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.
Global value chains are processes for production of goods and services that take place in several countries. Enterprises increasingly organize their production globally, breaking up their value chains into smaller parts supplied by a growing number of providers located worldwide. Outsourcing of business functions is a key feature of global value chains (GVCs) as European businesses increasingly globalize their production processes. As per defined in Implementing Regulation (EU) 2022/918, the 10 variables are as follows:
(1) employment by business function
(2-5) enterprises purchasing or supplying goods and services abroad
(6) the number of enterprises carrying out international sourcing
(7-8) jobs created and lost resulting from international sourcing
(9) motivations and barriers for sourcing, and
(10) events affecting GVC arrangements.
For variables 2-5, the dataset includes only legal units that report at least NOK 1'000'000 in goods or services purchased from or supplied abroad in the final year of the reference period.
All data are provided as absolute values and presented by NACE activity, business function, geographical partner area, size class, motivation and barrier types, and other breakdowns required by the regulation.
3.2. Classification system
The statistics are published according to the Standard Industrial Classification 2007 (SIC2007). SIC2007 complies with the EU standard, NACE Rev. 2. SN2007 forms the basis for the classification of units by most important activity in the Central Register of Establishments and Enterprises (CRE).
The country codes follow international standard ISO-3166-1 alfa-2.
3.3. Coverage - sector
NACE Rev.2. Sections B to N (business economy), and legal units with 50 or more employees and self-employed persons in the final year of the reference period are included.
3.4. Statistical concepts and definitions
International sourcing refers to the partial or full movement of business functions abroad within or outside the enterprise group, during 2021-2023. Business functions are divided into core and support categories. Furthermore, these categories are divided into specific areas ranging from production to research and development. As per mentioned in 3.2, legal units that exceeds the 1'000'000NOK threshold for goods/services purchased or supplied are included in data series 2-5. Motivations and barriers are ranked with an importance factor in a four-level scale.
3.5. Statistical unit
Legal unit.
3.6. Statistical population
The statistics Global value chains and outsourcing is based on a population consisting of enterprises in industries B to N:
B – Mining and quarrying
C – Manufacturing
D – Electricity, gas, steam and air conditioning supply
E – Water supply; sewerage, waste management and remediation activities
F – Construction
G – Wholesale and retail trade; repair of motor vehicles and motorcycles
H – Transportation and storage
I – Accommodation and food service activities
J – Information and communication
K – Financial and insurance activities
L – Real estate activities
M – Professional, scientific and technical services
N – Administrative and support service activities
The sample is limited to legal units with 50 or more employees and consists of approximately 2 000 legal entities.
3.7. Reference area
Norway as geographical area.
All geographical regions of the mainland are included in addition to Svalbard, Jan Mayen etc. for NACE categories.
3.8. Coverage - Time
2021-2023
3.9. Base period
Not applicable
Variables are compiled in absolute values.
2021-2023
6.1. Institutional Mandate - legal acts and other agreements
The data is compiled solely by the Division of External Trade Statistics in Statistics Norway.
7.1. Confidentiality - policy
The policy statement of Statistics Norway, states the following:
Official statistics must be published in such a way that the data cannot be traced back to the supplier of any data or to any other identifiable individual.
Exceptions may be made where this is not to the detriment of persons or to the unreasonable detriment of an establishment (ref. Statistics Act Section 2-6).
Data on individuals and establishments (where the identity is concealed) may be loaned for research purposes, but only under strict conditions, where privacy is protected.
7.2. Confidentiality - data treatment
Statistics Norway does not publish figures if there is a risk of the respondent’s contribution being identified. This means, as a general rule, figures are not published if small value countsform the basis of a cell in a table or if the contribution of a few respondents constitutes large part of the cell total.
In addition, secondary confidentiality is practised to maintain confidentiality.
8.1. Release calendar
Statistics Norway has a public release calendar for upcoming statistics, which informs users of upcoming publications and updated statistics published by Statistics Norway.
The main principles of Statistics Norway’s communication and dissemination are transparency, accessibility, comprehensibility, independence and confidentiality.
Ssb.no, including Statbank, is the main channel for the dissemination of statistics and analyses based on Statistics Norway’s statistical production and own research. The dissemination principles apply to statistics and analyses regardless of funding source.
Statistics Norway aims to be open about all aspects of statistics and analyses. Any errors in published material must be rectified as soon as possible, with clearly marked corrections.
Planned revisions of the statistics must be accounted for, and users must be alerted in advance of any major changes. Information on the quality of statistics and analyses is openly available in documentation (metadata).
No external users have access to the statistics and analyses prior to their publications, and the data is accessible simultaneously for everyone on ssb.no. However, there are a few exceptions to this principle of equal treatment of users.
The scope of exceptions should be limited and illuminated through documentation in About the statistics. Clients can access publications and other analyses before publication where this is part of a professional cooperation/quality assurance, but only for internal use. The media never have access to content prior to release, nor are given a press embargo.
Statistics and analyses must be notified well in advance in the Statistics Release Calender, and made available for everyone simultaneously, i.e. 8 o'clock on weekdays.
Possible deviations from the notification date are marked in the calendar and the statistics’ home page.
The statistics is published every three years, with a reference period of the three previous years. At this current time, the 2021-2023 statistics has not yet landed on a conclusive date for publishing in Statistics Norway.
10.1. Dissemination format - News release
At this current time, our articles have been published with the lastest publication being back in January 2023. The latest was for the period 2018-2020.
Collected and revised data are stored securely by Statistics Norway in compliance with applicable legislation on data processing. Researchers can be granted access by applying to the respective sections for the underlying microdata.
10.5. Dissemination format - other
Data transmissions to Eurostat is done through the reporting system eDAMIS (electronic Data files Administration and Management Information System).
The system for assuring the quality of Norwegian official statistics is based on the Statistics Act and in the European Statistics Code of Practice.
Annually, Statistics Norway prepares a report on the quality of all official statistics requested by Ministry of Finance. This report covers:
Coordination of official statistics
Quality evaluations of official statistics
Quality reviews of individual statistics
Recommendations for improvements
Follow up of quality reporting
Quality in administrative information systems
User orientation
Development of the quality assurance system
11.2. Quality management - assessment
Our assessment on the overall data quality is that the data is good. The response rate on the survey was approximately 98.5%, so most of the sample filled in the survey. Through our internal checks and dissemination, we contacted respondents regarding potential anomalies. There was a small degree of non-response, but it should not be of significance. All mandatory cells across all tables were fully reported, with certain cells marked with confidentiality.
Currently, we are in the process of moving our production over to the cloud. Since our GVC statistic production is in a transitionary period, it was done partially in Python and SAS EG. We had to tackle some technical issues related to this change. Fortunately, this did not affect the respondents, as this was only affecting the production side.
12.1. Relevance - User Needs
After the 2021-2023 results were published, we received questions and requests for deeper insight from politicians, enterprises and members of the public. Due to the nature of specific questions, we remained firm on our stance to confidentiality.
12.2. Relevance - User Satisfaction
We currently don't collect any information on user satisfaction.
12.3. Completeness
All mandatory cells across all tables were fully reported, with certain cells marked with confidentiality.
13.1. Accuracy - overall
Overall the accuracy of the responses were good, as well as 98.5% response rate. We would also get in contact with respondents if we deemed their survey responses suspicious. The population consists of approximately 4400 legal entities, where the sample consists of about 2000. Due to the nature of surveys, having a large sample relative to the population size results in lower uncertainty when estimating. The results of a statistical survey may nevertheless contain measurement and processing errors, non-response errors and sampling errors.
13.2. Sampling error
The statistics Global value chains and outsourcing is based on enterprises in industries B to N, limited to a sample of enterprises with 50 or more employees. The sample is drawn from the Central Register of Establishments and Enterprises (CRE). CRE is Statistics Norway's central register of all legal entities and enterprises in the private and public sectors in Norway.
The specific population used for sampling for each survey is made based on situational withdrawals from the Central Register of Establishments and Enterprises prior to the completion of the surveys. Furthermore, the selection is determined based on a stratified selection. The strata are defined based on industry classification, broken down by analytical areas and two different employee groups.
That being said, there could be a degree of uncertainty caused by the fact that the figures are produced based on a sample of units and not the entire population. Although, that has not been directly measured.
13.3. Non-sampling error
To reduce the degree of various measurement errors, we have been in contact with respondents whose responses were suspicious. Given the complexity of the survey, a respondent may misunderstand the questions and as such give incorrect answers.
Non-response is a possible error that could occur. The legal entity may have failed to respond to the survey, or partial non-response, i.e. the entity has failed to answer at least one question in the survey. In such cases, it may be appropriate to impute missing values to estimate them properly.
14.1. Timeliness
Data and metadata were transmitted to Eurostat within the specified legal deadline.
14.2. Punctuality
All of our internal milestones regarding the Eurostat transmission were all met. We have not yet decided on a date our national publication.
15.1. Comparability - geographical
We followed the specified classifications (NACE Rev 2.) and the GVC Implementing Act for the specified breakdowns. The data should, in theory, be comparable with other member states, so minor differences in dissemination should not affect comparability in a significant way.
15.2. Comparability - over time
The overall comparability of our previous surveys may be limited even though key definitions, business functions and geographical breakdowns have remained consistent. The contributing factors to this limitation are due to different number of reference years, sample size, questionnaire content and survey design. For example, in the current survey, motivations and barriers have been further expanded upon and the data has become more granular.
15.3. Coherence - cross domain
At this current time, we have not compared across domains.
15.4. Coherence - internal
Internal checks were done prior to sending the data to Eurostat. There were no reported inconsistencies with integrity rules for the GVC data transmitted.
We moved our survey to a new platform, and this allowed for a more streamlined experience with several quality of life improvements were implemented. In terms of burden, the response was mixed. Some found the survey simple to fill in while others found it more difficult. Certain respondents had to correspond with other departments within their enterprises to collect information on their activities.
Due to internal priorities, the survey was sent out slightly later than anticipated. As such, comments from respondents questioning the relevancy would occassionally arise.
Benchmark revisions are done according to Eurostat's recommendations. There are no planned revisions of previously published figures for Statistics Norway's publication.
18.1. Source data
The data is collected through statistical surveys. The sampling is derived from our statistical business register, which provides, if necessary, a live snapshot of every active legal entity for each month.
18.2. Frequency of data collection
Every 3 years in accordance with Implementing Regulation 2022/918.
18.3. Data collection
The information has been collected by means of a statistical survey. The legal units concerned (more than 50 persons employed) are completely enumerated.
Data is collected by electronic reporting on a mandatory basis. The respondents receive a survey for reporting every three years, with a reference period of the three previous years. Cover letters were used to inform about the background of the survey and motivate legal units to fill in the questionnaire.
18.4. Data validation
Our data is validated (degree of confidentiality, aggregation etc.) via logical checks and eDAMIS' own validation program prior to sending the data over to Eurostat.
18.5. Data compilation
Editing is defined here as checking, examining and amending data. Important sources for this work are the annual balance sheet, the Register of Establishments and Enterprises (VoF) and information from the enterprises themselves.
Considering this is a survey statistic, and values are estimated using a stratified expansion estimator. This estimator is made in the following way:
Within each stratum, average reported values are calculated based on the sample.
Because the sample is drawn stratified at random, these averages are good estimates of the actual averages in the various strata. Any extreme observations are kept out of the calculation of averages.
Within each stratum, the calculated averages are multiplied by the number of enterprises in the stratum (in the delimited population). This will be an estimate of the total value in the stratum.
The largest enterprises are always included, and we define these as fully counted enterprises. In this survey, all enterprises with 150 or more employees are full-count enterprises. These are not used as a basis for estimation and should therefore not represent other enterprises in the limited population. These enterprises enter the total only with their own value.
18.6. Adjustment
Not applicable.
No comment.
Global value chains are processes for production of goods and services that take place in several countries. Enterprises increasingly organize their production globally, breaking up their value chains into smaller parts supplied by a growing number of providers located worldwide. Outsourcing of business functions is a key feature of global value chains (GVCs) as European businesses increasingly globalize their production processes. As per defined in Implementing Regulation (EU) 2022/918, the 10 variables are as follows:
(1) employment by business function
(2-5) enterprises purchasing or supplying goods and services abroad
(6) the number of enterprises carrying out international sourcing
(7-8) jobs created and lost resulting from international sourcing
(9) motivations and barriers for sourcing, and
(10) events affecting GVC arrangements.
For variables 2-5, the dataset includes only legal units that report at least NOK 1'000'000 in goods or services purchased from or supplied abroad in the final year of the reference period.
All data are provided as absolute values and presented by NACE activity, business function, geographical partner area, size class, motivation and barrier types, and other breakdowns required by the regulation.
18 November 2025
International sourcing refers to the partial or full movement of business functions abroad within or outside the enterprise group, during 2021-2023. Business functions are divided into core and support categories. Furthermore, these categories are divided into specific areas ranging from production to research and development. As per mentioned in 3.2, legal units that exceeds the 1'000'000NOK threshold for goods/services purchased or supplied are included in data series 2-5. Motivations and barriers are ranked with an importance factor in a four-level scale.
Legal unit.
The statistics Global value chains and outsourcing is based on a population consisting of enterprises in industries B to N:
B – Mining and quarrying
C – Manufacturing
D – Electricity, gas, steam and air conditioning supply
E – Water supply; sewerage, waste management and remediation activities
F – Construction
G – Wholesale and retail trade; repair of motor vehicles and motorcycles
H – Transportation and storage
I – Accommodation and food service activities
J – Information and communication
K – Financial and insurance activities
L – Real estate activities
M – Professional, scientific and technical services
N – Administrative and support service activities
The sample is limited to legal units with 50 or more employees and consists of approximately 2 000 legal entities.
Norway as geographical area.
All geographical regions of the mainland are included in addition to Svalbard, Jan Mayen etc. for NACE categories.
2021-2023
Overall the accuracy of the responses were good, as well as 98.5% response rate. We would also get in contact with respondents if we deemed their survey responses suspicious. The population consists of approximately 4400 legal entities, where the sample consists of about 2000. Due to the nature of surveys, having a large sample relative to the population size results in lower uncertainty when estimating. The results of a statistical survey may nevertheless contain measurement and processing errors, non-response errors and sampling errors.
Variables are compiled in absolute values.
Editing is defined here as checking, examining and amending data. Important sources for this work are the annual balance sheet, the Register of Establishments and Enterprises (VoF) and information from the enterprises themselves.
Considering this is a survey statistic, and values are estimated using a stratified expansion estimator. This estimator is made in the following way:
Within each stratum, average reported values are calculated based on the sample.
Because the sample is drawn stratified at random, these averages are good estimates of the actual averages in the various strata. Any extreme observations are kept out of the calculation of averages.
Within each stratum, the calculated averages are multiplied by the number of enterprises in the stratum (in the delimited population). This will be an estimate of the total value in the stratum.
The largest enterprises are always included, and we define these as fully counted enterprises. In this survey, all enterprises with 150 or more employees are full-count enterprises. These are not used as a basis for estimation and should therefore not represent other enterprises in the limited population. These enterprises enter the total only with their own value.
The data is collected through statistical surveys. The sampling is derived from our statistical business register, which provides, if necessary, a live snapshot of every active legal entity for each month.
The statistics is published every three years, with a reference period of the three previous years. At this current time, the 2021-2023 statistics has not yet landed on a conclusive date for publishing in Statistics Norway.
Data and metadata were transmitted to Eurostat within the specified legal deadline.
We followed the specified classifications (NACE Rev 2.) and the GVC Implementing Act for the specified breakdowns. The data should, in theory, be comparable with other member states, so minor differences in dissemination should not affect comparability in a significant way.
The overall comparability of our previous surveys may be limited even though key definitions, business functions and geographical breakdowns have remained consistent. The contributing factors to this limitation are due to different number of reference years, sample size, questionnaire content and survey design. For example, in the current survey, motivations and barriers have been further expanded upon and the data has become more granular.
