Enterprises

Enterprise: An enterprise is defined as the smallest combination of legal entities that produce goods or services, and which to a certain extent have independent decision-making authority (autonomous). Legal entities that are not autonomous (e.g. auxiliary entities) are merged with the autonomous legal entity to which they provide services, so that they together form an enterprise.

Establishment: An establishment is defined as a locally delimited functional unit that mainly carries out activities within a specific industrial activity. An enterprise is a subunit of the enterprise. In Eurostat, this corresponds to "local kind-of-activity unit" (LKAU).

Activity: Activity is a measure of whether the enterprise was in operation or not, and to what extent. In the versions "enterprises with activity" and "registered enterprises", activity is characterized by employment or turnover during the year, and in the version "enterprises with employees", activity is characterized by the fact that the enterprise had employees during the year. In the "registered enterprises" version, the registration of the enterprise is also counted as activity.

New enterprise: New enterprises include all real establishing of enterprises, including first-time births, reactivations and other creations. Ownership changes are not considered new enterprises, since they are simply continuation of existing activity, whereas all other creations are.

Birth: Births include all new enterprises, except creations due to (change of ownership and) other creations.

First-time birth: An enterprise is considered as first-time born in the version "enterprise with activity", "enterprise with employees" and "enterprise with activity (excluding contract workers) " in the first year it had activity, and in version "registered enterprise" in the validity year of the first information about the enterprise in Statistics Norway’s Business register (VoF) or the year of foundation (applies to AS only) if this was earlier. Ownership changes are never regarded as first-time births, but other creations might be.

Reactivation: Reactivations are enterprises that become active again after at least 2 years of inactivity. Ownership changes are excluded.

Ownership change: A change of ownership takes place in the year an enterprise starts or resumes activity, if the following three additional criteria are met: 1) The enterprise takes over in the same year, or the year before, all its activities from only one other enterprise. 2) The other enterprise transfers all its activities to the first. 3) The other enterprise is closed down at the same time or becomes inactive for at least 2 years. A change of ownership is in practice a mere continuation of an existing activity, and is therefore not considered a new enterprise.

Other creations: Other creations (i.e. except ownership changes) take place in the year an enterprise is new, if at least one of the following three additional criteria is met: 1) The enterprise takes over businesses from several enterprises in the same year, or the year before. 2) The enterprise that transfers activity, transfers it to several enterprises. 3) The transferring enterprise continues to be active. Other creations involve the creation of a new organizational unit which, to a varying degree, continues an existing activity. These other creations are therefore considered new enterprises, but not births. The acquisition of establishments from other enterprises in years other than the founding year is not considered a new enterprise. For other creations in connection with the termination of an enterprise, see death.

Survival: An enterprise born in year t has survived in year t+n (n>=1) if it had activity in t+n and continuous inactivity for a maximum of 1 year in the period between t and t+n. Survivals can be divided into two categories: 1. Direct survival occurs when it is the same enterprise that holds the activity in year t+n. 2. Survival by change of ownership occurs when another enterprise has taken over the activity in line with the definition of ownership change. The original enterprise's industrial code, size group, legal form and regional affiliation are retained.

Dormant enterprises: Dormant enterprises are characterized by the fact that they are inactive for one whole calendar year, and then become active again. Enterprises that are dormant in the last year with final figures or in the year with preliminary figures are temporarily classified as not surviving in these years, as there is a lack of reliable information about activity in the following year at the time the statistics are published. At the next publication, when there is reliable information about activity in the wake-up year, the enterprises are transferred to the category of dormant enterprises in the year in which they were inactive. From the year of awakening, they are again counted as survivors.

Awakened enterprises: Awakened enterprises are enterprises that have become active again, after previously being dormant.

Closure: An enterprise closure occurs in the last year in which the enterprise had activity, or in the last year of activity before a period of at least two years without activity. If the criteria for change of ownership are met, the cessation of an enterprise is not regarded as closure, since the entire activity has in practice been continued. Other terminations are considered closures.

Death: An enterprise death is a closure where none of the activity is transferred to another enterprise. Enterprises where activity was taken over by other enterprises in connection with the termination, in line with the definition of other creations, are considered as closures but not deaths.

Growth: Growth is defined as an increase in turnover or the number of employed persons from the previous year.

High-growth enterprises: High-growth enterprises are defined as all enterprises in the business demography that over a period of three years have an average annual growth of at least 10 per cent. Growth is measured both in turnover, employment and the number of employees. The enterprises must have at least 10 employees in the start year of the growth period, and be established no later than the year before the start year of the period, i.e. be at least 4 years old.

High-growth micro-enterprises: High-growth micro-enterprises are high-growth enterprises with 1-9 employees in the start year of the growth period. High growth is defined in two different ways: 1) As for normal high growth enterprises. 2) Growth of at least 3.31 employed persons or employees – in practice at least 4 due to rounding of micro data – during the three-year period. So far, only high-growth micro-enterprises according to definition 1) have been published.

Potential high-growth enterprises: Potential high-growth enterprises are all enterprises with at least 10 employees in the start year of the period, which were established no later than the year before the start year of the period, i.e. are at least 4 years old. Similarly for potential high-growth micro-enterprises.

Gazelles: Gazelles are 4 or 5-year-old high-growth enterprises.

Micro-gazelles: Micro-gazelles are 4 or 5-year-old high-growth micro-enterprises.

Potential gazelles: Potential gazelles are 4 or 5-year-old potential high-growth enterprises. Similarly for potential micro-gazelles.

Employed persons: The employment figures show an average of the number of employed persons in the reference week of five selected months of the year. Employed persons are defined as persons with income-generating work of at least one hour's duration in the reference week, as well as persons with temporary absence from such work, and include employees and self-employed persons (i.e. employed owners). A person may be employed in several enterprises.

Employees: Employees are defined as anyone who receives compensation in the form of salary etc. for work of at least one hour's duration in the reference week, including contract workers. Self-employed persons are not included.

Contract workers: Contract workers include freelancers and people who receive a fee for work of at least one hour's duration in the reference week.

Turnover: Turnover is defined as the sum of remuneration for sales to customers, advances from the sale of merchandise and gross income from other commercial activities. Turnover includes rental income and commission income, but not public grants, value added tax or gains from the sale of fixed assets. In some industries, the concept of turnover is not clearly defined. This applies to financial activities and insurance (industry K), education (industry P), health and social services (industry Q) and activities of membership organization (part of industry S). In order to determine whether enterprises in these industries are active, if they lack employees, alternative sizes are used as a supplement, such as e.g. operating income.

Legal form: Equivalent to legal form in the Central Coordinating Register for Legal Entities (CCR). Here, a distinction is made between between sole proprietorships (ENK), limited liability enterprises (AS and ASA) and other legal forms such as NUF and general partnerships (ANS and DA). Tables based on founding year show the legal form in the year of foundation, while tables based on year of survival show the legal form in the year of survival.

Region: The enterprise is located in the county and municipality based on the enterprise’s business address. In enterprises with activity in several locations, the region will be where the head office is located. Tables based on founding year show the location in the year of foundation, while tables based on year of survival show the location in the year of survival.

Standard Industrial Classification SN2007

Standard industrial classification: Current Standard industrial classification (SN2007) in Statistics Norway, which is based on the EU standard NACE Rev.2. This can be found at https://www.ssb.no/en/klass/klassifikasjoner/6.

In addition, a separate industrial classification has been created in the business demography, which is based on SN2007 and the previous industrial classification standard SN2002, as well as the enterprises' legal form. This industrial classification has 15 categories for known industrial activity, as well as 1 category for unknown industry. The standard is consistent throughout the transition from SN2002 to SN2007 in 2007/2008.

Tables based on the founding year show the industry in the year of foundation, while tables based on survival year show the industry in the year of survival.

Classification of enterprises by size group

Classification of organization form

Name: Enterprises
Topic: Establishments, enterprises and accounts

Division for R&D, technology and business dynamics statistics

National and county. Certain tables are also available by municipality.

The enterprise statistics are published annually with preliminary figures for year t, except for closures where there are preliminary figures for the years t-1 and t-2, and new enterprises where the figures for year t are final. Final figures are published for the year before the (first) preliminary year. New registered enterprises are also published quarterly, approximately five weeks after the end of the quarter.

Enterprises by employee characteristics were published annually until 2017, but have been postponed until further notice.

Statistics on annual active enterprises, new establishments and closures are reported to Eurostat. Activity is defined as enterprises with employees, turnover or investments and enterprises with wage earners, as the case may be.

Figures on new businesses per quarter are also reported.

The figures that are reported to Eurostat correspond mainly to those that are published at www.ssb.no/foretak.

Processed microdata are temporarily stored as data sets in the SAS programming language and long-term stored as text files. Computer programs are stored on a Windows disk in Statistics Norway’s folder structure.

The first national publication was in December 2004, with enterprises defined as new in the year they were registered, and survived if they had activity in terms of employed persons and/or turnover. The first mandated reporting to the EU was in 2014 (reference year 2012), with enterprises defined as new respectively in the year they became active for the first time and any years in which they were reactivated after at least two years of inactivity, and as surviving all subsequent continuous years of activity. Activity was defined in two different ways, as having employed persons and/or turnover and as having employees, and series were reported based on each of the two definitions.

In 2023, the national annual statistics were revised, and the main change was that the statistics now contain four versions of the figures, with new enterprises and activity being defined differently – in line with the previous national statistics and the two the definitions of the EU – as well as a version where the break in timeseries due to the introduction of A-ordningen in 2015 has been reduced to the lowest possible level. The versions comprise:

• Registered enterprises: Creations defined by enterprise registration and new activity in terms of employment, turnover or investment, survival defined by continued activity.
• Active enterprises: Creations defined by new activity in terms of employment, turnover or investment, survival defined by continued activity.
• Employer enterprises: Creations defined by new activity in terms of employees, survival defined by continued activity (employees).
• Active enterprises (adjusted employer data): Definitions equal to version “active enterprises”, but with adjusted employer data.

Other changes includes that tables showing the survival and growth of new enterprises, as well as tables showing high-growth enterprises and gazelles, are now displayed here together with the rest of the business demography. New tables with modified content are provided for the annual figures, and the calculations have been revised, among other things to follow the EU's definition of enterprise, which leads to a change in the figures. In addition, together with figures on high-growth enterprises, figures are now published for enterprises with stable development and enterprises with a large decline. The changes are implemented all through the time series.

Preliminary figures for 2022, figures on defunct enterprises and turnover, as well as more detailed tables with e.g. regional distribution will be published in November 2023.

The quarterly statistics have not been revised, and are only published in one version that corresponds to "registered enterprises".

Users of the statistics are public administration, the researchers, business and the media. It is also used as microdata in the statistics Entrepreneurs in business enterprise sector.

Principles for equal treatment of users in releasing statistics and analyses

No external users have access to the statistics and analyses before they are published and accessible simultaneously for all users on www.ssb.no at 8 am. Prior to this, a minimum of three months' advance notice is given in the Statistics Release Calendar.

The population and the calculation of survival are, usually, the same as for entrepreneurs in business enterprise sector - a personal statistic of entrepreneurial activity, for the covered legal forms. However, after the restructuring of the enterprise statistics in 2023, there will be deviations until the entrepreneurs statistics are restructured, as well.

The statistics are coordinated with the structural business statistics (SBS) in several ways. For the enterprises covered by SBS, the same figures are used for turnover, employment and employees. In addition, the same definition of market-oriented activity is used as in SBS, based on industry, legal form, sector code, registration in CCR, status code and type of enterprise. The two latter variables are internal to Statistics Norway's business and enterprise register (VoF), and show, among other things, whether the enterprise was active, and whether it is an auxiliary enterprise, holding enterprise, etc. Small discrepancies between the two statistics may occur, as a result of the enterprise statistics using other data extractions from VoF than SBS. For enterprises that are not covered by SBS, turnover calculated in VoF based on register data is used, alternatively operating income, or operating costs for industry 64, from business statements or annual accounts is used as figures for turnover. In connection with the transition from the old to the new EU regulation for business statistics, during a transitional period minor deviations may occur between the enterprise statistics and SBS.

Employed persons and employees deviate somewhat in terms of definitions and methods from Statistics Norway's labor market statistics, and there will therefore be differences in the figures. The business demography follows the definitions used in SBS, and which are given in the EU regulation on business statistics (EBS regulation).

The number of enterprises differs from the figures published by the Brønnøysund registers (BR). Deviations from the Register of Business Enterprises (RBE), which is part of the BR, are due in particular to the fact that the RBE does not include all legal entities that run businesses, as sole proprietorships without employees are not compulsorily registered in the RBE. On the other hand, transfers and changes of ownership are excluded from some of Statistics Norway's tables, including the survival and growth of newly established enterprises. The figure also differs from the number of legal entities in the CCR, since CCR includes legal entities without commercial activities. In addition, enterprises in Statistics Norway's statistics are defined as autonomous units, and auxiliary legal units are therefore merged with the autonomous legal unit to which it provides auxiliary services. In BR the registered unit corresponds to the legal unit.

Not relevant

Commission Regulation (EC) No 251/2009 of 11 March 2009.

Commission Regulation (EC) No 439/2014 of 29 April 2014.

Calculation method and definitions in "Eurostat - OECD Manual on Business Demography Statistics. Methodologies and working papers. ISBN 978-92-79-04726-8".

The statistics include market-oriented activity in all industries according to the Standard for industrial classification (SN2007), except the primary industries (division 01, 02 and 03) and public administration and defense (division 84).

Market-oriented activity is delimited on the basis of industrial code, legal form, sector code, registration in RBE, status code and enterprise type. Norwegian subsidiaries of foreign enterprises (NUF) and voluntary organizations (FLI) that are not registered in the RBE in Brønnøysund do not carry out commercial activities and are therefore excluded from the statistics. Other body corporates (ANNA) are also excluded, unless they operate pension funds (industry 65.300). The same applies to other legal forms that obviously do not run commercial activities. The statistics also do not include enterprises that carry out administration and service provision in the state and social security administration (sector code 6100) or in county administration and municipalities (sector code 6500), even if the industry to which they belong is covered by the statistics. For industries that provide a mixture of public and private services, in particular power and water supply (division 35 and 36), education (division 85) and health and social services (division 86, 87 and 88), the figures may therefore appear to be incomplete. The central bank of Norway (sector code 3100) is also excluded. Enterprises within the state's business operations, state-owned enterprises, state enterprises, municipal business operations, independent municipal enterprises and state loan institutions, i.e. enterprises with sector code 1110, 1120, 1510, 1520 or 3900, are on the other hand included in the statistics, if they operate within the industries that is covered by it. The same applies to enterprises in all other sectors.

Enterprises that have not yet gained activity in terms of turnover or employment are not included in the "active enterprises" version. Similarly, enterprises that have not yet acquired employees are not included in the version "enterprises with employees".

Enterprises that are inactive for at least 2 years counts as inactive throughout the 5 year period, even if they are reactivated, while upon the reactivation it counts as a new separate enterprise. This means that surviving, dormant and inactive enterprises from a given cohort of enterprise births (i.e. first-time births or reactivations) always add up to the original number of births.

The business demography is based on administrative information about the legal units and the underlying establishments (LKAUs), as well as figures for employment, employees, turnover and investments. For several of the financial industries, turnover is a poor measure of activity, and alternative measures are then used instead. The administrative information is mainly obtained from Statistics Norway's business register (VoF), and both current and historical information is used. However, the date of foundation for limited liability enterprises and the registration of foundations and NUFs with regard to commercial activities is obtained from the CCR register in Brønnøysund.

Employment in final figures is obtained from VoF, which obtains most figures from Statistics Norway's structural business statistics (SBS). The employment figures in SBS are based on the register-based labor market statistics, but are further quality assured. For industries and individual enterprises not covered by SBS, employment in VoF is based on the number of employees or various estimation techniques. The number of employees corresponds to the number of employed persons minus the number of owners who work in the enterprise. The number of owners is derived from the legal form and a lower turnover limit of NOK 150,000 per owner in 2021, adjusted for inflation with the consumer price index (CPI). If turnover exceeds the limit, the number of owners in ENK is set equal to 1, and in DA, ANS and PRE equal to 2. The main basis for data on employees from 2015 onwards in SBS is A-ordningen, which is a coordinated digital collection of data on employment, income and tax deductions to the Tax Administration, the Norwegian Labour and Welfare Organization and Statistics Norway. Before 2015 the main source was AA-registeret, in which contract workers and employees in very minor positions (below approximately 4 hours per week) were not included, as well as other differences.

Turnover is also mainly obtained from VoF, which for most industries and legal forms has its figures from SBS and administrative sources (VAT register, tax return data for self-employed persons and accounting data). For the remaining industries and legal forms, turnover is obtained from accounting and tax return data or imputed from of e.g. operating income.

Investment data are obtained from the SBS.

There is no separate data collection in the business demography, only the reuse of administrative data and data from statistical surveys. The data used for final figures and the data on employees for preliminary figures are quality-assured as part of the production of the basic statistics, while the turnover data for preliminary figures is quality-assured by Statistics Norway only for industry sections G – N. The quality assurance includes direct contact with legal units, as well as a linking to the administrative registers being part of the Monitoring System for establishments in VoF. The quality assurance of NACE codes in VoF takes place, among others, through administrative sources, through contact with the enterprises in the production of the basic statistics and the administration of VoF, and by means of notifications about change in enterprises’ purpose to CCR.

The source data undergoes an extensive set of processing, before the enterprises are finally aggregated into tables. The processing includes:

• Obtaining and organizing register data.

• Extraction and linking of variables.

• Aggregation and consolidation of legal units into enterprises.

• Delimitation of the population for survival and growth.

• Marking the year of establishing (i.e. birth and other creations) and years with activity.

• Linking of establishments for identification of ownership changes and creations other than births.

• Linking of information (organisation number, industry, etc.), in cases of ownership changes, from the original to the receiving enterprise.

• Labeling of different types of establishing, survival and non-survival.

• Calculation of age.

• Delimitation of the populations for high-growth enterprises, high-growth micro-enterprises, gazelles and micro-gazelles.

• Calculation and categorization of growth and high growth.

• Marking of confidential data.

• Aggregation into tables.

In the high-growth statistics, it is the annual growth rate in the last 3-year period that determines whether the enterprise is a high-growth enterprise or a gazelle (i.e. young high-growth enterprise). The categories of growth, based on this rate, include:

• Very high growth: ≥ 20 percent growth

• High growth: ≥ 10 percent to < 20 percent growth

• Stable: ≥ -10 percent to < 10 percent growth

• Large decline: ≥ -20 percent to < -10 percent growth

• Very large decline: < -20 percent growth

Categories have been created based on employed persons, employees and turnover respectively. Enterprises within the categories "very high growth" and "high growth" are considered high-growth enterprises.

Four different versions of the final figures in the business demography are calculated, partly to apply different definitions of establishment and survival in line with the reporting to Eurostat and previous national business demography, and partly to create tables without breaks as a result of the introduction of A-ordningen in 2015. In the high-growth statistics, only the version "active enterprises" is calculated, because the difference between the various versions here is assumed to be very small.

The estimations will be described in more detail in a separate documentation.

Not relevant

Sensitive information about individual enterprises is protected according to the Act relating to official statistics and Statistics Norway (Statistics Act). This means that numbers are replaced with dots, if it is possible for an actor - directly or indirectly - to read out sensitive information about an enterprise from the number in question. This may apply, for example, if the figure is a sum based on only 1 to 3 enterprises, or if one or two enterprises make up a very high proportion of the sum (90 or 95 per cent respectively). The latter is called the dominance rule. The requirement to replace numbers with dots also applies in cases where a protected number can be read from the difference between the total and underlying individual categories. In such cases, an additional number must be protected, so-called secondary protected, in order for the first protected number to receive real protection.

In the business demography, only turnover needs protection based on the dominance rule, while the other content variables are protected solely based on the number of enterprises. To get the maximum benefit from protected tables, in many cases it can be decided with high probability which numbers are primary protected and which are secondary protected. If two values included in a sum are replaced with dots, and the time series of one value consists of relatively high numbers of enterprises and the time series of the other consists of very low numbers of enterprises, it is most likely the number in the low time series that is primary protected, and accordingly consists of 1 to 3 companies, while the number behind the other dot can be read out of the sum with fairly high accuracy.

The statistics on the survival of new enterprises and the other parts of the business demography are closely linked. Enterprises being survivors in year t are included in the population of the business demography in t. Enterprises that became defunct in t are among the enterprises that were not survivors in year t+1. The two statistics are based on the same basic data, and they use the same figures for turnover, employed persons and employees.

Since it can take time to determine whether an enterprise is active or not, some pragmatic adjustments have been made. In the national statistics on the survival of enterprises, a distinction is made between surviving, dormant and non-surviving enterprises. Dormant enterprises are defined as survivors when they resume the activity. However, they cannot be identified as dormant until the next cohort with final data shows that the activity was resumed. In the last cohort with final figures and in the preliminary figures, dormant enterprises are therefore included in the category "not survived". In tables of defunct enterprises, the time series is one year shorter than in the other tables, and the last two years in these tables are considered provisional. In the population statistics, an enterprise is considered active if it had registered activity that year. It corresponds to surviving enterprises, and enterprises of all ages are covered. The preliminary figures are recalculated at the next annual publication, when the final data base is available.

The restructuring of the business demography has brought the national publication of business demography more in line with the reporting of business demography to Eurostat. Four versions of the figures are published, one of which (registered enterprises) corresponds to the old national business demography, with some adjustments, while two others (active enterprises and enterprises with employees) correspond to the reporting to Eurostat, with some small adaptations.

The business demography is now designed so that changes in method, data sources and definitions can be implemented through the time series, as far back as there are micro data available. However, some breaks cannot be avoided. In the description below, a distinction is made between breaks due to changes in method, data sources and/or definitions, and permanent or temporary real changes as a result of, for example, changed laws and regulations.

Transitional rule E to the tax act in 2005: In connection with the tax reform from 1 January 2006, a transitional rule was introduced in 2005, which led to the creation of a number of holding enterprises. This resulted in a large real increase in the number of new enterprises within the financial industry this one year.

Change in rules for limited liability enterprises (AS) in 2012: From 1 January 2012 it became easier to establish a limited liability enterprise, partly because the requirement for equity capital was reduced. This resulted in a real and lasting increase in the number of limited liability enterprises starting in 2012.

Turnover for industry 06 (extraction of crude oil and natural gas): Up until the reference year 2014, the oil and gas extraction enterprises submitted ordinary income statements (RF-1167). Since then a new income statement (RF-1323, Tax return for enterprises covered by the Petroleum Tax Act) is used for the extraction enterprises. This affects turnover in industry 06 which from 2015 is based on calculated figures, but not whether an enterprise has turnover or not.

“A-ordningen” in 2015: As of 2015, the number of employed persons is based on owners plus employees from A-ordningen. Until 2014, the number of employees was based on owners plus employees from the AA register. The main difference is that the new employee data include contractors (freelancers and others who work for a fee), which the old employee data do not. This resulted in a persistent increase in the number of new enterprises due to a changed data source for the number of employees. A more detailed description of the changes upon introduction of A-ordningen is given in a separate article (Norwegian only) under the register-based employment statistics.

Until the publication of final figures for 2020, legal unit was used as a pragmatic approach to enterprises in the statistics, while from final figures 2021 and preliminary figures 2022 the official definition of an enterprise is followed. The change has been backcasted to older years in the micro data, and therefore does not cause a break in the time series.

The statistics are based on other statistics and on administrative registers, and have the same sources of error as these. In the survival statistics and high-growth statistics, errors will affect the figures several years later, because the statistics for a given year are based on the situation several years back (e.g. the survival of 5-year-old enterprises) and follow the enterprises over time (e.g. growth in the last 3 year period).

Typical register errors are missing or outdated information as a result of delays in registrations. Such delays are due to reports arriving late and changes being registered some time after they have occurred. It is primarily the backlog in the administrative information in VoF that are relevant for the business demography. Missing or outdated administrative information can lead to enterprises being misclassified or ending up in an "unknown" category, and in the "registered enterprises" version, late registration of enterprises can lead to the establishing year being pushed forward, and the enterprise thus receiving a too low age.

The quarterly figures for new enterprises are most affected by lag, since they are published the shortest time after the end of the reference period.

The remaining parts of the business demography is affected similar way, but here the percentage effect is far smaller. In the statistics on the survival and growth of new enterprises, the category "not survived" will include dormant enterprises in the last final reference year and in the preliminary year.

It is assumed to be some missing reporting of turnover among the very smallest enterprises, where the activity is marginal. There are relatively many of these enterprises, while at the same time they have small influence on the basic statistics of business demography (e.g. SBS), and thus the quality assurance is also less extensive here. However, in the business demography, where the number of enterprises is an important variable, marginal enterprises make up a significant proportion. If reporting is missing each year, both births, population and number of survivors are underestimated. If it is missing only for periods, on the other hand, births and deaths are overestimated due to reactivations, while the population and number of survivors are underestimated. Overall, it is assumed that missing reporting of turnover among the very smallest enterprises leads to a certain underestimation of the population and survival enterprises, while the effect on births is more uncertain.

When new enterprises are registered, they must report whether this is a newly started activity or a change of ownership. Where such information is missing, duplicate checks are carried out against existing enterprises to identify any changes in ownership. Ownership changes are also identified by following their establishments, i.e. corporate affiliations, from year to year. Enterprises where information about change of ownership is not reported and where the establishments also get a new number run the risk of not being caught in these routines. It is unknown how many changes of ownership this applies to, but it cannot be ruled out that the number of changes of ownership is somewhat too low and that the number of births is somewhat too high. It is nevertheless assumed that this source of error is small.

Number of employed persons, employed persons, etc. are calculated as an average over all 12 months of the year, even in the start-up year. These figures are thus lower for the birth year seen as a whole than the level of activity indicates. This is not strictly speaking a source of error, but it affects volume figures and distribution by size groups for the year of birth, as well as growth in the next year.

Not relevant