential data are reported by the participating countries. Therefore, no confidentiality policy applies.
7.2. Confidentiality - data treatment
Not applicable.
8.1. Release calendar
The intention of Eurostat is to publish all data on reference year N by the end of year N+1. The data release calendar for transport domain is available on the Eurostat website
Datasets checked for consistency are available free of charge through Eurostat’s Eurobase dissemination database.
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 - 'Accessibility and clarity') 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.
Annually. An intermediate update can occur 1 or 2 months after first release when late data are transmitted, or when errors are corrected.
10.1. Dissemination format - News release
Not applicable for this group of tables.
10.2. Dissemination format - Publications
Not applicable for this group of tables.
10.3. Dissemination format - online database
Datasets checked for consistency are available free of charge through Eurostat’s Eurobase dissemination database, under the theme ‘Transport’, node 'Railway transport', folder 'Railway transport - enterprises, economic performances and employment (rail_ec)'. Only aggregated annual data at national level are available.
Similar Common Questionnaire transport data for countries other than EU/EFTA/candidate countries and potential candidates are available from ITF and UNECE.
10.4. Dissemination format - microdata access
Not applicable; only aggregated national data are reported.
10.5. Dissemination format - other
Not applicable.
10.6. Documentation on methodology
The Glossary for transport statistics provides the definitions and lists the classifications used for the Common Questionnaire data collection. The Glossary is available on the websites of the three organisations.
10.7. Quality management - documentation
Not available.
11.1. Quality assurance
The Common Questionnaire is a voluntary data collection, not supported by legal acts. It is completed by the competent national authorities, who are responsible for the main quality assurance. National data are reported through EDAMIS portal. Upon transmission of a file, a validation process takes place in EDAMIS which contains a number of data checks and controls in two validation systems:
EDAMIS notifies the file sender if errors are found in the overall structure of transmitted Common Questionnaire file. This validation system is called Struval (structural validation).
EDAMIS notifies the file sender if errors are found in the content/data of the file (mainly differences between totals and the sum of their breakdown values). This validation system is called Conval (content validation).
All files transmitted to EDAMIS are first validated by Struval and then by Conval. In case a file fails to pass the level of Struval, it does not reach Conval. The data provider is notified about the error(s) at Struval level and needs to resubmit a corrected file. Similarly, if a file succeeds in passing Struval but is rejected at Conval level, the data provider is again notified about the error(s) and has to submit a corrected file.
Further plausibility checks, growth rate checks and data validation control are carried out by Eurostat in the form of quality reports by theme sent to the countries shortly after their data submission.
11.2. Quality management - assessment
Eurostat routinely assesses timeliness, completeness and accuracy and follows this up with the data providers.
Lists of licensed railway enterprises, according to ERA, are from time to time provided to countries.
Regarding the indicators about expenditures on rail infrastructure, a benchmarking exercise with similar data collected by the ITF took place in Spring 2026.
12.1. Relevance - User Needs
Key users of the statistics are: the European Commission, Directorate-General Mobility and Transport (DG MOVE); the United Nations Economic Commission for Europe (UNECE); and the International Transport Forum (ITF/OECD). These users are strongly involved in the design and development of the Common Questionnaire, thus user needs are generally considered to be well covered and integrated in the development process. Indicators for policy needs are constantly extracted from the Common Questionnaire, such as the indicators on rail and road infrastructure.
12.2. Relevance - User Satisfaction
Not available.
12.3. Completeness
Completeness for this group of indicators is average (~50%), higher for the number of enterprises by type (~66%), lower for the expenditures on the rolling stock (~25%). Eurostat regularly follows up completeness with the reporting countries (this group of indicators is ‘RAIL III’ in the annexed document on completion rates).
13.1. Accuracy - overall
The issue of accuracy for this group of indicators should be more related to concepts than to sampling..
13.2. Sampling error
There is no sampling issue, but there could be partial non-response. Data are easier to collect from a monopoly than from a widely-open market.
13.3. Non-sampling error
A difficulty for the number of enterprises is to identify the part of inactive licenced railway undertakings. For the investment and maintenance expenditures, issues can be the source of financing and the alternative between “Commitment Appropriations” and “Payment Appropriations”. .
14.1. Timeliness
The deadline for the data transmission is 9-10 months after the end of the reference period (calendar year). The expected deadline for the data dissemination is less than 12 months after the end of the reference period, i.e. about mid-December N+1.
14.2. Punctuality
The amounts of investment and maintenance expenditures are often available later than other indicators, in year N+2.
15.1. Comparability - geographical
GGeographical comparability is generally high due to harmonised definitions used by the national authorities to report data.
15.2. Comparability - over time
Comparability over time is generally good. The indicators on enterprises by type and investment and maintenance expenditures have been unchanged for 35 years. Eurostat addresses every year a 'request for clarifications' to countries, relying mainly on possible inconsistencies over time. Time series breaks that cannot be avoided are flagged.
15.3. Coherence - cross domain
The notions of “railway undertakings” and “integrated companies” together should be close to what is reported in structural business statistics under activities “49.11 Passenger heavy rail transport” and “49.2: Freight rail transport” of NACE Rev. 2.1.
15.4. Coherence - internal
There is a good internal coherence, due to Conval rules.
Each time a new data collection is launched on reference year N, normally in April N+1:
the CQ template of the reference year is pre-filled with the "static data" of the previous 5 years, which helps the countries to identify breaks or gaps in series, or only the need to revise some figures;
pre-filled CQ templates are availabe for revision for all reference years since 2013.
When Eurostat observes an inconsistency in time series, it encourages the countries to revise data since reference year 2013.
All reported errors (once validated) result in corrections of the disseminated data, as soon as possible.
A country can also ask for an ad hoc revision, so that nationally published data are consistent with data disseminated by Eurostat.
18.1. Source data
Data are collected and compiled by the competent national authorities (national statistical institutes or transport administrations). Only aggregated national annual data are reported to Eurostat.
18.2. Frequency of data collection
Annual.
18.3. Data collection
The data collection methodology is determined by the competent national authorities and may vary from country to country.
These national authorities send Excel files (1 per country, theme and year) to EDAMIS portal.
18.4. Data validation
Data validation is first carried out by the competent national authorities. Only aggregated national data are reported to Eurostat. National data are reported through the EDAMIS portal, which contains a number of data checks and controls (Struval rules for the structure validation, Conval rules for the content validation). Further plausibility checks, growth rate checks and data validation controls are carried out by Eurostat in the form of quality reports / requests for clarifications addressed to countries.
18.5. Data compilation
The data collection and compilation methodology is determined by the competent national authorities and varies from country to country.
Eurostat compiles EU total only.
18.6. Adjustment
Eurostat converts expenditures into millions of euro.
The (rail_ec) group of tables provides data about railway enterprises by type of enterprise; and about investment and maintenance expenditures on railway vehicles or rail infrastructure.
There are no railways in Cyprus, Malta and Iceland. There is only a minor railway network in Liechtenstein, but the railway integrated company is Austrian.
The Common Questionnaire is not supported by a legal act, but is based on a goodwill agreement with the participating countries; its completeness may therefore vary from country to country. The Glossary for transport statistics was published since 1994 (1st edition) with the purpose of assisting countries during this data collection
Data is generally provided by the national statistical institutes or responsible transport administrations.
Eurobase tables based on the Common Questionnaire cover annual data from the EU Member States, EFTA countries, candidate countries and potential candidates to the EU. Data from other participating countries are available through the ITF and the UNECE websites. In total, comparable transport data collected through the Common Questionnaire are available for close to 50 countries worldwide.
11 May 2026
The definitions used in the Common Questionnaire are specified in the Glossary for transport statistics. The Glossary is an official publication in EN, FR and DE. The UNECE produces also a version in Russian.
As for rail enterprises and economic performances, the definitions are listed between entries A.III-01 and A.III-17 in the Glossary.
The statistical units observed are the national railway enterprises.
The railway enterprises and their economic performances are observed on an annual basis.
Only aggregated national data are reported by the participating countries; no micro data or confidential data are transmitted.
Member States of the European Union; EFTA states; European Union Candidate countries and potential candidate(s). Other countries are covered by the ITF and UNECE.
Data are collected on a yearly basis.
The issue of accuracy for this group of indicators should be more related to concepts than to sampling..
The ‘Railway enterprises by type of enterprise’ are measured in number of enterprises.
The ‘Expenditure on rolling stock and railway infrastructure’ are measured in million Euros.
The data collection and compilation methodology is determined by the competent national authorities and varies from country to country.
Eurostat compiles EU total only.
Data are collected and compiled by the competent national authorities (national statistical institutes or transport administrations). Only aggregated national annual data are reported to Eurostat.
Annually. An intermediate update can occur 1 or 2 months after first release when late data are transmitted, or when errors are corrected.
The deadline for the data transmission is 9-10 months after the end of the reference period (calendar year). The expected deadline for the data dissemination is less than 12 months after the end of the reference period, i.e. about mid-December N+1.
GGeographical comparability is generally high due to harmonised definitions used by the national authorities to report data.
Comparability over time is generally good. The indicators on enterprises by type and investment and maintenance expenditures have been unchanged for 35 years. Eurostat addresses every year a 'request for clarifications' to countries, relying mainly on possible inconsistencies over time. Time series breaks that cannot be avoided are flagged.
