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.
DG Eurostat, the statistical office of the European Union
1.2. Contact organisation unit
G3: Business cycle; Short-term statistics
1.3. Contact name
Restricted from publication
1.4. Contact person function
Restricted from publication
1.5. Contact mail address
2920 Luxembourg LUXEMBOURG
1.6. Contact email address
Restricted from publication
1.7. Contact phone number
Restricted from publication
1.8. Contact fax number
Restricted from publication
2.1. Metadata last certified
12 September 2024
2.2. Metadata last posted
12 September 2024
2.3. Metadata last update
12 September 2024
3.1. Data description
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.
From reference year 2021 onwards, the Prodcom data includes:
the physical volume of production sold during the survey period
the physical volume of production under sub-contracted operations during the survey period
the value of production sold during the survey period
the value of production under sub-contracted operations during the survey period
for some products, the volume of actual production during the survey period
The Prodcom data is obtained by the National Statistical Institutes (NSIs) who conduct a survey of enterprises.
Eurostat calculates EU totals at EU-15, EU-25, EU-27_2007 and EU-27_2020 level (depending on the year) from the national data.
For more details please see here below:
1. "INDICATORS metadata.xlsx" dimension metadata used online (COMEXT, Data Browser) is available in EN, FR, DE.
The Prodcom survey is based on the Prodcom List, consisting of about 4000 products. The 8-digit codes used in the PRODCOM List are derived from the 6-digit CPA headings and hence the 4-digit NACE. From 2008 onwards the Prodcom code is linked to NACE Rev. 2 and to CPA 2008 (to CPA 2.1 from 2016). 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 was revised every year until 2017. Since then, the Prodcom List is revised every 2 or 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.
The Geonomenclature is used to identify reporting countries.
3.3. Coverage - sector
Prodcom covers sections B, C and E (38.32) of NACE Rev.2.
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.
The NSI in each reporting country carries out a survey of industrial production in that country, collates the results and transmits them to Eurostat. Eurostat calculates EU totals and publishes the national and EU data together with the related external trade data.
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.
Reporting countries have a major difficulty in identifying all the enterprises that produce a product. Although they can use the Business Register to find enterprises whose primary or secondary activities are linked to the product being surveyed, they can never be sure that they have identified all producers of the product.
The products: the Prodcom List defines the products to be included in the survey.
There are approximately 4000 headings representing manufactured products and some industrial services, in the following NACE Rev.2 sectors:
07 - Mining of metal ores
08 - Other mining and quarrying
10 - Manufacture of food products
11 - Manufacture of beverages
12 - Manufacture of tobacco products
13 - Manufacture of textiles
14 - Manufacture of wearing apparel
15 - Manufacture of leather and related products
16 - Manufacture of wood and of products of wood and cork, except furniture; manufacture of articles of straw and plaiting materials
17 - Manufacture of paper and paper products
18 - Printing and recording services
19 - Manufacture of coke and refined petroleum products
20 - Manufacture of chemicals and chemical products
21 - Manufacture of basic pharmaceutical products and pharmaceutical preparations
22 - Manufacture of rubber and plastic products
23 - Manufacture of other non-metallic mineral products
24 - Manufacture of basic metals
25 - Manufacture of fabricated metal products, except machinery and equipment
26 - Manufacture of computer, electronic and optical products
27 - Manufacture of electrical equipment
28 - Manufacture of machinery and equipment n.e.c.
29 - Manufacture of motor vehicles, trailers and semi-trailers
30 - Manufacture of other transport equipment
31 – Manufacture of furniture
32 - Other manufactured products
33 - Repair and installation services of machinery and equipment
38 (38.32) - Recovery of sorted materials
The reporting countries: those countries that conduct the survey and transmit the results to Eurostat. The countries are identified by the Geonomenclature codes.
The value and the volume of production. The value is expressed in national currency in thousands 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.
EBS variables on industrial production include the following data:
Variable 251001: The sold production is the production produced on own account and may be carried out under the primary or secondary activities of the enterprise. It includes production sold (invoiced) between different kind-of-activity units belonging to the same enterprise.
Variable 251002: The production under sub-contracted operations, which has been sold (invoiced) to the principal in line under the conditions for sub-contracted operations (CPA rev. 2.1 Handbook part 6 - Outsourcing). The production may be carried out under the primary or secondary activities of the enterprise.
Variable 251003: The actual production includes any production which, either in the kind-of-activity unit itself, or in another kind-of-activity unit belonging to the same enterprise are intended for sale, or are processed into another product, or are fitted into another product and or are put into stock.
3.5. Statistical unit
The enterprise (within one country) is the observation unit surveyed by the National Statistical Institute.
The production for each Prodcom product from all surveyed enterprises in the country is aggregated before the results are sent to Eurostat.
3.6. Statistical population
The survey population of the reference period shall be enterprises whose principal activity or one of its secondary activities is listed in section B, C or E of the classification of economic activities in the European Community (NACE Rev.2).
3.7. Reference area
European Union + Norway + Iceland
Candidate countries: Albania, the Republic of North Macedonia (MK), Montenegro (ME) and Serbia (RS)
Potential candidates: Bosnia and Herzegovina (BA)
Each country provides the production statistics relating to their country: the Member States, BA, ME, MK, RS and EEA-EFTA countries (Norway and Iceland) transmit Prodcom data to Eurostat.
3.8. Coverage - Time
Since 1995 for annual data.
3.9. Base period
Not applicable.
Production is measured by value (in thousands Euro) and by the volume (kg, m2, number of items, etc) that is appropriate for the product.
The sold production value is collected in thousands of national currency (as stated in the Commission Implemented Regulation (EU) 2020/1197) and disseminated in Euro in the Eurostat Data Browser.
For more info for the volume - please see the below annex.
Annual data (for all products); reference year 2023.
6.1. Institutional Mandate - legal acts and other agreements
EUR-Lex - 32020R1197 - EN - EUR-Lex (europa.eu) 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.
EUR-Lex - 32019R2152 - EN - EUR-Lex (europa.eu) 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.
EU implementing regulation 2022/2552 on technical specifications of data requirements for the detailed topic industrial production statistics establishing the industrial products classification breakdown.
COMMISSION REGULATION (EU) 2019/1933 of 6 November 2019 establishing the ‘Prodcom list’ of industrial products referred to in Council Regulation (EEC) No 3924/91 (Text with EEA relevance)
COMMISSION REGULATION (EU) No 2119/2017 of 22 November 2017 establishing the ‘Prodcom list’ of industrial products provided for by Council Regulation (EEC) No 3924/91
COMMISSION REGULATION (EU) No 1872/2016 of 6 October 2016 establishing for 2016 the ‘Prodcom list’ of industrial products provided for by Council Regulation (EEC) No 3924/91
COMMISSION REGULATION (EU) No 1711/2015 of 17 September 2015 establishing for 2015 the ‘Prodcom list’ of industrial products provided for by Council Regulation (EEC) No 3924/91
COMMISSION REGULATION (EU) No 824/2014 of 4 July 2014 establishing for 2014 the ‘Prodcom list’ of industrial products provided for by Council Regulation (EEC) No 3924/91
COMMISSION REGULATION (EU) No 936/2013 of 12 September 2013 establishing for 2013 the ‘Prodcom list’ of industrial products provided for by Council Regulation (EEC) No 3924/91
COMMISSION REGULATION (EU) No 907/2012 of 20 August 2012 establishing for 2012 the ‘Prodcom list’ of industrial products provided for by Council Regulation (EEC) No 3924/91
COMMISSION REGULATION (EU) No 830/2011 of 27 July 2011 establishing for 2011 the ‘Prodcom list’ of industrial products provided for by Council Regulation (EEC) No 3924/91
COMMISSION REGULATION (EU) No 860/2010 of 10 September 2010 establishing for 2010 the Prodcom list of industrial products provided for by Council Regulation (EEC) No 3924/91
COMMISSION REGULATION (EU) No 163/2010 of 9 February 2010 establishing for 2009 the Prodcom list of industrial products provided for by Council Regulation (EEC) No 3924/91
COMMISSION REGULATION (EC) No 36/2009 of 11 July 2008 establishing for 2008 the Prodcom list of industrial products provided for by Council Regulation (EEC) No 3924/91
COMMISSION REGULATION (EC) No 1165/2007 of 3 September 2007 establishing for 2007 the Prodcom list of industrial products provided for by Council Regulation (EEC) No 3924/91
COMMISSION REGULATION (EC) No 294/2007 of 20 February 2007 establishing for 2006 the 'Prodcom list' of industrial products provided for by Council Regulation (EEC) No 3924/91
COMMISSION REGULATION (EC) No 317/2006 of 22 December 2005 establishing for 2005 the 'Prodcom list' of industrial products provided for by Council Regulation (EEC) No 3924/91
COMMISSION REGULATION (EC) No 210/2004 of 23 December 2003 establishing for 2004 the "Prodcom list" of industrial products provided for by Council Regulation (EEC) No 3924/91
COMMISSION REGULATION (EC) No 347/2003 of 30 December 2002 establishing for 2003 the .Prodcom list. of industrial products provided for by Council Regulation (EEC) No 3924/91
6.2. Institutional Mandate - data sharing
None.
7.1. Confidentiality - policy
EUR-Lex - 02009R0223-20241226 - EN - EUR-Lex (Chapter V - Statistical confidentiality) of 11 March 2009, 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
Reporting countries mark some data as confidential and Eurostat is legally bound to suppress those data from dissemination. However, Eurostat is able to publish EU totals containing confidential data if this can be done without revealing the confidential items within the total. A Confidentiality charter 2022 agreed with the Member States specifies the criteria which must be met before an EU total can be considered safe to be published.
In the past, unsafe EU totals were suppressed from dissemination. However, a procedure is now used to apply "controlled rounding" to such totals. This has the effect of giving a range within which the true total lies, without being so precise that the confidential data is revealed. For instance, instead of publishing a real value of 52.178, a value of 52.000 with base value of 500 might be published, indicating that the true value lies somewhere between 51.500 and 52.500.
8.1. Release calendar
The deadline for Member States to transmit data to Eurostat is 30 June of the reference year + 1. Eurostat first publishes this data in mid-July, two weeks after the deadline. However reporting countries may submit revisions at any time, so Eurostat also republishes the data in the middle of any month when new data has been received since the last publication. In practice this means that data is republished in most months, sometimes even for earlier years.
The revised NACE Rev. 2 classification is applicable to the PRODCOM Statistics since 2008 reference year. For the reference years 1995 to 2007 the PRODCOM data collected by a NACE Rev. 1.1 PRODCOM List classification has been converted where possible to NACE Rev.2. For 2008 onwards the PRODCOM statistics are collected and supplied by the reporting countries in NACE Rev. 2. classification.
Note: The conversion is incomplete since it can only be done when there is a direct correspondence between the old and new codes or where two or more codes are merged into one.
Before sending data to Eurostat, Member States apply a number of micro and macro plausibility checks at different aggregation levels, between historic and current data and with other sources. After the official data transmission, Eurostat provides the validation report process through EDIT to Member States.
11.2. Quality management - assessment
Data is collected and structured according to agreed specifications that are required to transmit data in a specific data format (SDMX). Eurostat performs standardized validation services (STRUVAL and CONVAL) that support the automatisation of the data validation process.
STRUVAL service verifies the structure of the dataset. CONVAL service completes the validation of the content of statistical datasets based on validation rules and constraints formulated by Eurostat.
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
The European Commission and the national governments need data to monitor industry and markets and to develop their corresponding policies. To meet these requirements, Eurostat and Member States have developed the Prodcom system and disseminate data which allow international comparisons between all Member States and other countries. The enterprises benefit from data provided by the Prodcom system which allow them to evaluate markets and opportunities for development.
The most important categories of users are:
the European Commission,
the national governments and their national authorities (i.e. public institutions, central and local administrations),
the enterprises and trade associations,
the research institutions and students,
the media.
12.2. Relevance - User Satisfaction
Confidential data: a lot of data has to be suppressed due to confidentiality, reducing the value of the statistics. Eurostat introduced a system of controlled rounding of EU totals that would otherwise be suppressed, greatly increasing the number of published EU totals.
Missing data: in the past, in cases where a country was unable to report data for a product, not only was the national data unavailable, but the EU total could not be calculated either. Eurostat recommended a system of estimating missing data so that affected EU totals can be calculated and published. This measure, combined with the controlled rounding of EU totals, means that EU totals are now published for all products.
Data availability: the deadline for transmission of data for reference year t is t + 6 months. Given that the enterprises cannot be asked to report their annual production until the relevant year is complete, this is a very tight deadline for the reporting countries, which must wait for the responses from enterprises, chase up non-response, compile the national data and apply quality checks. Eurostat is able to publish full data in mid-July of year t + 1.
Data accessibility: Data are available online in Eurostat's Database (Data Browser).
12.3. Completeness
Following the receipt of data from reporting countries, a "compliance score" is calculated. This combines the timeliness of the transmission with the percentage of data reported. The EBS Harmonised Compliance Monitoring and Reporting is applicable for the PRODCOM data transmission in 2023.
In the past, nearly all countries include 100% of the data in their transmission file, with all exceeding 85%.
13.1. Accuracy - overall
The Member States decide themselves which sampling design they apply to collect data for Prodcom survey.
The Commission Implementaing Regulation (EU) 2020/1197 [1] (EBS Regulation) is the main source for maintaining the sampling frame of business surveys. Consequently, the sampling frame for the Prodcom survey is drawn from the EBS Regulation (EU) 2020/1197 and contains all enterprises, authorities and organisations that carry out any target industrial activity during the reference period and within the economic territory of each country.
The overall national sample of enterprises has to be designed in such a way that it leads to representative results at the level of the national economy and of each product included in the Prodcom List (i.e. heading level).
Non-probabilistic sampling is applied in Prodcom surveys and the sample is a cut-off sample. Very few countries draw up a probabilistic sample according to the simple random sampling design for small enterprises. The sample of small enterprises is drawn up especially for those NACE classes where it is impossible to cover at least 90% of national production.
Sampling errors do not occur in the Prodcom survey, as these statistics are collected by non-probability sampling which means a full coverage survey with a cut-off threshold (i.e. cut-off sample). The part of the sample drawn up by a simple random sampling design is so negligible that it is not needed to collect the corresponding CVs (CV =coefficient of variation computed as a ratio between the square root of the estimate of the sampling variance and the estimated value).
13.3. Non-sampling error
Coverage errors
Coverage represents the extent to which the target population coincides with the frame population.
The coverage for the Prodcom survey is impossible to determine because the targeted population should be defined in terms of all enterprises which manufacture goods included in the Prodcom List. No reporting country has the resources to produce a corresponding sampling frame. Consequently, since the European Business Statistics (EBS) does not offer information on products but only on economic activities, the coverage of Prodcom survey is generally assessed by using enterprise turnover as a reference.
The approach most used by reporting countries to assess the coverage of Prodcom survey is in terms of turnover (or total production value) as the ratio between
total turnover (or total production value) of all observational units in NACE Rev. 2 07-33 and 38.32 in the Prodcom sample and,
total turnover (or total production value) of all relevant enterprises (i.e. enterprises with primary and secondary activity within NACE Rev. 2 07-33 and 38.32 in the SBS population.).
Additionally, there are coverage errors due to the divergence between the quality of the SBR as the source for the sampling frame and the effective enterprise population, namely:
under-coverage: there are targeted population units which are not accessible via the sampling frame;
over-coverage: there are units accessible via the sampling frame which do not belong to the target population (i.e. out of scope units);
multiple listing: some units may be present more than once in the sampling frame
By comparing the Member States' Quality Reports we can estimate that the overall coverage error at EU27 level is under 10%.
Measurement errors
The main reason that a measurement error can occur is erroneous information on the products produced by an enterprise. The enterprises are not using the Prodcom List for their own needs and it could be hard for an enterprise to update its internal registries in order to report according to Prodcom demands. More precisely, a measurement error by the respondent may concern:
wrong industrial product codes;
wrong measurement unit or reporting a volume unit inconsistent with the Prodcom list;
wrong conversion from one measurement unit to another;
which heading has to be reported (sold and/or actual);
reporting of work under subcontracting, and so on.
The approach most used to minimise the measurement errors is to use customised questionnaires by pre-printing Prodcom codes as far as possible. Also, particular attention is paid to providing the respondent with professional guidance and keeping a close contact with them.
As a very efficient and powerful approach to minimising errors in general and measurement errors in particular, Member States are developing electronic questionnaires and using internet facilities as far as possible. However this depends heavily on the resources and internal environment of each reporting country.
Processing errors
Processing errors occur during the process from the point when data are collected to the point when they are ready to be analysed: coding data entry, data editing, imputation, and so on.
It is not possible to present a general description of data processing for Prodcom surveys due to the high diversity in practices in the reporting countries. All reporting countries fulfil similar tasks. The difference is in the logical order and means (i.e. scanning, manually entry, automatically entry, fax use, software used, etc.).
Therefore, a lot of micro and macro plausibility checks are performed at different aggregation levels, between historical data and actual data and with other sources (i.e. business surveys, STS1 and SBS2 especially). These operations are highly interconnected and interdependent and could be grouped as follows:
logical checks;
checking coherence at a single product level by comparing the value of different indicators (i.e. mean value, minimum/maximum value, median value) for a given variable (i.e. 251001, 251002, 251003);
checking coherence between different variables (i.e. sold/sub-contracted/actual production) at different aggregate levels (i.e. observational unit, observational unit and product, product, and so on);
checking coherence between monthly/annual data;
checking coherence with external sources (i.e. STS data, SBS data, foreign trade data).
There are also cases where some micro data (e.g. sensitive units and/or products) are individually checked by the statisticians on the basis of their experience and information about the individual unit and the economic branch concerned.
Missing data is completed with extra enquiries by phone or e-mail, with estimates based on previous data or turnover data or with data obtained from the internet home pages of enterprises.
As with measurement errors, the past experience of most of reporting countries has shown very clearly that processing errors can be minimised to a great extent when electronic facilities (i.e. electronic questionnaires, web-based questionnaire, scanning, electronic/online checks, and so on) are used.
Non-response errors
Non-response results from a failure to collect complete information on all units in the selected sample except non-eligible units (i.e. dead units, out-of-scope units and non-contacts). There are two types of non-response. Firstly, a sampled unit that is contacted may fail to respond; this is called "unit non-response". Secondly, the sampled unit may respond incompletely to the questionnaire; this is called "item non-response". The unit non-response is adjusted by imputation methods and auxiliary information where it is available. The item non-response is adjusted either by imputation methods or estimation methods.
Normally, no imputation or estimation methods should be applied for new observation units (i.e. units included for the first time in the survey and no time data series is available for them) which are not in the position to respond within the reporting period.
1. STS = Short Term Statistics.
2. SBS = Structural Business Statistics
14.1. Timeliness
NSIs are required to transmit production data to Eurostat within 6 months of the end of the reference year for annual data.
14.2. Punctuality
Over the last few years, punctuality of data transmission has improved enormously. Now, all reporting countries transmit data by the deadline. This enables Eurostat to publish the PRODCOM data for the first time in July, 3 weeks after the deadline.
15.1. Comparability - geographical
By its nature, there is no objective way to compare levels of production of a product in different countries. However, Eurostat assumes that there will be some correlation of unit price (value/volume) between countries.
Eurostat calculates the unit price for each product for each country and takes the median value over the observations for two years. When new data arrives, the unit prices are compared with the corresponding medians and outliers are identified.
Large deviations from the median are not necessarily the result of an error, but the reporting country is informed of such cases and asked to check the data. This sometimes results in corrections to the data, but sometimes a reasonable explanation can be given for the deviation.
15.2. Comparability - over time
Eurostat compares corresponding data over successive years, to identify big increases or decreases in production. A sudden change from one year to the next does not necessarily signal an error, but if the level reverts to the preceding norm in the following year there is almost certainly an error. Such cases are notified to the reporting country for checking and correction if necessary.
15.3. Coherence - cross domain
The value of production for all the products in the same 4-digit NACE class can be summed and compared with national turnover in that class according to Structural Business Statistics. This exercise requires access to confidential data.
The Commission Implementaing Regulation (EU) 2020/1197 [1] (EBS Regulation) underlines necessity to specify the data requirements for the production of business statistics in order to produce comparable data between Member States and achieve harmonisation across all Business Statistics domains.
For most products it is possible to obtain the related External Trade data from the Comext system, by taking data for all the Combined Nomenclature headings that match one Prodcom heading. Trade data reported by the same country is aggregated for all matching CN headings and for all partners, to obtain single import and export figures for each Prodcom heading. (For EU totals, where production and trade data for all EU countries is aggregated, only trade with countries outside the EU is included. This ensures that we always show the trade crossing the external borders of the territory in question.)
The main purpose of this is to publish the trade and production data together in Europroms, but it also enables us to calculate apparent consumption, defined as production + imports - exports. Occasionally the result of this calculation is negative. In reality this is not possible, so the presence of a negative result indicates a problem with either the production or external trade figures.
By their nature, production statistics can only be obtained by surveying producer enterprises; the data is not available from administrative sources. Therefore the statistics rely on questionnaires completed by enterprises.
The requirements imposed by European regulation have been considerably reduced in recent years: the number of headings has been reduced from nearly 7000 to less than 4000, the amount of detail per heading has been reduced and quarterly and monthly surveys have been discontinued. However, Member States often continue to conduct more detailed or more frequent surveys to satisfy national needs, so the reduction in requirements at European level does not always result in a reduction in the burden for enterprises.
For reference periods before 2021, the survey covered the physical volume of production and the value and volume of production sold. According to the requirements in the EBS General Implementing Act (1197/2020), the following three variables for industrial production are specified: 251001, 251002 and 251003. The "additional" variable 251002 on production under sub-contracted operations shall be completely seprated from the sold production variable.
Several surveys have been conducted by Eurostat and DG Enterprise to measure the burden imposed by statistical surveys, including Prodcom.
17.1. Data revision - policy
To further specify the general Eurostat revision policy, the following revision policy has been established for annual industrial production statistics (PRODCOM).
Eurostat accepts all new versions of data transmitted by reporting countries. When reporting errors or anomalies to NSIs, Eurostat expects them to send revised data to resolve the problems identified.
17.2. Data revision - practice
The revision practice effectively corresponds to the revision practice of the domain listed under sub‑concept 17.1 (data revision – policy).
All reported errors (once validated) result in corrections of the disseminated data.
Reported errors are corrected in the disseminated data as soon as the correct data have been validated.
Estimated data (flagged by E) are replaced with final data if transmitted and validated at a later date. Data of low reliability are not published. European aggregates are updated for consistency with new country data.
Whenever new data are provided and validated, the already disseminated data are updated. European aggregates are updated at the same time when country data revision is published.
Data are usually revised for the last reference year. When data error or anomaly occurs, the country revises the relevant reference year.
There are routine revisions aiming at adjusting annual time series to have temporal consistency with the data published at the national level.
Depending on the country, the data are considered to be final after the first transmission or after the first or second revision.
Routine revisions are documented in the revision calendar. The revision file is transmitted to Eurostat in the specified period.
For data of reference year 2022, the revision impact on the EU aggregate between the first data released in July 2023 and the first revision published in March 2024 was of 3.8644% for the total value of production sold, 2.9691% for the value of production sold and 25.2228% for the value of production under subcontracted operations.
Revision Jul Yr+1 / Mar Yr+1
Total sold production value
Sold production value
Value produced under subcontracted operations
Reference year 2019
- 0.1982 %
-
-
Reference year 2020
0.5471%
-
-
Reference year 2021
- 0.5613%
0.0995%
2.5705%
18.1. Source data
Prodcom data are collected by the National Statistical Institutes among enterprises. The data are collected through statistical survey, together with any other sources or use of estimates, if it is deemed appropriate by them.
The survey is conducted by questionnaire, targeted to enterprises likely to be producing particular products. This can be established from the activity classification of enterprises in the business register. There are targeted the enterprises having the principal or one of the secondary activity in Sections B, C and E of NACE Rev.2.
Some countries survey only a sample of the enterprises in the target population and then gross-up the results.
18.2. Frequency of data collection
The frequency specified by European legislation is for an annual survey. Some countries conduct more frequent surveys to satisfy national requirements.
18.3. Data collection
Prodcom data and metadata are produced by the National Statistical Institutes of the EU-27 (excluding CY, MT, LU), the European Free Trade Association (EFTA) countries (Norway, Island), Albania (starting with 2022), Bosnia and Herzegovina (starting with 2011), FYRO Macedonia (starting with 2009), Montenegro (starting with 2011), Serbia (starting with 2011), Turkey (only between 2005 and 2010) and transmitted to Eurostat by using EDAMIS.
There is one Prodcom coordinator in each reporting country in charge of transmission of these data. All organisations and providers are defined in EDAMIS.
The transmitted dataset is named "PRODCOM_A_A_XX_YYYY_0000_V0001.CSV (SDMX-CSV or XML format)" ([XX] country code and [YYYY] reference year, V0001 – version number).
The transmitted national metadata is named "PRODCOM_ESMS_A_XX_YYYY_0000" ([XX] country code and [YYYY] reference year).
Eurostat sends reminder(s) in case of missing transmission after the deadline (t+6 months for dataset; t+8 months for national metadata files).
The dataset might include confidential data and therefore, the standard encryption facility of EDAMIS is available for National Statistical Institutes (public/private keys managed by unit G3).
18.4. Data validation
Eurostat facilitates the Pre-Validation services for reporting countries to test their datasets before the Official Transmission. It eliminates the need to implement the validation services on local premises and it replicates the exact same validation rules, which are used for Official Transmission. The data provider submits the dataset PRODCOM_A_A via EDAMIS and the structure validation is checking the structure and organisation of the dataset. The next content validation service enables the content validation (PRODCOM-specific) of the statistical data files based on a set pre-definied validation rules. Afterwards, the dataset is succefully processed in Eurostat database. In case of other errors or unexpected results (post-validation checks), the responsible domain manager contacts the person who sent the data (identified in EDAMIS).
Several plausibility checks are applied (by the National Statistical Institutes and Eurostat) to Prodcom data. Eurostat consults the NSIs in case of errors or anomalies.
18.5. Data compilation
Apart from disseminating national Prodcom data, several different European aggregates meaningful for a specific year (starting with 1993 reference year) are disseminated on the Eurostat website.
Estimates are made for missing data to ensure that EU totals can be calculated for all products. A set of rules, agreed with the Member States, is applied to ensure that no aggregates containing confidential data are published if this would allow a user to derive the national confidential data. For those aggregates that are unsafe to publish according to these rules, rounding is applied so that an approximate value within a range can be published without revealing the confidential national data included in the aggregate.
The amount of production of each product in the Prodcom List is recorded, normally by value and volume.
18.6. Adjustment
Reporting countries give value data in national currency ('1000) and Eurostat converts this to Euro where necessary.
Eurostat updates the published data once per month, to include any new or revised data received in the preceding month.
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.
From reference year 2021 onwards, the Prodcom data includes:
the physical volume of production sold during the survey period
the physical volume of production under sub-contracted operations during the survey period
the value of production sold during the survey period
the value of production under sub-contracted operations during the survey period
for some products, the volume of actual production during the survey period
The Prodcom data is obtained by the National Statistical Institutes (NSIs) who conduct a survey of enterprises.
Eurostat calculates EU totals at EU-15, EU-25, EU-27_2007 and EU-27_2020 level (depending on the year) from the national data.
For more details please see here below:
1. "INDICATORS metadata.xlsx" dimension metadata used online (COMEXT, Data Browser) is available in EN, FR, DE.
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.
The NSI in each reporting country carries out a survey of industrial production in that country, collates the results and transmits them to Eurostat. Eurostat calculates EU totals and publishes the national and EU data together with the related external trade data.
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.
Reporting countries have a major difficulty in identifying all the enterprises that produce a product. Although they can use the Business Register to find enterprises whose primary or secondary activities are linked to the product being surveyed, they can never be sure that they have identified all producers of the product.
The products: the Prodcom List defines the products to be included in the survey.
There are approximately 4000 headings representing manufactured products and some industrial services, in the following NACE Rev.2 sectors:
07 - Mining of metal ores
08 - Other mining and quarrying
10 - Manufacture of food products
11 - Manufacture of beverages
12 - Manufacture of tobacco products
13 - Manufacture of textiles
14 - Manufacture of wearing apparel
15 - Manufacture of leather and related products
16 - Manufacture of wood and of products of wood and cork, except furniture; manufacture of articles of straw and plaiting materials
17 - Manufacture of paper and paper products
18 - Printing and recording services
19 - Manufacture of coke and refined petroleum products
20 - Manufacture of chemicals and chemical products
21 - Manufacture of basic pharmaceutical products and pharmaceutical preparations
22 - Manufacture of rubber and plastic products
23 - Manufacture of other non-metallic mineral products
24 - Manufacture of basic metals
25 - Manufacture of fabricated metal products, except machinery and equipment
26 - Manufacture of computer, electronic and optical products
27 - Manufacture of electrical equipment
28 - Manufacture of machinery and equipment n.e.c.
29 - Manufacture of motor vehicles, trailers and semi-trailers
30 - Manufacture of other transport equipment
31 – Manufacture of furniture
32 - Other manufactured products
33 - Repair and installation services of machinery and equipment
38 (38.32) - Recovery of sorted materials
The reporting countries: those countries that conduct the survey and transmit the results to Eurostat. The countries are identified by the Geonomenclature codes.
The value and the volume of production. The value is expressed in national currency in thousands 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.
EBS variables on industrial production include the following data:
Variable 251001: The sold production is the production produced on own account and may be carried out under the primary or secondary activities of the enterprise. It includes production sold (invoiced) between different kind-of-activity units belonging to the same enterprise.
Variable 251002: The production under sub-contracted operations, which has been sold (invoiced) to the principal in line under the conditions for sub-contracted operations (CPA rev. 2.1 Handbook part 6 - Outsourcing). The production may be carried out under the primary or secondary activities of the enterprise.
Variable 251003: The actual production includes any production which, either in the kind-of-activity unit itself, or in another kind-of-activity unit belonging to the same enterprise are intended for sale, or are processed into another product, or are fitted into another product and or are put into stock.
The enterprise (within one country) is the observation unit surveyed by the National Statistical Institute.
The production for each Prodcom product from all surveyed enterprises in the country is aggregated before the results are sent to Eurostat.
The survey population of the reference period shall be enterprises whose principal activity or one of its secondary activities is listed in section B, C or E of the classification of economic activities in the European Community (NACE Rev.2).
European Union + Norway + Iceland
Candidate countries: Albania, the Republic of North Macedonia (MK), Montenegro (ME) and Serbia (RS)
Potential candidates: Bosnia and Herzegovina (BA)
Each country provides the production statistics relating to their country: the Member States, BA, ME, MK, RS and EEA-EFTA countries (Norway and Iceland) transmit Prodcom data to Eurostat.
Annual data (for all products); reference year 2023.
The Member States decide themselves which sampling design they apply to collect data for Prodcom survey.
The Commission Implementaing Regulation (EU) 2020/1197 [1] (EBS Regulation) is the main source for maintaining the sampling frame of business surveys. Consequently, the sampling frame for the Prodcom survey is drawn from the EBS Regulation (EU) 2020/1197 and contains all enterprises, authorities and organisations that carry out any target industrial activity during the reference period and within the economic territory of each country.
The overall national sample of enterprises has to be designed in such a way that it leads to representative results at the level of the national economy and of each product included in the Prodcom List (i.e. heading level).
Non-probabilistic sampling is applied in Prodcom surveys and the sample is a cut-off sample. Very few countries draw up a probabilistic sample according to the simple random sampling design for small enterprises. The sample of small enterprises is drawn up especially for those NACE classes where it is impossible to cover at least 90% of national production.
Production is measured by value (in thousands Euro) and by the volume (kg, m2, number of items, etc) that is appropriate for the product.
The sold production value is collected in thousands of national currency (as stated in the Commission Implemented Regulation (EU) 2020/1197) and disseminated in Euro in the Eurostat Data Browser.
For more info for the volume - please see the below annex.
Apart from disseminating national Prodcom data, several different European aggregates meaningful for a specific year (starting with 1993 reference year) are disseminated on the Eurostat website.
Estimates are made for missing data to ensure that EU totals can be calculated for all products. A set of rules, agreed with the Member States, is applied to ensure that no aggregates containing confidential data are published if this would allow a user to derive the national confidential data. For those aggregates that are unsafe to publish according to these rules, rounding is applied so that an approximate value within a range can be published without revealing the confidential national data included in the aggregate.
The amount of production of each product in the Prodcom List is recorded, normally by value and volume.
Prodcom data are collected by the National Statistical Institutes among enterprises. The data are collected through statistical survey, together with any other sources or use of estimates, if it is deemed appropriate by them.
The survey is conducted by questionnaire, targeted to enterprises likely to be producing particular products. This can be established from the activity classification of enterprises in the business register. There are targeted the enterprises having the principal or one of the secondary activity in Sections B, C and E of NACE Rev.2.
Some countries survey only a sample of the enterprises in the target population and then gross-up the results.
Annual (for all products) as required to publish revised data.
NSIs are required to transmit production data to Eurostat within 6 months of the end of the reference year for annual data.
By its nature, there is no objective way to compare levels of production of a product in different countries. However, Eurostat assumes that there will be some correlation of unit price (value/volume) between countries.
Eurostat calculates the unit price for each product for each country and takes the median value over the observations for two years. When new data arrives, the unit prices are compared with the corresponding medians and outliers are identified.
Large deviations from the median are not necessarily the result of an error, but the reporting country is informed of such cases and asked to check the data. This sometimes results in corrections to the data, but sometimes a reasonable explanation can be given for the deviation.
Eurostat compares corresponding data over successive years, to identify big increases or decreases in production. A sudden change from one year to the next does not necessarily signal an error, but if the level reverts to the preceding norm in the following year there is almost certainly an error. Such cases are notified to the reporting country for checking and correction if necessary.
