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.
Prodcom statistics contain data on national production of manufactured goods. The data is collected using the Combined Nomenclature (CN) but is translated to Prodcom before transmission to Eurostat. The collected variables are:
251001. Sold production
251002. Production under sub-contracted operations
251003. Actual production
For variable 251001 (Sold production): national currency (thousands) and (except for industrial services) quantity as defined in the PRODCOM list in force at the end of the reference period.
For variable 251002 (Production under sub-contracted operations): (except for industrial services) national currency (thousands) and quantity as defined in the PRODCOM list in force at the end of the reference period.
For variable 251003 (Actual production): quantity as defined in the PRODCOM list in force at the end of the reference period.
The response rate was somewhat lower than normal due to a decision to not use financial penalties for non-respondents (participation in the survey is still mandatory).
3.2. Classification system
Goods are classified according to a national version of CN (the Combined Nomenclature). Some CN codes are divided into two or more codes to maintain the link to PRODCOM. Codes for industrial services are also added and they are based on SNI2007 which is the Swedish national classification based on NACE Rev.2. For submission to Eurostat, data is converted from the national version of CN to Prodcom codes using the key between the two lists of commodities.
3.3. Coverage - sector
The survey covers industrial products, according to PRODCOM, from activities listed in sections B, C and E in the Statistical classifcation of economic activities in the European Community, NACE (Rev. 2).
3.4. Statistical concepts and definitions
Total value of sold industrial production.
Value of sold industrial production from goods produced under subcontracting outside of national territory. The goods have been taken home to the national territory before being sold.
Merchanting.
Other income.
Value of sold industrial production specified on own production, subcontracting (fee received), goods produced under subcontracting on national territory, maintenance and repairs, assembly work, other.
Value of sold industrial production specified using codes from national version of CN.
Volume of sold industrial production specified using codes from national version of CN.
Volume of actual industrial production specified using codes from national version of CN.
3.5. Statistical unit
Kind-of-Activity-Unit (KAU) is observation unit and reporting unit.
3.6. Statistical population
The population for the statistics are all kind of activity units (KAU's) with main or secondary activity in mining and quarrying, manufacturing or materials recovery, and having at least 20 employees. The population is selected annually from the Statistical Business Register.
3.7. Reference area
The statistics cover products produced in Sweden.
3.8. Coverage - Time
Since 1996 annual data is published in Statistics Sweden's database.
3.9. Base period
Not applicable.
For volumes: measure units are specific to each PRODCOM code. For codes that are not included in PRODCOM the volume unit is ton.
For value (sales): 1000 SEK (Swedish Krona).
Year 2024.
6.1. Institutional Mandate - legal acts and other agreements
This survey is carried out in compliance with the following regulations:
At EU level:
Regulation (EU) 2019/2152 of the European Parliament and of the Council of 27 November 2019 on European business statistics, repealing 10 legal acts in the field of business statistics (Text with EEA relevance)
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 (Text with EEA relevance)
Commission Regulation (EU) 2022/2552 of 12 December 2022 laying down the technical specifications of data requirements for the detailed topic industrial production statistics establishing the industrial products classification breakdown, pursuant to Regulation (EU) 2019/2152 of the European Parliament and of the Council and amending Commission Implementing Regulation (EU) 2020/1197, as regards the coverage of the product classification (Text with EEA relevance)
At National level: There is an obligation to provide information for this survey according to the Official Statistics Act (SFS 2001:99). The statistics are also regulated according to the Official Statistics Ordinance (SFS 2001:100) and Statistics Sweden's regulations (SCB-FS 2021:33) changed by (SCB -FS 2024:26).
6.2. Institutional Mandate - data sharing
Not applicable at Member States level.
7.1. Confidentiality - policy
The Swedish Prodcom survey follows Statistics Sweden's general confidentiality policy. At national level confidentiality is regulated by SFS 2009:400. It regulates which information from authorities may not be disclosed due to confidentiality. Within Statistics Sweden’s statistical operations, there is normally so-called absolute secrecy, which means that no information about individuals is disclosed, regardless of the reason. There are a few, narrow exceptions to this absolute secrecy.
At EU level confidentiality is regulated by Regulation (EC) No 223/2009. It stipulates the need to establish common principles and guidelines ensuring the confidentiality of data used for the production of European statistics and the access to those confidential data with due account for technical developments and the requirements of users in a democratic society.
7.2. Confidentiality - data treatment
The Swedish PRODCOM survey must apply active confidentiality according to Statistics Sweden policy. We first use the threshold rule and then the p%-rule for disclosure control on all PRODCOM codes. About 65 percent of all PRODCOM codes with sold value are confidential because of disclosure risk. We have recently started to use waivers where we collect consent from enterprises with disclosure risks. After consents about 35 percent of the PRODCOM codes are confidential.
8.1. Release calendar
The publication date appears in the release calendar. The date is confirmed in the weeks before.
Statistics are always published at 8:00 a.m. on the day announced in the release calendar. No one outside of Statistics Sweden can access the statistics before they are published.
The data is published in the national database (t+6 m provisional data, t+16 m final data) on headings from national version of CN.
10.1. Dissemination format - News release
Not available.
10.2. Dissemination format - Publications
No publications are published since reference year 2021.
10.3. Dissemination format - online database
Data is published in the Statistics Sweden online database on headings from the national version of CN.
10.4. Dissemination format - microdata access
Researchers in Sweden can access micro data for approved research projects through MONA (Microdata Online Access).
We use time series to check the estimates of sold value, quantities and unit values on product groups. We use plausibility checks to study the relations between estimates of sold value and quantities. For each product group the estimates are based on collected data and imputed micro data. The estimates are graded into one single quality measurement calculated by the degree of collected data and on the imputation method. For “not sufficiently good estimates” the product groups are marked up and the estimates are not published.
The sold values are also checked and compared, on enterprise level, with the industrial turnover from the SBS survey.
The estimates from the Prodcom survey are sometimes compared, on aggregated level, with estimates from the short term statistics (Turnover statistics, IPI) and the Foreign trade statistics.
12.1. Relevance - User Needs
The most important users of our data are:
Eurostat: Publishes values and quantities on PRODCOM codes.
National Accounts (NA):Uses data for GDP calculations. Extra information is collected to meet demands of NA. E.g. total turnvnover and CN codes not included in PRODCOM are collected.
United Nations (UN):Publishes values and quantities. For some codes UN requests a different quantity unit than is collected. This demand is not met.
Researchers:Uses data for research. No special considerations are given. There has been no indication of unmet user needs.
Department of Agriculture:Uses data to calculate food consumption. No special considerations are given. There has been no indication of unmet user needs.
Statistics Sweden's Producer Price Indices (PPI):Uses data for to calculate weights for PPI. No special considerations are given. There has been no indication of unmet user needs.
Statistics Sweden's Business Register:Uses data to classify the NACE of local units. No special considerations are given. There has been no indication of unmet user needs.
SBS survey:Uses data to estimate the value of industrial production for enterprises that are not surveyed. No special considerations are given. There has been no indication of unmet user needs.
12.2. Relevance - User Satisfaction
There are no user satisfaction surveys. There is an annual meeting with users from the National Accounts.
12.3. Completeness
All PRODCOM codes are covered in the national version of CN.
13.1. Accuracy - overall
The Swedish Prodcom survey uses the Swedish Business Register as a sampling frame. All the enterprises with 20 or more employees are obliged to report their industrial production. Enterprises below the threshold which carry out industrial activities and non-responding enterprises are model estimated. The model estimation is executed in two steps. First, estimates an enterprise total sold value. Second, for each enterprise, distributes the total sold value on product groups and imputes quantities. In the first step the model estimation is quite integrated and coordinated with estimates in the Swedish SBS survey.
We think that the coverage error in the Prodcom survey is of minor size. The major sources of errors arises from the model estimation and measurement errors (mainly from classifying products to CN codes). In the future we want to investigate if it is possible to estimate the mean square error (at least the errors from the model imputation) on product groups in order to improve the quality measurement (see 11.2).
The response rate was somewhat lower than normal due to a decision to not use financial penalties for non-respondents (participation in the survey is still mandatory).
13.2. Sampling error
Not relevant for this statistics.
13.3. Non-sampling error
The main non-sampling error is the measurement error concerning classification at the most detailed CN level, as respondents do not always report sales according to the correct codes. Furthermore, data on quantities are generally less reliable than those on values, as some respondents estimate quantities and other do not answer, so that estimations must be made in the statistical production process.
Unit non-response - rate Un-weighted: 13.34 % Size-weighted: 9.12 %
Multiple listings Multiple listings are estimated to not exist.
14.1. Timeliness
Preliminary data is published t+6 months after the reference year. Final data is published t+16 months after the reference year.
14.2. Punctuality
The publications are timely and fully compliant with the Eurostat regulation and timetable.
15.1. Comparability - geographical
Comparative PRODCOM data for EU countries is available on the Eurostat website.
Geographical comparability at national level is not applicable as there is no regional data on production in Sweden.
15.2. Comparability - over time
Data on national production is available in publications from 1913. From 1959 to 1987 data was published on CCC-codes (Customs Cooperation Council Nomenclature). For 1988-1995 data was published on HS-codes (Harmonized Commodity and Coding System). Since 1996 data is available online on a national version of CN codes (Combined Nomenclature). In 2007 there was a major revision of NACE that had some impact on the national version of CN and which activities that are included in the survey.
For reference year 2021 the new EBS manual was implemented. KAU replaced Enterprise as observation unit and a new variable "Production under sub-contracted operations" was added. Using KAU can lead to "new" production being made visible (sales between KAU's belonging to the same Enterprise). Production under sub-contracted operations was previously included in the variable Sold production. The new variable Total sold production is comparable to the old Sold production variable.
15.3. Coherence - cross domain
Consistency checks are done in correlation with STS & SBS data, in order to ensure the accuracy of the data. Cross-checks are made between SBS, external trade and STS at micro and macro level.
15.4. Coherence - internal
PRODCOM statistics are internally coherent by their nature. Each level of aggregated data is compiled directly from micro data.
The annual cost for producing the Swedish production statistics is 5,400,000 SEK (approx. 580,000 EUR).
The estimated burden for responding enterprises is approximately 4,000,000 SEK (approx. 430,000 EUR). This estimation is based on an average time of 60 minutes per respondentand an estimated cost per hour.
An electronic online questionnaire is used to collect data. Data for the previous year is pre-printed in the questionnaire. For reference year 2021 there were several changes made to the integrated controls to minimize the burden for the respondents.
17.1. Data revision - policy
Statistics Sweden revises published figures in accordance with the revision policy for Statistics Sweden.
Data is published two times for each reference year. No other major revisions are done after specific time intervals.
17.2. Data revision - practice
The preliminary data in the national database is revised 10 months after they are published. Data for earlier years are revised when needed.
First data transmission of preliminary data: T+6
Data revision/transmission of final data: T+16
18.1. Source data
Approximately 4,000 units are part of the survey, selected by a general cut-off (min. 20 employees). Data is estimated for units below the cut-off.
18.2. Frequency of data collection
Annual.
18.3. Data collection
The data is collected using an electronical questionnaire.
Respondents break down the value of industrial activities in Sweden in:
Goods of own manufacture
Contract processing (sub-contracting)
Repairs and maintenance
Assembly and installation
Other processing
Manufacture from another enterprise
These categories are then further broken down by CN codes. For each CN code value, sold quantity (for most codes), total produced quantity (for some codes).
18.4. Data validation
The web-questionnaire contains automatic validation of reported data. If data are very different from those reported by the same company for the last period, the respondent will be prompted to check again the data and to provide an explanation for the change.
In Statistics Sweden the data are validated through a number of controls that check against last period as well as against other companies reporting under the same codes (unit price control). Reporting under product codes that are not typical for the NACE (activity) of the reporting unit are also controlled for.
Data are also checked against other sources, especially SBS, both at micro (enterprise) and macro (NACE) level. Data is also checked against export statistics at micro level.
18.5. Data compilation
After the imputations (see 13.1) have been executed, there will be no missing values on enterprises and their study variables. Micro data are aggregated to product groups. No further processing is done besides to perform disclosure controls and calculate quality measures on product groups.
18.6. Adjustment
Not available.
No comments.
Prodcom statistics contain data on national production of manufactured goods. The data is collected using the Combined Nomenclature (CN) but is translated to Prodcom before transmission to Eurostat. The collected variables are:
251001. Sold production
251002. Production under sub-contracted operations
251003. Actual production
For variable 251001 (Sold production): national currency (thousands) and (except for industrial services) quantity as defined in the PRODCOM list in force at the end of the reference period.
For variable 251002 (Production under sub-contracted operations): (except for industrial services) national currency (thousands) and quantity as defined in the PRODCOM list in force at the end of the reference period.
For variable 251003 (Actual production): quantity as defined in the PRODCOM list in force at the end of the reference period.
The response rate was somewhat lower than normal due to a decision to not use financial penalties for non-respondents (participation in the survey is still mandatory).
25 July 2025
Total value of sold industrial production.
Value of sold industrial production from goods produced under subcontracting outside of national territory. The goods have been taken home to the national territory before being sold.
Merchanting.
Other income.
Value of sold industrial production specified on own production, subcontracting (fee received), goods produced under subcontracting on national territory, maintenance and repairs, assembly work, other.
Value of sold industrial production specified using codes from national version of CN.
Volume of sold industrial production specified using codes from national version of CN.
Volume of actual industrial production specified using codes from national version of CN.
Kind-of-Activity-Unit (KAU) is observation unit and reporting unit.
The population for the statistics are all kind of activity units (KAU's) with main or secondary activity in mining and quarrying, manufacturing or materials recovery, and having at least 20 employees. The population is selected annually from the Statistical Business Register.
The statistics cover products produced in Sweden.
Year 2024.
The Swedish Prodcom survey uses the Swedish Business Register as a sampling frame. All the enterprises with 20 or more employees are obliged to report their industrial production. Enterprises below the threshold which carry out industrial activities and non-responding enterprises are model estimated. The model estimation is executed in two steps. First, estimates an enterprise total sold value. Second, for each enterprise, distributes the total sold value on product groups and imputes quantities. In the first step the model estimation is quite integrated and coordinated with estimates in the Swedish SBS survey.
We think that the coverage error in the Prodcom survey is of minor size. The major sources of errors arises from the model estimation and measurement errors (mainly from classifying products to CN codes). In the future we want to investigate if it is possible to estimate the mean square error (at least the errors from the model imputation) on product groups in order to improve the quality measurement (see 11.2).
The response rate was somewhat lower than normal due to a decision to not use financial penalties for non-respondents (participation in the survey is still mandatory).
For volumes: measure units are specific to each PRODCOM code. For codes that are not included in PRODCOM the volume unit is ton.
For value (sales): 1000 SEK (Swedish Krona).
After the imputations (see 13.1) have been executed, there will be no missing values on enterprises and their study variables. Micro data are aggregated to product groups. No further processing is done besides to perform disclosure controls and calculate quality measures on product groups.
Approximately 4,000 units are part of the survey, selected by a general cut-off (min. 20 employees). Data is estimated for units below the cut-off.
The data is published in the national database (t+6 m provisional data, t+16 m final data) on headings from national version of CN.
Preliminary data is published t+6 months after the reference year. Final data is published t+16 months after the reference year.
Comparative PRODCOM data for EU countries is available on the Eurostat website.
Geographical comparability at national level is not applicable as there is no regional data on production in Sweden.
Data on national production is available in publications from 1913. From 1959 to 1987 data was published on CCC-codes (Customs Cooperation Council Nomenclature). For 1988-1995 data was published on HS-codes (Harmonized Commodity and Coding System). Since 1996 data is available online on a national version of CN codes (Combined Nomenclature). In 2007 there was a major revision of NACE that had some impact on the national version of CN and which activities that are included in the survey.
For reference year 2021 the new EBS manual was implemented. KAU replaced Enterprise as observation unit and a new variable "Production under sub-contracted operations" was added. Using KAU can lead to "new" production being made visible (sales between KAU's belonging to the same Enterprise). Production under sub-contracted operations was previously included in the variable Sold production. The new variable Total sold production is comparable to the old Sold production variable.
