Comparability

Comparability - geographical

All EU member states produce Prodcom statistics. Manufacturers' sales of goods is the Danish Prodcom statistics and can be compared to those of other EU member states. Manufacturers' sales of goods differ in a few aspects from the Prodcom statistics described in the regulation and commonly implemented. Firstly, industrial services are not covered in the detail prescribed by the Prodcom list. This is only up to the year 2020. In 2021 the statistics changed and is now fully comparable with the Prodcom list. Though, coverage is not defined as production on the physical territory of Denmark, but by the economic ownership of goods sold and produced by Danish enterprises (cf. Geographical coverage).

Comparability over time

Historically, the statistics date back to 1905 as a survey measuring the industrial production. In 1944, the name was changed to Industrial Production Statistics. In 1967, the name became 'Product statistics for industry', and the statistics no longer measured production, but sales instead. From 1968, the survey has been conducted quarterly, before it became an annual survey. The present name, 'Manufacturers' sales of goods' was introduced in 2007.

The statistics use two sets of classifications: one for products and one for economic activities. Changes in the classifications affect comparability over time.

The classification of products is the 8-digit Combined Nomenclature (CN). The CN is changed every year, which normally only affects 100-600 of the groups. Years with major CN revision typically affect up to 2,000 groups.

CN codes have 8 digits and are regulated by the EU. Prior to 2008, the CN codes were, for selected product categories (e.g. wind mills), further divided as Danish national codes using a 9th and 10th digit. This was discontinued to reduce response burden.

Industrial classifications are changed only at long intervals. The current industrial classification, Dansk Branchekode 2025, was introduced in the statistics from 2025 and retrospectively applied back to 2021. Prior to 2025, the statistics were collected and published according to Dansk Branchekode 2007 (and earlier industrial classifications) and can be found in archived tables in the StatBank.

For the period 2000–2008, the statistics were reclassified according to Dansk Branchekode 2007, meaning that comparable industry-distributed data are available from 2000 onwards.

There are special circumstances for 2025 that mean the data are not directly comparable with 2021–2024:

Kind of activity units (KAU's): In 2025, Statistics Denmark carried out a consolidation of KAU’s, which generally means that there are fewer but broader KAU’s in 2025 than previously. This affects which local units reporting enterprises must include in their reporting for these statistics. For some respondents, this may mean that they must include their sales of trading goods in 2025 even though this was not previously required. It may also mean that a larger share of the enterprise’s total sales must be included in 2025 compared with earlier years. This affects the comparability of data between 2025 and previous years, as data for 2021–2024 are merely distributed according to Dansk Branchekode 2025, while the underlying data were collected under Dansk Branchekode 2007 and on more numerous and narrower KAU’s. However, the impact is assessed to be minimal to negligible in this context.

New definition of manufacturing enterprises and factoryless goods production: The new industrial classification of Danish enterprises – Dansk Branchekode 2025 (DB25) – entered into force on 1 January 2025. The new classification includes a new definition of factoryless goods production. This new definition may affect enterprise reporting in two different ways:

• Some enterprises that were not previously classified as manufacturing enterprises will now be classified within manufacturing industries under DB25. These enterprises may therefore become subject to reporting obligations based on this new classification.
• Some goods that previously had to be reported as trading goods must now instead be reported as own goods produced by a subcontractor (factoryless manufacturing).

The expanded definition of manufacturing reflects increasing globalization within manufacturing industries. Under the new definition, ownership of raw materials and/or production facilities is no longer the sole criterion for determining whether an enterprise is classified as a manufacturer.

Under DB25, a factoryless goods producer (FGP) is defined as a principal that: - controls the production of a good by carrying out the initiating activities and supplying the technical specifications (intellectual assets and rights) required to produce the good, BUT outsources all or part of the material transformation required to produce the good.

The new FGP definition in DB25 affects the definitions of production methods:

• Traditional manufacturing – unchanged definition: The enterprise owns the factory, raw materials, and finished goods. Revenue from traditional manufacturing continues to be reported as revenue from own-produced goods.
• Factoryless manufacturing – new, broader definition, applicable from 2025: The enterprise has its products manufactured at an external factory but owns the raw materials and finished goods. In this case, the enterprise effectively rents the labor force of the external factory and/or has its products manufactured externally while supplying the design, owning the intellectual property rights, or otherwise controlling the production process. Revenue from factoryless manufacturing where the enterprise owns the design, rights, etc. to externally manufactured goods must — regardless of ownership of raw materials — be included under goods produced by subcontractors. This revenue must be distributed across commodity codes, and quantities must be reported. Prior to this expansion of the ownership definition to include intellectual property rights, such production had to be reported as trading goods.

As a result, from 2025 onwards there are industries and type of turnover categories that show developments reflecting not only changes in sales, but also changes in the reporting population and ownership definitions compared with earlier years. The years prior to 2025 have not been grossed up to reflect the population or ownership definition used in 2025, as the statistics are based solely on reported commodity code data, etc., for the respective years. In particular, the turnover categories Sales of goods produced by subcontractors and Sales of trading goods are affected by the new ownership definition under DB25. Users should therefore exercise particular caution when comparing developments between 2025 and earlier years.

Comparability between archived tables and new tables:

Publication of 2025 data takes place in four new StatBank tables: VARER10, VARER30, OMS10 and INTJEN10.

At the same time, six StatBank tables have been discontinued: VARER, VARER1, VARER2S, VARER3, OMS5 and OMS6.

Since the statistics became annual in 2025, it is natural that the quarterly tables VARER, VARER2S, and OMS5 have been discontinued, as data are no longer collected quarterly. However, because DB25 entered into force and commodity codes from 2025 onwards are 8-digit rather than 10-digit, the annual tables VARER1, VARER3, and OMS6 were also discontinued from 2025 onwards.

It is still possible to create comparable time series between the new and archived tables as follows:

• VARER10 and VARER3 can be merged, disregarding the fact that VARER10 uses 8-digit commodity codes instead of 10-digit codes. The last two digits are often 00. In VARER10, the years 2021–2024 have been converted to 8-digit codes, and since VARER3 contains data up to and including 2024 using 10-digit codes, the linkage is straightforward for constructing long time series covering 1995–2025, subject to other reservations described in earlier statistical documentation.
• VARER30 and VARER3 can also be merged. The only reason for the existence of two tables is that 2025 data were collected under DB25 rather than DB07. For SITC groups, this has no impact, which is why data in VARER30 have also been recalculated back to 2021.
• OMS10 and OMS6 differ in that they were collected under two different industrial classifications, DB25 and DB07. There is no direct one-to-one comparability between the two, due to differences in ownership definitions and thus the underlying reporting populations, as previously described.

There are also differences in turnover categories between the two tables:

• Total turnover in both OMS6 and OMS10 is comparable.
• Sales of own commodities, total (OMS6) and Sales of commodities, total (OMS10): Sales of own commodities are the sum of production on own account and production outsourced to others. Sales of goods, total, are the sum of production on own account, production outsourced to others and production as subcontractor.
• Sales of production on own account in OMS6 and OMS10 are comparable.
• Sales of production outsourced to others in OMS6 and OMS10 are comparable, though subject to differences in ownership definitions under DB07 and DB25.
• Sales of production as subcontractor (OMS10): These sales are not separately identified in OMS6, but correspond to a combination of values on commodity codes under sales of production as subcontractor plus approximately one-third of values on service codes, together constituting Contract work for other enterprises in OMS6.
• Invoiced discounts, freight charges etc., not distributed to commodities are identical in OMS6 and OMS10.
• Total Industrial services: In OMS6, this equals the sum of all values on service codes plus the sum of values on commodity codes under sales of production as subcontractor. In OMS10, it equals only the sum of all values on service codes. Thus: Industrial services, total (OMS6) = Industrial services, total (OMS10) + values for Sales of production as subcontractor (OMS10).
• Construction work for others enterprises (OMS6) is comparable with service code values in INTJEN10 containing the wording “Installation of”.
• Reconditioning and repair work for others enterprises (OMS6) is comparable with service code values in INTJEN10 containing the wording “Repair and maintenance of”.
• Contract work for other enterprises (OMS6) is comparable with service code values in INTJEN10 not containing the wording “Installation of” or “Repair and maintenance of”, plus Sales of production as subcontractor in OMS10 (values only).
• Commercial (resale) turnover in OMS6 and OMS10 are comparable, though subject to differences in ownership definitions under DB07 and DB25.
• Other turnover in OMS6 and OMS10 is comparable.

Coherence - cross domain

The statistics may be compared to a number of other statistics that describe the manufacturing industry. In all cases there are however differences in units, definitions and/or coverage, which need to be considered when comparing. The main related statistics are:

• Industrial production and turnover: The statistics is a monthly index based on a sample. The statistical unit used is the Kind of Activity unit, which is also used in the manufacturer's sales statistics. The industrial production and turnover statistics does however not include commercial resale turnover.

• External trades in goods: Exports and imports data are available according to the same commodity codes as manufacturer’s sales, so data can be combined. One should be aware that the enterprise is used as the statistical unit in the external trade statistics. Another difference between the two statistics is that the manufacturers sales of goods includes goods produced abroad under subcontracting for Danish enterprises, while the external trade statistics only includes goods that cross the Danish border. On the other hand, commercial resale turnover is divided on commodity codes in the external trade statistics, which is not the case in the manufacturer's sales statistics.

• Purchases and sales by firms: The statistics are based on the enterprises’ VAT reporting to the Danish tax authorities. The variable Total sales consists of domestic sales added export sales, and is comparable with the total turnover in the manufacturer’s sales. Inconsistencies between the two statistics is typically due to the fact that the Purchases and sales statistics has the enterprise as the statistical unit and that the two statistics have different definitions on turnover. Another reasons for inconsistencies is that the main purpose of the purchases and sales statistics is to collect information on VAT payments, while the statistical use of the data is secondary. Furthermore, sales that are subject to VAT in another EU country, is exempt from VAT in Denmark and will not be included in Purchases and sales by firms.

-Accounts statistics: The statistics provides key economic figures, also for manufacturing industries. The accounts statistics is based on the enterprise as the statistical unit.

In addition, for mining and quarrying, it is possible to compare to the annual statistics on mining and quarrying (volumes, not values)

The difference between manufacturers' sales and production as measured in the National Accounts is mainly different definitions and calculations in the National Accounts to ensure full coverage (for units with less than 10 employees) and consistency.

Coherence - internal

Not relevant for these statistics.