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 contains Prodcom statistics on production of manufactured goods together with related external trade data. The data characteristics described in this documentation refer to the Prodcom production data. For external trade data characteristics, please see documentation for Comext.
The Prodcom data includes:
the physical volume of production sold during the survey period
the value of production sold during the survey period
for some products, the volume of total production during the survey period
the physical volume of production under sub-contracted operations sold during the survey period
the value of production under sub-contracted operations sold during the survey period
The Prodcom data is obtained by the National Statistical Institutes (NSIs) who conduct a survey of enterprises.
3.2. Classification system
The Prodcom survey is based on the Prodcom List, consisting of 4044 products. The 8-digit codes used in the List are based on the 6-digit CPA headings and hence the 4-digit NACE rev 1.1. From 2008 onwards the Prodcom code is linked to CPA 2008 and NACE Rev. 2. The link to NACE enables the NSIs to use the Business Register to identify the enterprises likely to be manufacturing the product.
The Prodcom List is revised every 2-3 years.
The Prodcom codes normally relate to one or more Combined Nomenclature headings, thus enabling external trade data to be related to production data.
There are 4.044 headings representing manufactured products and some industrial services, in the following NACE Rev.2 sectors:
Mining of metal ores
Other mining and quarrying
Mining support service activities
Manufacture of food products
Manufacture of beverages
Manufacture of tobacco products
Manufacture of textiles
Manufacture of wearing apparel
Manufacture of leather and related products
Manufacture of wood and of products of wood and cork, except furniture; manufacture of articles of straw and plaiting materials
Manufacture of paper and paper products
Printing and reproduction of recorded media
Manufacture of coke and refined petroleum products
Manufacture of chemicals and chemical products
Manufacture of basic pharmaceutical products and pharmaceutical preparations
Manufacture of rubber and plastic products
Manufacture of other non-metallic mineral products
Manufacture of basic metals
Manufacture of fabricated metal products, except machinery and equipment
Manufacture of computer, electronic and optical products
Manufacture of electrical equipment
Manufacture of machinery and equipment n.e.c.
Manufacture of motor vehicles, trailers and semi-trailers
Manufacture of other transport equipment
Manufacture of furniture
Other manufacturing
Repair and installation of machinery and equipment
3.3. Coverage - sector
The Coverage of the PRODCOM survey is calculated in terms of turnover value as follows:
Total turnover value for local KAUs in NACE 07, 08, 10-33 in the enterprises in the PRODCOM sample (SBS figures) /Total turnover value for all local KAUs in NACE 07, 08, 10-33 in the SBS population
The turnover value of these local KAUs is measured in the Structural Business Statistics (SBS). The Production Value is lower than the figures shown in the below table, but coverage is equivalent for both Production Value and Turnover Value.
The coverage rate can be estimated on a more detailed basis. As the average coverage is above 90 % in the Norwegian PRODCOM Survey 2007 some NACE 4-digit classes have a coverage rate below the 90 % coverage claim of the Regulation and some detailed NACE classes have a coverage rate beyond the 90 % threshold.
3.4. Statistical concepts and definitions
The purpose of the statistics is to report, for each product in the Prodcom List, how much has been produced in the reporting country during the reference period. This means that Prodcom statistics relate to products (not to activities) and are therefore not strictly comparable with activity-based statistics such as Structural Business Statistics. The NACE codes on which Prodcom codes are based merely serve to identify the enterprises that should be surveyed in order to determine the amount of production of the product.
Prodcom differs from external trade statistics in that the latter can be thought of as event-based: each time a product crosses a border it is registered as a 'trade', and if the same product crosses borders several times it is recorded as several trades.
On the other hand a product is never produced more than once. It is important to avoid double counting, since this inflates the reported quantity of the product produced. As a general principle, when a production process takes as input a material that does not match the description of the product, and produces as output something that does, then production of the product should be recorded. On the other hand if the processing merely works on a product without changing the heading under which it is classified, it should not be recorded, since this would result in double counting.
This means that the link to the turnover of an enterprise is tenuous, since some of the enterprise's activity does not result in new products and should not be recorded in Prodcom.
The products: the Prodcom List defines the products to be included in the survey.
The value and the volume of production. The value is expressed in thousands national currency but where necessary, converted to Euro by Eurostat. The volume is expressed in a unit specified for each product.
The volume unit: the measurement unit, such as kilograms, square metres etc. used to indicate the volume of goods produced.
Production type: Either
Sold Production (the value and volume of the product sold by the enterprise) or
Total Production (the volume of all production of the product, including both the proportion that is sold and the proportion that is retained by the enterprise for adding to stocks, using in further processing etc.). From 2006 onwards, for those products where Total Production is required, Sold Production by value and volume must be reported as well.
As from 2021, the value and the volume of production under sub-contracted operations are produced and published. The value is expressed in national currency but where necessary, converted to Euro by Eurostat. The volume is expressed in a unit specified for each product.
Sold Production under sub-contracted operations (the value is the fee the enterprise has received for production done under-sub-contracted operations and the volume is the volume of the product the enterprise has produced under sub-contracted operations) or
Total Production under sub-contracted operations. Under sub-contracted operations, the unit that produces the good does not own the main material inputs, and ther unit that owns the main material input is not resident in Norwegian economic territory.
3.5. Statistical unit
The reporting unit is the enterprise for enterprises where all the local KAUs classified within NACE 07, 08, 10-33 are classified within the same 5-digit subclass. For the other enterprises the reporting units are kind of activity unit or local kind of activity unit respectively.
3.6. Statistical population
Population according to the Regulation
According to the PRODCOM Council Regulation and Commission Regulation the Survey population must cover:
Enterprises with NACE principal or secondary activities listed in NACE B, C.
Undertakings employing at least 20 people
At least 90 % of national production per NACE Class
Definition of Norwegian PRODCOM population
The PRODCOM survey population is framed on basis on information in the Central Register of Establishments and Enterprises at Statistics Norway (CEE).
The Observation units are selected annually on basis on the CEE at the turn of the year.
The population used for finding the observation units is defined as: All local KAUs that where active at least 3 months during the reference year.
The population of the survey consists of the enterprises that are sampled.
3.7. Reference area
Industrial production on national level
3.8. Coverage - Time
The periode observed is one clander year.
Data available back to 1995.
Website with data 2002 to 2007 based on NACE Rev.1.1 and NACE Rev.2 from 2008.
For volume: Phsical mesurement unit required by the Nomenclature of Industrial Production, spesific for each product.
For value: Thousand NOK, excl. VAT.
2024 (The survey is performed annually)
6.1. Institutional Mandate - legal acts and other agreements
The survey is performed with basis in the Norwegian Statistical Act. This means that the survey is compulsory for the units within the sample. The units not responding are fined.
Regulation (EC) No 223/2009 on European statistics (recital 24 and Article 20(4)) of 11 March 2009 (OJ L 87, p. 164), 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 data is confidential (and marked by):
Three or less with production of the product.
One enterprise has 90 % or more of the production.
Two enterprises have 95 % or more of the production.
8.1. Release calendar
Statistics Norway's website has a calender with all the upcomming publications. All publications must be announced three months prior to publication.
For PRODCOM the practices is to publish the last workingday of June the year after the referance year.
Statistics Norway always release the data to all at the same time. To be able to send data to Eurostat before the deadline (T+6M), all none confidential data are released a few days prior on the Statistics Norway website.
Prodcom data are submitted annually to Eurostat and published annually.
Several actions contribute to maximize the quality of the statistics:
All questionnaires are checked manually at micro-level. They are also checked manually against the figures / questionnaires from the preceding year and against the last year questionnaire from the structural business statistics. There is also a check against an annualized monthly turnover statistic.
For enterprises that have reported suspicious data, the contact persons (as asked for in the questionnaires) are telephoned or contacted by e-mail. The information given is checked against the Central Register of Establishments and Enterprises.
Furthermore, the Internet is used as a source of information.
11.2. Quality management - assessment
Statistics Norway applies checks by comparing unit values across historical and current data and with additional checks when SBS are published. Any anomalies identified are reported to the declarant enterprise for explanation or correction.
Eurostat applies its own data checks by comparing unit values across countries, and comparing historical and current data and with additional checks. Any anomalies identified are reported to the declarant country for explanation or correction. In addition, if users question the accuracy of data, the declarant country is asked to investigate it.
12.1. Relevance - User Needs
National accounts make use of industrial commodity statistics.
Researchers and analysts also use the data.
The data are also used to update the Business register to give the enterprises the right NACE Rev2. code.
12.2. Relevance - User Satisfaction
Not available.
12.3. Completeness
Statistics Norway provides to Eurostat all data requested by the PRODCOM List.
Transmission have been accepted, and the whole dataset is okay.
13.1. Accuracy - overall
The data covers about 90 % of the overall population (from the SBS). There is great effort to get this as close to 90 % and over for each NACE Rev.2 class. However because of the pressure to limit the burden on the enterprises, no enterprise with 10 employees or less are in the survey. This may give some bias to cost per product being less than the real cost (only larger companies report data).
Non-response is checked and imputation is done when needed.
Each questionnaire is gone through by staff at Statistics Norway to check.
13.2. Sampling error
The survey is a cut-off survey. As a result there is no probability sampling errors.
13.3. Non-sampling error
Undercoverage
Enterprises with no local KAUs with more than 20 persons employed (for some NACE divisions 10 or 15) are not covered by the survey. Due to structural differences, some industries are characterized by a relatively high share of the production value coming from small enterprises. The reasons for not including small enterprises is consideration to the respondent burden for especially small enterprises and because the share of total production value in small enterprises is relatively small (in average 10 % of total industrial turnover).
Workplaces making industrial activities may not be registered in the register due to the share of this activity being below the registration threshold set or due to erroneous information in the Register.
Overcoverage
Enterprise or KAUs being part of the frame although they do not produce industrial commodities will find it very difficult to split the turnover up into detailed PRODCOM commodity groups. These units will very likely contact Statistics Norway and the misunderstanding will be corrected.
Another overcoverage error more likely to occur is observation units reporting sale of merchandises as sale of own products. We regularly detect and correct this type of overcoverage but the size is unknown.
Multiple listings
No units are present more than once in the frame. An identity number relating to the name and address of the unit characterizes each and every workplace.
Incorrect auxiliary information
Two types of auxiliary information in the CEE may lead to errors in the frame of the survey:
Incorrect NACE activity information may lead to a unit not apparent in the frame.
Incorrect information in CEE concerning number of employees may also lead to a unit not apparent in the frame.
Actions taken for the assessment of undercoverageS
Small enterprises
Statistics Norway does not want to increase the burdens of the enterprises. We will however consider the possibilities to make estimations for small enterprises for some NACE divisions in the future.
Observation units not returning questionnaire
The response rate is very close to 100 per cent, due to a thoroughly reminder procedure and the Statistical act that gives us the possibility to fine the enterprises not responding. These respondents will be given fines, which however do not exempt respondents from reporting the data to Statistics Norway. Normally about 75 enterprises are reported to the fined annually because the questionnaire has not been received for the PRODCOM survey. The units missing are manually imputed using information from other sources (VAT, account information etc.)
Causes to measurement errors are normally categorized as:
Survey instrument: the form, questionnaire or measuring device used for data collection may lead to the recording of wrong values.
Respondents may, consciously or unconsciously, give erroneous data.
Causes of measurement errors
Last year PRODCOM codes are pre-printed on the questionnaire
This may lead to the enterprises only using the pre-printed codes, even though the production may have been changed. We have tried to reduce this problem by grouping some of the codes. In addition, when there are changes of NACE codes we preprint both all the PRODCOM codes from the old NACE code and all the codes from the new NACE code.
Only sale of own products is registered
According to the Regulation the survey only covers sale of own production. Respondents are asked to make a distinction between commodities of own manufacture and commercial turnover or resale turnover (merchandises). Not all respondents are quite aware of this distinction although it appears from the questionnaire and from the guidelines enclosed. On the questionnaire it is specified what kind of accounting figures that should be specified and the respondent should be aware of the difference between the categories sale of own products and commercial turnover (resale turnover or sale of merchandises).
Thousand-errors
The value figures are supposed to be in NOK 1.000, but sometimes they are given in NOK 1. These errors are however most often easy to detect.
Mechanisms used for reducing measurement error
Micro data are checked systematically by the following procedures when registered:
Comparing with registered data from the same respondent previous period.
Comparing unit price among respondents with same commodity.
Summing controls ensures alignment between value of detailed commodity sales and total sale.
Non-response rate
In our registration system we use codes for the status of the questionnaires. In the table below we have used the following codes:
Questionnairesreceived
The questionnaires either received by electronic reporting or reporting on the sent out paper questionnaires
Taken out
Units have been taken out of the survey for reasons of bankruptcy or merger.
Exceptions
Units handled in a special way.
Returned
Questionnaire returned because they have not reached the respondent. Often happens with enterprises that have gone bankrupt or no longer exists.
No response
Enterprises from which we have gotten no response
Status of questionnaires
2024
Number of enterprises
Per cent
Questionnaires received
2282
99.5 %
Taken out
0
0 %
Exceptions
0
0 %
Returned
0
0 %
No response
52
2.2 %
Total number of enterprises in the survey
2334
100.0 %
Item non-response rate
6.5 %
The response rate is about 65% three weeks after the end of the reference month. At the moment of the final production of the statistics, five weeks after the end of the reference month, there is a response rate of 99%. The weighted non-respons for 2024 is calculated to 0.87 per cent
Causes of unit non-response
The enterprise has gone bankrupt, has been merged into an other enterprise or other kinds of liquidation.
The enterprises have not finished making the accounts at the time of the deadline of the questionnaire.
The respondent refuses to respond.
Causes of item non-response
One of the reasons for missing volume declarations is that the volume unit used in PRODCOM is not always suitable for the commodity that the enterprise is producing.
Some enterprises may have problems specifying the production on the different PRODCOM codes and some of them may then report the production on fewer codes than they are meant to.
Mechanisms used for reducing non-response
The non-response rate is quite low due to several mechanisms used for reducing it:
Statistical act
General reminder procedure
Enterprises not responding are fined
Flexible reception of data
General reminder procedure
The general reminder procedure for declarations to the PRODCOM Statistics is initiated with one written reminder, and then a resolution of fine. In addition important enterprises are reminded either by e-mail or telephone in between the reminder and the resolution of fine.
Due to our work developing the new system of editing, the deadlines for the 2008 survey have been late. The deadlines will be adjusted for the 2009 statistics.
Written reminders procedure
PRODCOM 2024
Number of enterprises
Sent out
Deadline
Population (available number of respondents)
2334
11 MAR 2025
11 APR 2025
Reminder and notice about fine
413
18 APR 2025
02 MAY 2025
Enterprises fined
28
03 June 2025
03 June 2025
The resolution of fine letter is only sent to respondents, and is later sent to the Governmental Collection Office if no reply is received.
Reminder procedure only reducing unit non-response
At present, the reminder procedure is only effectuated for missing declarations measured by value. If the respondent has submitted a questionnaire indicating the value in NOK, but not any volume data, the enterprise will be contacted by phone or e-mail to obtain the volume data.
Non-respondents are fined
In very few cases, none of the above measures has any effect. If the respondent does not respond within 8-9 months after the reference year, the respondent is fined.
No enterprises are fined just because of item non-response.
About 40 respondents are fined every year.
14.1. Timeliness
Provisional data are published 6 months following the end of the reference periode.
Final data for the year before are also updated based on SBS data.
14.2. Punctuality
For the reference year 2024, the provisional data were transmitted to Eurostat on 30 June 2025.
15.1. Comparability - geographical
Statistics are comparable across the EU and Norway (same methodology for PRODCOM data is used).
15.2. Comparability - over time
The data are comparable over time making certain reservations:
Change due to Classifications
The PRODCOM List is changed once a year mainly due to changes in the CPA and the industrial classification. The number of changes varies very much from year to year.
Change due to number of employees
During the calendar year no updating due to the number of employees is made except for large enterprises.
Change due to NACE activity
Once a year the population is updated with enterprises, where there are changes of the local KAUs activity into NACE 07, 08 10-33 or out of NACE 07, 08, 10-33. During the calendar year updating due to changes in activities are only made for large enterprises.
Aggregates of the data are made each year. These aggregates are used to compare the trends in the National accounts with the trends for the PRODCOM for the last two years.
As from 2021, data on production under sub-contracted operations were collected separately from data on production on own account.
15.3. Coherence - cross domain
Consistency checks are done in correlation with SBS and STS data.
15.4. Coherence - internal
Not available.
Not available.
17.1. Data revision - policy
When the annual accounts are public, and we publish annual statistics for industry, we will correct major errors in PRODCOM.
Most likely, more answers will come in over the autumn when more people return to work after leave.
17.2. Data revision - practice
Final data revised on the basis of SBS data for the referance year. Will be done before reference year + 18 months.
No other standard revision.
Ad-hoc revisions may occur later then reference year + 18 months, but only if the revision is absolutely necessary (substantial impact on aggregates).
18.1. Source data
Based on a survey.
The statistics covers enterprises with at least one local KAU (kind-of-activity unit) classified in NACE Rev.2 07, 08, 10-33 (NACE B + C) with at least 20 persons employed. For some NACE divisions the threshold is lower than 20 persons employed. In the tables below you will find these NACE divisions and their thresholds. The population is selected on the basis of the registered number of employees, approximately 1 year before the actual period of the survey.
NACE divisions with threshold 10 persons employed
08.111
14.110
22.110
25.610
28.490
08.113
14.130
22.290
25.730
29.310
10.850
15.200
23.110
26.520
32.120
13.300
16.240
23.190
26.700
32.200
13.910
16.290
23.700
27.900
32.400
13.921
18.130
24.200
28.230
32.500
13.929
18.200
25.290
28.290
32.990
13.990
20.520
25.500
28.300
33.190
NACE divisions with threshold 15 persons employed
10.840
17.290
25.940
28.410
10.860
22.230
26.800
30.120
14.140
25.710
28.250
30.990
18.2. Frequency of data collection
The Survey is carried out annually. Norway does not collect monthly steel-data with reference to the 1 percent threshold rule in the Regulation.
18.3. Data collection
Questionnaire design
To make it easier for the respondent to know what kind of production it is supposed to specify, the questionnaires consists of two parts. The respondent is asked to specify accounting information for; sale of self-produced commodities, repair work, assembly and installation work, contract work, sale of commodities (commercial) and other income. The respondent is then asked to specify the production consisting of the sum of the first four figures. Some codes for industrial services have been included in the national version of the PRODCOM list, so that the whole sum may be specified.
Last year codes used by the enterprise are pre-printed on the questionnaires. In addition we have made groups of codes related to each other. If an enterprise has used one code within the group, all the codes of this group are pre-printed on next year questionnaire.
Electronic questionnaire
From the reference year 2003 it has been possible for respondents to report data electronically to the PRODCOM survey. In the table we present figures for the use of electronic questionnaires.
From the reference year 2012 the only chanel to report data are the electronic PRODCOM survey.
Number of enterprises using the electronic questionnaire The electronic questionnaire is designed in a way that it is easy to recognize when the respondent compare it with the questionnaire on paper. One reason is that most of the respondents are familiar with the questionnaire on paper and this makes the shift to the electronic questionnaire easier.
Processes put in place for controlling and reducing processing errors.
Electronically reporting
It is assumed that processing errors will be reduced when enterprises themselves report their data electronically. There are several possibilities of handling the controlling processes of the data over to the respondent himself, which will reduce processing errors in Statistics Norway. Electronically reporting has been a possibility for the respondents from the 2003 PRODCOM statistics.
Development of better imputation methods
A method of imputing missing volume data by using a model for detecting outliers before making the estimates has been used from the 2004 survey. The introduction of this method has given more and better volume estimates.
Call backs and collection of data from other sources
The PRODCOM survey is very detailed and there are no sources available with the same type of detailed data. We have however started to compare the data with data from external trade on a 5-digit CPA-level to see if the same codes have been used.
Non-response concerning missing volume data is taken care of on a continuing basis but it is not possible to reduce it completely if only for that reason that for a number of headings there is only one respondent for whom it may not be possible to declare the required volume.
18.4. Data validation
Several plausibility checks are applied (by the Statistics Norway and Eurostat) to Prodcom data.
Eurostat consults the NSIs in case of errors or anomalies.
18.5. Data compilation
Processing of data between collection and production of statistics. The questionnaires run through the following processes: The data from the electronic questionnaires are automatically registered and loaded into an oracle database every morning.
After the process of registration data, the micro data are controlled on a micro level.
Control lists are generated to check:
If total value of PRODCOM goods is equal to turnover
If invalid PRODCOM codes have been used
If less than 50% of an enterprise's production falls within its NACE activity class
If prices on goods that differ by more than 100% from the mean price of the PRODCOM commodities concerned.
Trend in total turnover on detailed NACE level is studied and a comparison to former periods is made. The figures are also compared with the estimates made in the National Accounts.
Processing errors
Errors from electronic reporting are minor to manual registration errors
The figures from the electronic questionnaires are the same as the enterprises have entered. We assume that manual registration causes much more and larger errors than the additional errors made by the respondent entering the data into the electronic questionnaire. As the share of enterprises using the electronic questionnaire is increasing, the processing errors from registration will be reduced.
Imputation
The response rate is quite high due to the strict way non-responses are handled. The Survey is compulsory. This means that enterprises not returning the questionnaire in due time is sent 2-3 reminders and after that the respondents that do not respond are fined. The need for imputation of value figures is therefore limited. The imputation is made using figures from other sources, accounting figures and figures from the VAT statistics. We also use the questionnaire from last year or a questionnaire from a similar enterprise.
Estimation/imputation methods used Estimation of unit non-responses
For the missing units, figures are calculated on the basis of the account figures or turnover (VAT-statistics) and the last year survey (if available).
Estimation of item non-responses
The most important estimation of item non-response is the estimation of volume, due to the fact that we have a high degree of response for volume data.
The existing method of imputing missing volume quantities is based on the common ratio model:
Vj = pMj + ej , ej - N(0, Mjσ2) for each unit (j)
where V = value, M = volume and p = price for a given PRODCOM code.
For each PRODCOM code (i), the price (p) is estimated by i
The estimated price is accepted if
2
For each unit (j) with volume quantities (M) missing, the volume is imputed by:
The imputation is only accepted if less than half of the total volume of the PRODCOM code has been estimated. In addition there are disclosure restrictions for publishing volume and values for the PRODCOM codes.
We know that outliers can have effect on regression estimators, both their value and standard deviation. In order to improve the method for price estimation we have used an automatic method for detecting outliers.
Units were defined as outliers if there were at least 4 units which had reported both value and volume within the same PRODCOM code and the unit could be assumed influential by the hat diag or had a large residual. The criterions for outlier diagnostics have been tested graphically for PRODCOM codes where traditional imputation method failed. Criterions used here, are:
Studentized residual
Observation with absolute value of Studentized residual greater than 3, are defined as outliers.
Influential observations - hat diag
The leverage of an observation in the common ratio model (Hat diag in SAS) is
which sums up to 1.
Observations with < , are defined as outliers, k = 4.
Outliers are excluded in the estimation of the volume figures.
Non-response rate (%) = 6.5%
Imputation (%) = Not availed for 2023, but approx 18% of the products are imputation
18.6. Adjustment
Not applicable
Metadata update is updated with 2025-dates in 2.1, 2.2 and 2.3
3.9 Base period is updated to 2021 from 2015
13. 3 Non-response error; Updated with weighted non-response for 2024; 0.87 per cent
Prodcom contains Prodcom statistics on production of manufactured goods together with related external trade data. The data characteristics described in this documentation refer to the Prodcom production data. For external trade data characteristics, please see documentation for Comext.
The Prodcom data includes:
the physical volume of production sold during the survey period
the value of production sold during the survey period
for some products, the volume of total production during the survey period
the physical volume of production under sub-contracted operations sold during the survey period
the value of production under sub-contracted operations sold during the survey period
The Prodcom data is obtained by the National Statistical Institutes (NSIs) who conduct a survey of enterprises.
29 August 2025
The purpose of the statistics is to report, for each product in the Prodcom List, how much has been produced in the reporting country during the reference period. This means that Prodcom statistics relate to products (not to activities) and are therefore not strictly comparable with activity-based statistics such as Structural Business Statistics. The NACE codes on which Prodcom codes are based merely serve to identify the enterprises that should be surveyed in order to determine the amount of production of the product.
Prodcom differs from external trade statistics in that the latter can be thought of as event-based: each time a product crosses a border it is registered as a 'trade', and if the same product crosses borders several times it is recorded as several trades.
On the other hand a product is never produced more than once. It is important to avoid double counting, since this inflates the reported quantity of the product produced. As a general principle, when a production process takes as input a material that does not match the description of the product, and produces as output something that does, then production of the product should be recorded. On the other hand if the processing merely works on a product without changing the heading under which it is classified, it should not be recorded, since this would result in double counting.
This means that the link to the turnover of an enterprise is tenuous, since some of the enterprise's activity does not result in new products and should not be recorded in Prodcom.
The products: the Prodcom List defines the products to be included in the survey.
The value and the volume of production. The value is expressed in thousands national currency but where necessary, converted to Euro by Eurostat. The volume is expressed in a unit specified for each product.
The volume unit: the measurement unit, such as kilograms, square metres etc. used to indicate the volume of goods produced.
Production type: Either
Sold Production (the value and volume of the product sold by the enterprise) or
Total Production (the volume of all production of the product, including both the proportion that is sold and the proportion that is retained by the enterprise for adding to stocks, using in further processing etc.). From 2006 onwards, for those products where Total Production is required, Sold Production by value and volume must be reported as well.
As from 2021, the value and the volume of production under sub-contracted operations are produced and published. The value is expressed in national currency but where necessary, converted to Euro by Eurostat. The volume is expressed in a unit specified for each product.
Sold Production under sub-contracted operations (the value is the fee the enterprise has received for production done under-sub-contracted operations and the volume is the volume of the product the enterprise has produced under sub-contracted operations) or
Total Production under sub-contracted operations. Under sub-contracted operations, the unit that produces the good does not own the main material inputs, and ther unit that owns the main material input is not resident in Norwegian economic territory.
The reporting unit is the enterprise for enterprises where all the local KAUs classified within NACE 07, 08, 10-33 are classified within the same 5-digit subclass. For the other enterprises the reporting units are kind of activity unit or local kind of activity unit respectively.
Population according to the Regulation
According to the PRODCOM Council Regulation and Commission Regulation the Survey population must cover:
Enterprises with NACE principal or secondary activities listed in NACE B, C.
Undertakings employing at least 20 people
At least 90 % of national production per NACE Class
Definition of Norwegian PRODCOM population
The PRODCOM survey population is framed on basis on information in the Central Register of Establishments and Enterprises at Statistics Norway (CEE).
The Observation units are selected annually on basis on the CEE at the turn of the year.
The population used for finding the observation units is defined as: All local KAUs that where active at least 3 months during the reference year.
The population of the survey consists of the enterprises that are sampled.
Industrial production on national level
2024 (The survey is performed annually)
The data covers about 90 % of the overall population (from the SBS). There is great effort to get this as close to 90 % and over for each NACE Rev.2 class. However because of the pressure to limit the burden on the enterprises, no enterprise with 10 employees or less are in the survey. This may give some bias to cost per product being less than the real cost (only larger companies report data).
Non-response is checked and imputation is done when needed.
Each questionnaire is gone through by staff at Statistics Norway to check.
For volume: Phsical mesurement unit required by the Nomenclature of Industrial Production, spesific for each product.
For value: Thousand NOK, excl. VAT.
Processing of data between collection and production of statistics. The questionnaires run through the following processes: The data from the electronic questionnaires are automatically registered and loaded into an oracle database every morning.
After the process of registration data, the micro data are controlled on a micro level.
Control lists are generated to check:
If total value of PRODCOM goods is equal to turnover
If invalid PRODCOM codes have been used
If less than 50% of an enterprise's production falls within its NACE activity class
If prices on goods that differ by more than 100% from the mean price of the PRODCOM commodities concerned.
Trend in total turnover on detailed NACE level is studied and a comparison to former periods is made. The figures are also compared with the estimates made in the National Accounts.
Processing errors
Errors from electronic reporting are minor to manual registration errors
The figures from the electronic questionnaires are the same as the enterprises have entered. We assume that manual registration causes much more and larger errors than the additional errors made by the respondent entering the data into the electronic questionnaire. As the share of enterprises using the electronic questionnaire is increasing, the processing errors from registration will be reduced.
Imputation
The response rate is quite high due to the strict way non-responses are handled. The Survey is compulsory. This means that enterprises not returning the questionnaire in due time is sent 2-3 reminders and after that the respondents that do not respond are fined. The need for imputation of value figures is therefore limited. The imputation is made using figures from other sources, accounting figures and figures from the VAT statistics. We also use the questionnaire from last year or a questionnaire from a similar enterprise.
Estimation/imputation methods used Estimation of unit non-responses
For the missing units, figures are calculated on the basis of the account figures or turnover (VAT-statistics) and the last year survey (if available).
Estimation of item non-responses
The most important estimation of item non-response is the estimation of volume, due to the fact that we have a high degree of response for volume data.
The existing method of imputing missing volume quantities is based on the common ratio model:
Vj = pMj + ej , ej - N(0, Mjσ2) for each unit (j)
where V = value, M = volume and p = price for a given PRODCOM code.
For each PRODCOM code (i), the price (p) is estimated by i
The estimated price is accepted if
2
For each unit (j) with volume quantities (M) missing, the volume is imputed by:
The imputation is only accepted if less than half of the total volume of the PRODCOM code has been estimated. In addition there are disclosure restrictions for publishing volume and values for the PRODCOM codes.
We know that outliers can have effect on regression estimators, both their value and standard deviation. In order to improve the method for price estimation we have used an automatic method for detecting outliers.
Units were defined as outliers if there were at least 4 units which had reported both value and volume within the same PRODCOM code and the unit could be assumed influential by the hat diag or had a large residual. The criterions for outlier diagnostics have been tested graphically for PRODCOM codes where traditional imputation method failed. Criterions used here, are:
Studentized residual
Observation with absolute value of Studentized residual greater than 3, are defined as outliers.
Influential observations - hat diag
The leverage of an observation in the common ratio model (Hat diag in SAS) is
which sums up to 1.
Observations with < , are defined as outliers, k = 4.
Outliers are excluded in the estimation of the volume figures.
Non-response rate (%) = 6.5%
Imputation (%) = Not availed for 2023, but approx 18% of the products are imputation
Based on a survey.
The statistics covers enterprises with at least one local KAU (kind-of-activity unit) classified in NACE Rev.2 07, 08, 10-33 (NACE B + C) with at least 20 persons employed. For some NACE divisions the threshold is lower than 20 persons employed. In the tables below you will find these NACE divisions and their thresholds. The population is selected on the basis of the registered number of employees, approximately 1 year before the actual period of the survey.
NACE divisions with threshold 10 persons employed
08.111
14.110
22.110
25.610
28.490
08.113
14.130
22.290
25.730
29.310
10.850
15.200
23.110
26.520
32.120
13.300
16.240
23.190
26.700
32.200
13.910
16.290
23.700
27.900
32.400
13.921
18.130
24.200
28.230
32.500
13.929
18.200
25.290
28.290
32.990
13.990
20.520
25.500
28.300
33.190
NACE divisions with threshold 15 persons employed
10.840
17.290
25.940
28.410
10.860
22.230
26.800
30.120
14.140
25.710
28.250
30.990
Prodcom data are submitted annually to Eurostat and published annually.
Provisional data are published 6 months following the end of the reference periode.
Final data for the year before are also updated based on SBS data.
Statistics are comparable across the EU and Norway (same methodology for PRODCOM data is used).
The data are comparable over time making certain reservations:
Change due to Classifications
The PRODCOM List is changed once a year mainly due to changes in the CPA and the industrial classification. The number of changes varies very much from year to year.
Change due to number of employees
During the calendar year no updating due to the number of employees is made except for large enterprises.
Change due to NACE activity
Once a year the population is updated with enterprises, where there are changes of the local KAUs activity into NACE 07, 08 10-33 or out of NACE 07, 08, 10-33. During the calendar year updating due to changes in activities are only made for large enterprises.
Aggregates of the data are made each year. These aggregates are used to compare the trends in the National accounts with the trends for the PRODCOM for the last two years.
As from 2021, data on production under sub-contracted operations were collected separately from data on production on own account.
2 
< 
, are defined as outliers, k = 4.