Federal Statistical Office (DESTATIS)

Unit E22: Short-term statistics in the industry, Production of the industry

Statistisches Bundesamt

Gustav-Stresemann-Ring 11

D-65180 Wiesbaden

The PRODCOM survey is conducted by the 14 statistical offices of Germany's 16 federal states (Bundeslaender). It is a monthly and quarterly survey of the value and volume of products being produced during the reference period by local units in Germany. The annual PRODCOM statistics consists of the aggregated monthly and quarterly results. It is carried out in compliance with Commission Implementing Regulation (EU) 2020/1197 of 30 July 2020, laying down technical specifications and arrangements under Regulation (EU) 2019/2152 of the European Parliament and of the Council on European business statistics and the Regulation (EU) 2019/2152 of the European Parliament and the Council of 27 November 2019 on European business statistics.

The PRODCOM survey provides information on PRODuction of the European COMmunity and is compiled from data received from German manufacturers. 

Data of sold production on own account and the sold production produced under sub-contracted operation are collected as part of the national industrial production is not published at the national level in this breakdown.

• The European Classification of Economic Activities (NACE Rev. 2). This is the EU system for classifying economic activities.
• The Community Classification of Products by Activity (CPA) and the derived PRODCOM List 2021
• The German Classification of Industrial Products (Güterverzeichnis für Produktionsstatistiken 2019) - the German version of the PRODCOM list.

To view the national classification, please see https://www.klassifikationsserver.de/klassService/jsp/common/content.jsf

The survey covers all local units belonging to a legal unit with 20 or more employees, which are totally or primarily engaged in industrial production and industrial services in the Mining, Quarrying, and Manufacturing industries (NACE Rev.2 sections B andC). Additionally, all local units with 20 or more employees belonging to an enterprise (legal unit) outside the NACE Rev.2 sections B and C are also covered. The survey covers 93 percent of total production. The estimated covered production comes from a Eurostat project that estimated the production value of all enterprises with less than 20 employees using administrative data and a regression model.

The purpose of this statistic is to report for each product in the PRODCOM List the amount and value of production carried out within the territory of the reporting country during the reference year. This means that PRODCOM relates to products (not 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 on aid to identify the units that should be surveyed to determine the product's production.

The PRODCOM List defines the products to be included in the survey. There are approximately 4.000 headings representing manufactured products and some industrial services.

The observation unit is local unit.

The reporting is the legal unit. 

The survey population consists of all local units belonging to a legal unit with 20 or more employees, which main activity is in mining and quarrying or in manufacturing (NACE Rev.2 sections B and C). Furthermore, the survey population includes all local units with 20 or more employees belonging to a legal unit outside the NACE Rev.2 sections B and C.

PRODCOM statistics are available on the national level.

In the national dissemination, industrial production is also only available nationally.
Each of the 14 regional statistical offices in Germany carrying out the survey publishes their production statistics on the regional level of the Federal States. The regional statistical offices are responsible for this dissemination.
Regional data beyond the regional level of the federal states is not available.

The reference period of PRODCOM statistics is one year.

German industrial products statistics date back to 1948.

Not applicable.

For volumes: The volume is expressed in a volume unit specified for each product, such as kilograms, square meters etc. It is used to indicate the volume of the goods. For a comprehensive enumeration of units used in national publications, see the following list of units of measurement: https://www.destatis.de/DE/Methoden/Klassifikationen/Gueter-Wirtschaftsklassifikationen/klassifikation-gp-19.html.

For values: The production value is measured in thousands of euros.

Reference period is a calender year. Latest data refers to the year 2022.

This survey is carried out in compliance with the following regulations:

At the EU level: - Regulation (EU) 2019/2152 of the European Parliament and of the Council of 27 November 2019 on European business statistics.

                    - Commission Implementing Regulation (EU) 2020/1197 of 30 July 2020 laying down technical specifications and arrangements under Regulation (EU) 2019/2152 of the European Parliament and of the Council on European business statistics.

At the National level: Bundesstatistikgesetz (BStatG) (date 22 January 1987), Gesetz über die Statistik im Produzierenden Gewerbe (ProdGewStatG) (date 21 March 2002);

Not applicable.

The German confidentiality policy is based on:

National Level:

• Federal Statistics Law (BStatG), § 16 
• Statistics Law on manufacturing industry (ProdGewStatG), § 10
• Restriction of Competition Act (GWB), § 47

Note: If national data have previously been classified as confidential to Eurostat, they should not subsequently be disseminated in other public documents.

Germany is applying the principle of active confidentiality.

The following data are confidential:

  1. individual statistical data (except anonymized individual data for scientific purposes);

  2. statistical data which can be matched in a way that enables the identification of a specific statistical unit;

  3. statistical information which aggregates data for less than three legal units;

  4. statistical information that aggregates data with one or two dominant legal units. The rule is: If the difference between the total value of a cell, and the second largest single value exceeds the largest single amount by less than p %.

  5. Cross-over confidentiality: If the dissemination of one PRODCOM Code would reveal data of the German production codes (GP) in the national dissemination, than the PRODCOM code has to be confidential as well.

The policy of German Statistics on confidentiality of PRODCOM data is that a specific code is flagged as confidential if at least one of the mentioned criteria apply to the PRODCOM code.

About 31% of all results of the German data for PRODCOM are confidential.

Data of the German industrial products statistics are released regularly and without a particular announcement of the release date.

Not applicable.

Not applicable.

National dissemination:

• Monthly national publication of a selection of product aggregates on users' behalf.
• Quarterly national publication for the Sold Production and Total Production (volume and value) for all the national production codes of the GP, obtained from the monthly and quarterly survey.
• Annual national publication that aggregates the monthly and quarterly results.
• In addition, the monthly collected data (local units with 50 or more employees belonging to the manufacturing industry) are part of the monthly calculated production index.

Eurostat dissemination:

•  PRODCOM statistics are submitted annually to EUROSTAT and are also published annually

There can be ad-hoc press releases or a dissemination referring to the “number of the week”, like: https://www.destatis.de/DE/Presse/Pressemitteilungen/Zahl-der-Woche/2022/PD22_31_p002.html

Till 1952 production data for specific important production codes/product aggregates are published in the German Statistical Yearbook in paper format.

Starting with 2002, data have been made available online at https://www-genesis.destatis.de/genesis/online

Since the introduction of the Research Data Centre (RDC) in 2007, scientists can get access to official statistical microdata: www.forschungsdatenzentrum.de (German only)

The entry for the German industrial products statistics is at the portal to Germany’s official statistics: https://www.destatis.de/DE/ZahlenFakten/Wirtschaftsbereiche/IndustrieVerarbeitendesGewerbe/IndustrieVerarbeitendesGewerbe.html

The content includes a combination of HTML webpages for narrative, charts and graphs, with data being provided in usable formats such as CSV and Excel. Additionally, information on quality and methodology, FAQ (frequently asked questions) and a glossary can be found here.

For users, the main documentation on methodology and quality is the quality report in German:

https://www.destatis.de/DE/Methoden/Qualitaet/Qualitaetsberichte/Industrie-Verarbeitendes-Gewerbe/produktionserhebungen.html

See 11.1

The Quality guidelines of the German Statistical Office ensures a high quality of data through a variety of systematic quality assurance measures, such as the quality standards in German official statistics: https://www.destatis.de/EN/Methods/Quality/QualityAssurance/QualityAssurance.html

The German Statistical Office applies a comprehensive quality management system based on Total Quality Management (TQM).

The underlying quality guidelines are the "European Statistics Code of Practice" and the "Quality Standards of Official Statistics".

Primary users of PRODCOM/industrial products statistics data are:

• Besides Eurostat and other international users like the United Nations, data is necessary for national purposes and therefore required by ministries and other governmental bodies.
• Also, non-governmental users can be found like associations representing different economic branches or industries, the media or researchers and students.
• The data is also used in other statistics and internal processes, such as national accounts or price statistics.

Every second year we have a committee of experts where users, leading associations from different industries, and representatives of the field of industry statistics of the Federal Statistical Office come together to discuss statistical issues to improve statistics. Production statistics are also a part of these regular meetings. This committee of experts is the only user feedback we get since there is no user satisfaction survey for these statistics.

The PRODCOM data delivered to Eurostat is not fully complete:
This is because we have a national classification that differs in some aspects from the PRODCOM List. For some PRODCOM codes, we cannot deliver the volume of total and sold production because the national classification has a different measuring unit which cannot be converted into the measuring unit asked by PRODCOM List. Also, the national classification cannot change as fast as PRODCOM list does. When there are new products or new segmentations of products (e.g., splits), it is possible that the national classification does not include those new products or new segmentations. Because the production survey is based on the national classification and not the PRODCOM list, for some codes, we cannot deliver figures at all, or we have to estimate the shares of a national product code for PRODCOM data. In 2019 we had a significant update of the national classification where Germany adopted the national classification to PRODCOM list asgood as possible. But national user needs are considered as well. So we cannot apply the PRODCOM list in every case. Generally, the national classification is more detailed than the PRODCOM list (9 digit code). That means we also have additional information on some products not available on the Prodcomlevel.

As this survey is a census of local industrial units belonging to an enterprise with 20 or more employees obtained from the national statistical Business Register, all companies in the relevant NACE groups of sections B and C are surveyed.

1. Response-rate:

At the end of the reporting year, the response rate is not applicable because due to imputation for non-response, there are no local units with missing data in the dataset. If imputation is not considered for the response rate, then the imputation rate and response rate are equal in the German system of compiling production statistics.

2. Item non-response-rate

The item non-response does not apply. In Germany, we have an online survey. If a reportable enterprise reports under a certain national product code, every variable has to be filled out in an online survey. If one item is missing, the data cannot be transmitted, and the questionnaire cannot be finalized. If a local unit doesn't report under one product code, although the local unit is producing it, it is hard to detect it as a "missing item". In case a local unit no longer reports a specific product code, we contact the local unit and ask whether it has been forgotten, for example, or if it is no longer being produced.

3. Imputation-rate

In 2022, the imputation rate of the PRODCOM survey for Sales production value was 2.2% (2020: 2.3%) and for Sales production volume was 2.1% (2020: 2.2%).

The survey is done as a cut-off survey; all units above a certain threshold are included. Since there is no grossing up for units below this threshold, there is no sampling error.

Possible errors may be:

1. Local units may be missing if they are not registered as part of the manufacturing industry, even though they belong to the manufacturing industry (undercoverage).
2. In the case of delivering incorrect data or the incorrect classification of products, measurement errors and data processing errors may occur. Therefore, a plausibility check is taken place.
3. In case of unit non-response or delayed answers, estimation values are inserted in the monthly production survey.

PRODCOM data are submitted to Eurostat within t + 6 months.
The German industrial production statistics data are published quarterly within t + 3 months.

The statistics are published on the scheduled date. At least in the last three years, there have been no late publications.

The production survey is performed with the same methods for all German Federal States. The geographical comparability is given fully on the national level.
The results of PRODCOM are comparable to those of the European Union states.

The comparability over time is given in between one reference year. Due to methodological changes, there is a particular dynamic in
the production statistics limiting the comparability over time:

1. In general revision in NACE and changes such as product innovations and the general revision of NACE will regularly lead to changes in the Germany product classification (GP) about every seventh year. An extensive revision of NACE in 2008 has resulted in changes to the Germany product classification (GP) for the 2009 survey and onwards. In 2019, we introduced the Germany product classification (GP) 2019. There are new products and new segmentations of products. In general, the classification adapted more to the PRODCOM list. New codes or new arranged codes that arise with new versions of classification comparability over time is limited. With new or rearranged codes appearing with new versions of the classification, comparability decreases over time. The same applies to the PRODCOM list, which changes about every second year, with sometimes minor and sometimes major changes.

2. The basic survey population can change because of legal changes in statistical law to reduce enterprises' burden: E.g. Since 2007 in some small structured branches, the threshold for local units increased from 10 to 20 employees.

3. Every fifth year, there is a revision of the HS (Harmonized System), which will also result in more significant changes to CN and Prodcom. The latest major revision was in 2012.

4. Production in wage labor is identified by a "2" as the 10th digit of the actual 9-digit national commodity number. Thus, there is no separate commodity number for this type of production

In general there are no other official statistics where production is surveyed.

1. Coherence with Short-Term Statistics (STS):

There are consistency checks with the monthly report of local units of NACE sections B and C because both statistics have the same survey population. The German Federal States, responsible for surveying the local units, compare the sold production with the turnover (monthly / quarterly) on the micro data level. If there is an unusual divergence between these two values, the local unit is contacted to check the data, correct the data, or explain the differences.

2. Coherence with Structural Business Statistics (SBS):

There are no consistency checks with SBS because definitions and methods are too different. Enterprises are surveyed instead of local units, and the SBS for NACE sections B and C is a sample that is extrapolated to calculate results for the whole target population. These and other reasons explain why the returns and sold products can differ. Since significant differences are usual, the comparison does not take place.

3. Coherence with trade:

There are no consistency checks with trade statistics. Data of production statistics are comparable with the external trade statistics results concerning the product types' definitions. So the calculation of the domestic availability of products is possible with some restrictions.

4. Coherence with national accounts:

There are no consistency checks with national accounts. The national accounts use the results of the production statistics. The national accounts are one primary user of the production statistics.

National Production data is coherent within itself. The online survey and other validation checks ensure this internal coherence.

The burden on respondents for the industrial production statistics was calculated to be 16.6 million euros in 2017. Altogether, about 45.000 local units were reported for the production statistics in 2017. About 24.300 are reporting quarterly and 20.700 monthly.
Since using electronic data transmission methods offered to the respondents (IDEV, sStatistik.Core), the burden for the enterprises was reduced significantly. The cost-reducing effect of electronic data transmission was calculated to be 1.9 million euros. 

There is no cost calculation available for the production statistics concerning the burden of the statistical agencies in Germany.

The data submitted at the end of June is final and will be revised if necessary.

When relevant, the Federal Statistical Office will provide revised versions of the previous year.

The survey population is selected from the statistical Business Register. The Register contains all local German units with their industry classification, number of employees, and other information.

A cut-off threshold generates the survey population. All local units belonging to a legal unit with 20 or more employees classified as NACE divisions 07 through to 33 are included (NACE Rev.2 sections B and C). Also covered are local units that belong to legal units of other NACE sections with 20 or more employees. The design reduces the burden on smaller businesses.

Data are collected monthly from local units with 50 or more employees and quarterly with 20 or more employees. The latter is, therefore, not part of the monthly survey.

National data collection is decentralized and organized by the 14 regional statistical offices. The data is collected through an online survey. The online survey and instructions for companies are available at
https://erhebungsportal.estatistik.de/Erhebungsportal/# The online system includes consistency checks and validation of the data.
Each respondent receives an individual document based on the previous year's report. The reported production code of the previous period is thus pre-assigned.

The data are validated through several validation checks, for example, consistency checks over time for production value and volume for each respondent and consistency checks over time for the unit price of each code. Reporting under product codes that are not typical for the NACE code activity of the reporting unit is also controlled.
Data validation checks are performed on a regional and national level. Conspicuous values are checked along with the reporting unit.

There is an internal IT system for validation checks and editing data of the respondent. Once the data has been scanned, edits are made, and any problems or discrepancies with the data are checked and corrected if necessary.
The PRODCOM data are generated based on the national production survey. The annual national data with the national codes on 9-Digit levels are transformed to the 8-Digit PRODCOM code level and checked again to ensure their quality before transmitting to EUROSTAT. As the PRODCOM survey population is selected from the statistical Business Register, any issues detected are feedback to enhance the register’s quality.
In the event of technical or methodological changes, the staff is trained to be familiar with the survey processes.

At the national level, the production index is published monthly. The production index contains the production of the NACE sections B, C, D, and F.

No comment.