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.
Statistics Denmark, Social Statistics, Labour and Income
1.3. Contact name
Confidential because of GDPR
1.4. Contact person function
Confidential because of GDPR
1.5. Contact mail address
1.6. Contact email address
Confidential because of GDPR
1.7. Contact phone number
Confidential because of GDPR
1.8. Contact fax number
Confidential because of GDPR
2.1. Data description
Statistics Denmark collects Earnings data from the public sector as well as from all private enterprises with at least 10 fulltime employees/full time equivalents (FTE). These Earnings data are combined with other labour cost data, collected through an annual survey, in order to calculate the total amount of labour costs. The scale of coverage is at a level that is regarded as satisfying when it comes to user needs.
2.2. Classification system
Not available. New concept added with the migration to SIMS 2.0. Information (content) will be available after the next collection.
2.3. Coverage - sector
The statistics cover corporations and organizations, as well as the public sector.
2.4. Statistical concepts and definitions
Not available. New concept added with the migration to SIMS 2.0. Information (content) will be available after the next collection.
2.5. Statistical unit
The statistical unit is the individual job, which is defined as a person employed with a specific employer and engaged in a specific occupation.
2.6. Statistical population
The statistical population is all persons employed in companies or organisations with ten or more employees, as well as all persons employed in the public sector. Labour costs for the private sector are based on the annual structure of earnings statistics and the survey of other labour costs of the private sector, which is also annual. The survey on other labour costs for enterprises in the private sector is based on a special sample of enterprises with 10 or more employees, and the sample is drawn with the use of the Danish Business Register. The sample is stratified to cover enterprises in different size groups (number of employees) and class of industry.
The framework used in selecting the target population is Statistics Denmark’s Central Business Register. The Central Business Register contains information on all enterprises and local units in Denmark (covering both the private and public sector). Each enterprise is identified by means of an 8-digit registration number (CVR number), which is the same number used in the administrative registers operated by the tax authorities, etc. The local units are identified by a 10-digit local unit code number. The Central Business Register is continuously updated, which implies that there are only minor problems in achieving close to complete coverage. Information on firms and economic activity of the local units, geographical location and legal ownership is collected from the Central Business Register. Furthermore, the register contains information on total employment measured in full time units.
2.7. Reference area
The reference area is Denmark.
2.8. Coverage - Time
The time coverage of the statistics is the year 2020.
2.9. Base period
Not available. New concept added with the migration to SIMS 2.0. Information (content) will be available after the next collection.
For details please see sections 3.1 to 3.6.
3.1. Source data
Total labour costs are compiled on the basis of the labour cost survey, which comprises two sub-surveys, the annual structure of earnings survey and the survey of other labour costs for the private sector. The primary data for the total labour costs are collected in collaboration with the Danish Employers' Confederation and the Danish Employers' Association of the Financial Sector, which collect information from their affiliate business enterprises and the information, is then made available to Statistics Denmark. Statistics Denmark collects information from non-affiliate business enterprises and organisations. When possible the information is collected from ATP.
3.2. Frequency of data collection
Data is collected annually.
3.3. Data collection
The data collection system is designed to explore the potentials of the latest computer technology for efficient and rational collection and reporting of data. In this respect, a Standard for Electronic Statistical Reporting has been developed, which is used by computer agencies and respondents utilising standard systems and by business enterprises which have set up a payroll system. In addition to this, Statistics Denmark has prepared an electronic questionnaire as an alternative method of reporting data on paper questionnaires.
3.4. Data validation
Data is validated in a number of ways throughout the different production processes of both the sub-samples it is made of. The final validation, which directly relates to the total labour costs, is done at an aggregate level for all the NACE industries and size classes. Comparisons are made with previous years reults to ensure conformity and quality of the results. Eurostat also performs validation checks as data is received from the transmitting countries.
3.5. Data compilation
Total labour costs is based on data transmissions from business enterprises covering 58 percent of the target population. Due to the stratification of the sample, there are large differences between the degrees of coverage within the groups of stratification. Consequently, the number of jobs in the statistics has been raised. This is conducted by classifying the business enterprises by industry and number of employees. Subsequently, the degree of coverage is ascertained for each industry group, which forms the basis for calculating the raising factor allocated to each enterprise in the group in question.
3.6. Adjustment
Not applicable.
4.1. Quality assurance
Statistics Denmark follows the principles in the Code of Practice for European Statistics (CoP) and uses the Quality Assurance Framework of the European Statistical System (QAF) for the implementation of the principles. This involves continuous decentralized and central control of products and processes based on documentation following international standards. The central quality assurance function reports to the Working Group on Quality. Reports include suggestions for improvement that are assessed, decided and subsequently implemented.
4.2. Quality management - assessment
The quality of the statistics is considered to be high. Some of the fundamental information in other costs rest upon voluntary agreements, and as a consequence are therefore somewhat difficult to assess the validity of. Still, the impact these information have on the total labour costs are minimal, and the uncertainty caused by these is considered to be limited. A potential source of error is the lack of response from the business enterprises. However, the response rate for other labour costs is close to 100 per cent.
For details please see sections 5.1 to 5.3.
5.1. Relevance - User Needs
The main users of the Danish Labour Cost Statistics are researchers, private business enterprises, ministries, counties, municipalities, national organisations (employer- as well as employee organisations) and international organisations. Most of the user needs are related to earnings and not the total labour costs. Statistics Denmark is not familiar with any users not being satisfied with the content and output of the LCS data.
5.2. Relevance - User Satisfaction
There has not been performed any evaluation/survey of user satisfaction regarding the earnings or total labour costs.
5.3. Completeness
Not available.
5.3.1. Data completeness - rate
Not available.
There are no estimation of accuracy available at the moment. But in general the statistics on labour costs are considered as both very accurate and reliable.
6.1. Accuracy - overall
Despite being a census the accuracy is affected by different types and degrees of errors such as coverage-, measurement- and non-response errors.
The margins of statistical errors are especially linked to hours of work. Data reported on paid absence can be subject to inaccuracies. In addition to this, there may be errors in the periodic delimitation, which are essential to the compilation of hours worked as well as the agreed working time. However, efforts are continuously made to improve the data quality through feedback to enterprises and through update and improvement of the production systems.
In addition, there is some inaccuracy linked to the variables on the number of employees in the labour cost survey. This is primarily because the underlying surveys are intended to capture trends in earnings and labour costs and not the size or the structure of the labour force. The variables on the number of employees is calculated based on how many records we receive from the enterprises, which means that improvements in the pay-roll systems of the enterprises can have implications on the figures. It is reckommended that other sources are used for reliable information on the actual size of the labour force in Denmark.
6.2. Sampling error
The sample survey is a source of uncertainty. The sample survey used for other labour costs is created on the basis of extracts from the Central Business Register of business enterprises with more than nine employees working full-time. The sample is stratified in accordance with the size of the enterprises and economic activity. In 2020 the sample consisted of 3448 enterprises. 1847 of these received a questionnaire from Statistics Denmark, which corresponds to about 54 pct of the sample. 42 pct. of the questionnaires were mailed by the Danish Employers' Confederation, while the remaining 4 pct. were mailed by the Danish Employers' Association of the Financial Sector. About 100 pct. of the mailed questionnaires were incorporated in the production of statistics. The questionnaires that were not used in compiling the statistics were omitted for different reasons, e.g. the business enterprises discontinued their activities during 2020, or the number of employees decreased to less than ten during the year.
There exists no overall assessment of the uncertainty caused by sampling errors.
6.2.1. Sampling error - indicators
By means of the annual survey of structural earnings, Statistics Denmark is able to provide information on all variables, except for the variables D2, D3, D4, D5 and a very small part of D12. This information is obtainable from the survey of other labour costs. Other labour costs are obtained via a sample survey, which naturally implies that the survey may be subject to sampling errors. Other labour costs make up, on average, only about 4 pct. of the total labour costs.
6.3. Non-sampling error
The size or form of non-sampling error has not been estimated or assessed yet.
6.3.1. Coverage error
When analysing the range of principal economic activities used in the LCS 2020 the result shows that about 19,000 private enterprises with at least 10 full time employees is in the target population. The amount of employees in the study population private enterprises covers 97 per cent of the amount in the target population private enterprises. A measurement of the public sector coverage has not been made.
Due to errors in the data reported by the enterprises, some of the reported data have not been used, and enterprises which discontinued or commenced their activities during the year are normally not included. Conversely, some enterprises not legally obliged have reported data for the statistics.
The national survey on other labour costs only covers NACE B-N. The information on other costs for NACE O-S (the central and local government) are estimated or calculated from other administrative sources. For the central government, this is done through a match with data from Statistics Denmark’s Office for Government Finance.
It is more problematic for the local government; here the same sort of data does not exist. The local governments report total levels, which cannot be split up into wages and salaries and other costs alone. Therefore, Statistics Denmark has to estimate most of the other costs for the local governments. Information regarding compensations received by employers due to sickness or maternity based on register-information exists and can be linked to data on wages and salaries with unique ID numbers identifying employee and employer. The rest of the components on statutory social security cost are estimated with more or less precision as a consequence.
6.3.1.1. Over-coverage - rate
There has not been performed any estimation of the over-coverage rate on the surveys behind the LCS.
6.3.1.2. Common units - proportion
Not aplicable.
6.3.2. Measurement error
The data reported to Statistics Denmark, which form the basis for the LCS, have been collected annually since 1994, and since then Statistics Denmark has systematically contacted business enterprises in cases where they have reported inadequate or erroneous data to Statistics Denmark. This continuous contact with business enterprises has improved data reports.
In addition to the continuous contact with enterprises, there is also a close cooperation between Statistics Denmark and the largest pay systems providers, which have led to constant improvements of the data reporting systems.
As the primary data for the Danish LCS are reported at the level of employees, there is a possibility that the data contains some errors. The employees with errors are deleted from data, and this means that the data reported by the enterprises is only usable for a part of their employees. Furthermore some enterprises does not (more or less intentionally) submit their earnings data.
Table 1 (see in Annex below) shows the fraction of the target population used to produce the statistics. The fraction is calculated as the actual number of full-time employees included in the survey compared to the amount registered by the central business register. The cells not completed in the scheme below mean that the specific section and size class does not exist in the data.
As mentioned earlier, only a very small number of enterprises do not deliver the requested data, because otherwise they will be imposed with a fine. As a result, the response rate for enterprises transmitting data for the annual structure of earnings survey and the labour cost survey is in both cases considered to be at least 99 percent.
There is a possibility of errors in the received observations (see above). Many of these errors concern missing values and can be fixed by imputation based on information from the rest of the observations.
Because of the disposal of some of the reported data, there are relatively big differences in the degree of coverage in the various industries. Consequently, the occupations included in the statistics have been weighted. The enterprises are classified by industry and size on the basis of data from the Central Business Register. The degree of coverage is established for each industrial size group and using this as basis a weight is calculated, which is subsequently linked to each occupation in the group in question. Because of this and the fact that the response rate of enterprises is relatively high, the impact of non-response errors is considered to be limited.
6.3.3.1. Unit non-response - rate
Not available.
6.3.3.2. Item non-response - rate
Not available.
6.3.4. Processing error
All enterprises respond to the contact made by Statistics Denmark to collect earnings data, otherwise they will be imposed a fine. This implies that nearly all units to which Statistics Denmark has made contact, for the purpose of making use of the enterprises’ data, submit information to Statistics Denmark. The reason why there is, after all, a divergence between the target population and the number of enterprises covered by the statistics, is attributed to the processing of data. The quality of some of the data that are submitted to Statistics Denmark is questionable, and they are sorted out during the course of the processing of data, and consequently do not appear in the statistics.
6.3.4.1. Imputation - rate
Imputation has not been performed directly on any of the variables included in the LCS, but it has been done on some sub-variables that form the basis for a few of the variables in the LCS. The sub-variables in question are number of hours not worked due to sickness and holiday payments (for both full-time and part-time employees). In the case of number of hours not worked due to sickness for full-time employees, the approximate rate of imputation is 60 percent. For the rest of the variables the imputation rate is about 30 percent. None of these imputations should affect the accuracy of the variables ‘Annual labour costs’ or ‘Gross earnings in the reference month’. Instead there is a possibility that the imputation of the number of hours not worked to some extent have an impact on the accuracy of the number of hours worked.
6.3.5. Model assumption error
The information on other costs, the survey of the variables D2, D3, D4, D5 and a minor part of D12 are based on a sample of enterprises in the private sector. Due to the sample size, it is impossible to stratify other costs at the same level of detail as with the remaining variables. This implies that Statistics Denmark has been forced to estimate the above variables to some of the strata.
Table 2 gives an overview of which strata there has been collected information on other costs. The table contains 3 values. ’OK’ means that in the sample survey of other costs, there are one or more enterprises in the respective stratum that belongs to the sample. ’.’ implies that there is not one enterprise in the entire population with local units in the respective stratum. ’Missing’ implies that in the population there are enterprises with local units in the respective stratum, but none which are represented in the sample survey of other costs.
On the basis of a general linear model, the cells with missing-values have been estimated. This estimation has been conducted using both the respective NACE group and the respective size group as explanatory variable.
For NACE O-S the other costs have been estimated using registers already collected by Statistics Denmark and there is full coverage of NACE O-S and the size classes. It is however important to know that the other costs for NACE O-S only cover costs that can be found in administrative sources and legal costs that can be calculated.
There are normally no delays in the national publishing of the statistics or in the delivering of data on labour costs to Eurostat.
7.1.1. Time lag - first result
Not aplicable.
7.1.2. Time lag - final result
Not aplicable.
7.2. Punctuality
Earnings data
In December 2019 enterprises were informed of their legal obligation to report earnings data to Statistics Denmark for the reference year 2020. Deadline for submitting the data reports to Statistics Denmark was January 2021.
Normally, in cases where an enterprise failed to report data even after a fourth reminder in June 2021 the enterprise would have been reported to the police, who subsequently would deal with the respective enterprises. This proces was halted for reference year 2020 due to covid-19.
The production process of the national earnings statistics began during spring 2021. The earnings statistics was published March 2022 due to postponement as a result of lockdown and covid-19.
Other costs
In March 2020 enterprises were informed of their legal obligation to report other costs data to Statistics Denmark for the reference year 2020. Deadline for submitting the data reports to Statistics Denmark was April 2021.
In cases where an enterprise fails to report data even after a fourth reminder, the enterprise is passed on to the police, which will subsequently deal with the respective enterprise.
Due to covid-19 the production process began for the national statistics in late 2021 and the results were released Feburary 2022.
Immediately after this, the work on establishing/revising the LCS production systems was initiated, including, e.g. quality checks, plausibility checks, etc. Data were submitted to Eurostat in June 2022. There has been no major redelivery of data since then.
7.2.1. Punctuality - delivery and publication
Not available.
As mentioned in the section on accuracy, the quality of the variables concerning number of employees can be questioned. This is because the information is based on how many records are received from the pay-roll systems of the enterprises, and as the quality of the enterprises transmission of data stedily improves, this can have an influence on the amout of records received. Furthermore as the labour cost survey is based on two surveys that are mainly dealing with earnings or labour costs, the focus is on variables relating to these concepts. The information received on the number of employees is therefore more a by-product of the two surveys, and not based on an actual labour force survey. It is reckommended that other sources are used for reliable information on the actual size of the labour market in Denmark.
8.1. Comparability - geographical
The regulations ask for information on the number of local units in the universe (E1) and the number of units in the sample (E2). The Danish figures show the exactly same number in the variables E1 and E2. This is not a picture of how things are in our register, but a picture of how things should be, and consequently a manipulation of the variable E2.
Following several talks with Eurostat concerning the LCS 2000, it was decided that this was the most appropriate thing to do. The problem is as follows:
Statistics Denmark receives an identification number (CVR number) from all enterprises and an identification number for each local unit associated with the enterprise. These two numbers are used in identifying the enterprise in the Statistical Business Register, where NACE, number of full-time employees, address, etc. are linked. In a few instances, it is impossible to identify the local unit in the Statistical Business Register, and the persons linked to the non-identifiable local unit(s) are subsequently allocated with the enterprise but without a local unit.
If Statistics Denmark were to report the data that are actually recorded in our register on earnings, E2 would be lower than E1. The problem is that this would not reflect the reality, but only reflect some internal technical problems involved in matching two registers.
8.1.1. Asymmetry for mirror flow statistics - coefficient
Not aplicable.
8.2. Comparability - over time
In general it is possible to compare the LCS 2012 with LCS 2008. However a few small changes have been made regarding the placement over certain variables from the Danish Other Labour Cost survey.
Reimbursements for expenses regarding paid absence, maternity leave etc. has been moved from D5 (Subsidies received by the employer) to D1221 (guaranteed remuneration in the event of sickness).
Contributions/provisions for company pension schemes (benefit plans) has been moved from D3 (Other expenditures paid by the employer) to D1222 (Employers’ imputed social contributions for pensions and health care).
Since 1995, changes in the methods or variable definitions to be reported from the respondents have not been made. In contrast, there have been major improvements in the coverage and quality. As mentioned earlier, Statistics Denmark collects data every year for the national SES and LCS. This implies that the respondents are familiar with the way in which to report data, both with regard to regularity of the survey, but also due to a relatively efficient way of contacting the enterprises in cases of insufficient data reports.
8.2.1. Length of comparable time series
Not available.
8.3. Coherence - cross domain
Coherence with the hours actually worked in the main job in the labour force survey (LFS) 2020.
As Table 3 shows there are rather big differences between the LCS and the Labour Force Survey (LFS) when it comes to the number of hours worked per job when broken down by NACE sections C-S. There are several reasons why different figures occur for the two surveys.
First of all, it is of great importance to take note that the LFS is a survey of households, while the Danish LCS is a survey of enterprises. There may be great difference in the way in which employees and employers, respectively, consider various conditions, with respect to both the individual person and the enterprise. For example, to which NACE section does the enterprise belong, how long the employee actually works and whether the individual employee is full-time or part-time employed, etc.
The number of actually worked hours is probably where the greatest difference is. The LFS is built on information gathered from the employees whereas the LCS is a survey from the employers’ side as it is collected from the enterprise. This means that information on e.g. the number of hours worked can be different as the employee and not the employer takes a record of hours worked but not paid. Although this might not explain all the difference, it can still have a large impact on the results.
Furthermore, the LFS contains information on the main job whereas the LCS contains information on all jobs. E.g. one person has a full-time job of 1,500 hours actually worked and a part-time job of 800 hours actually worked. This person will in the LFS give the result 1,500 as only the main job is used. In the LCS the outcome is quite different. For simplicity let us assume that all the grossing-up factor is 1, then the person with the two jobs will have 1,150 hours on average per job in the table below. This is quite a big difference from the 1,500 in the LFS.
Coherence with wages and salaries from the Structure of Business Statistics (SBS)
Table 4 shows that there are also rather big differences between wage and salary per employee according to the LCS and the SBS. The explanation for these differences is presented below.
The SBS and the LCS statistics are compiled on the basis of two different populations. The SBS statistics are based on enterprises, whereas the LCS statistics are compiled on the basis of the local unit. In the case of large groups of companies, local units are frequently classified to various NACE groups. As the sum of wages and salaries is derived from the number of persons employed in each local unit, the size of the sum is, of course, affected by this. For example, if an enterprise belonging to NACE I have several of the local units belong to NACE C, all employees will appear under NACE I in the SBS while in the LCS they will appear in both NACE C and I.
As a consequence, the comparison made in Table 4 is not necessarily of the same group of employees. This is also reflected in the statistics, where e.g. the number of employees belonging to NACE C is 309,435 in the SBS and 255,488 in the LCS. The reason why the figures in Table 4 differ is therefore, at least to some extent, a consequence of the difference in the way the data is collected.
Coherence with the Labour cost index (LCI)
As there are rather big differences between the two statistics when it comes to objectives and methodologies used, it is neither expected nor attempted at making them coherent. As an example, the labour cost index is calculated on identical companies, which exist in both periods, while this is not the case with the LCS. The LCI in Denmark is produced in a way that partially removes fluctuations in earnings caused by structural changes on the labour market. The LCS is on the other hand a structural statistics where these changes are a natural part of the figures. This issue was also presented at the Labour costs workshop on 5-6 May 2015. The document for this presentation is attached as an annex in the metadata.
Coherence with the hours actually worked in the main job in the labour force survey (LFS) 2020.
The Danish NA do not apply the LCS results as a data source in their compilations. National Accounts make some adjustments regarding information on wage and employment to harmonize the National Accounts with the accounting and production statistics. National Accounts do not use the LCS data. For these reasons the comparability of the NA and LCS can be very low in certain groups. Please note that the total figure for NACE B-S is, despite the low rate of comparison, rather close for the two statistics.
As the LCS is identical to the combination of two annual national statistics – the annual structural earnings and the survey of other labour costs – the LCS results delivered to Eurostat are not published. Instead, the two annual national statistics are published.
9.1. Dissemination format - News release
Not available.
9.2. Dissemination format - Publications
As Statistics Denmark collects information on labour costs on an annual basis, there is at least one publication of the statistics each year. Note here that unlike the data transmitted to Eurostat, the annual domestic statistics on labour costs only cover the private sector.
Each year the statistics is available at a detailed level at Statistics Denmarks Statbank page normally around October. At the same time as the statistics become available at the Statbank page, a news release with graphs, tables and text is released on the labour costs. It is also included in the Statistical Yearbook published by Statistics Denmark each year.
Key figures are also published in Nyt fra Danmarks Statistik (News from Statistics Denmark) and in Statistisk Årbog (Statistical Yearbook).
An English declaration of contents can be found at DST website. If a greater level of detail or tabular cross-tabulations is required, they can be produced on request.The register is at the level of individual employees and may be used in connection with compiling more detailed statistics or in coupling data from other statistics.
Other costs
When it comes to the survey on other labour costs, key figures are available on the subject page on Labour Costs. In addition, the figures are published in Nyt fra Danmarks Statistik (News from Statistics Denmark) and in Statistisk Årbog (Statistical Yearbook). More detailed figures are also available in English from Statbank Denmark, tables SAO01-SAO04.
If a greater level of detail or tabular cross-tabulations is required, they can be produced on request. The English declaration of contents can be found at DST website.
9.3.1. Data tables - consultations
Not available.
9.4. Dissemination format - microdata access
Not aplicable.
9.5. Dissemination format - other
For the time being, no information is sent to the reporting units when the statistics are published.
9.6. Documentation on methodology
There exists a national quality declaration for the statistics on total labour costs with comprehensive description of metadata and methodological issues. The quality declaration can be read in english at this website.
9.7. Quality management - documentation
Not available.
9.7.1. Metadata completeness - rate
Not available.
9.7.2. Metadata - consultations
Not available.
Not available.
11.1. Confidentiality - policy
Not available.
11.2. Confidentiality - data treatment
Not available.
There is nothing else to add to the quality declaration on the Danish structural statistics on labour costs.
Statistics Denmark collects Earnings data from the public sector as well as from all private enterprises with at least 10 fulltime employees/full time equivalents (FTE). These Earnings data are combined with other labour cost data, collected through an annual survey, in order to calculate the total amount of labour costs. The scale of coverage is at a level that is regarded as satisfying when it comes to user needs.
Not Applicable
Not available. New concept added with the migration to SIMS 2.0. Information (content) will be available after the next collection.
The statistical unit is the individual job, which is defined as a person employed with a specific employer and engaged in a specific occupation.
The statistical population is all persons employed in companies or organisations with ten or more employees, as well as all persons employed in the public sector. Labour costs for the private sector are based on the annual structure of earnings statistics and the survey of other labour costs of the private sector, which is also annual. The survey on other labour costs for enterprises in the private sector is based on a special sample of enterprises with 10 or more employees, and the sample is drawn with the use of the Danish Business Register. The sample is stratified to cover enterprises in different size groups (number of employees) and class of industry.
The framework used in selecting the target population is Statistics Denmark’s Central Business Register. The Central Business Register contains information on all enterprises and local units in Denmark (covering both the private and public sector). Each enterprise is identified by means of an 8-digit registration number (CVR number), which is the same number used in the administrative registers operated by the tax authorities, etc. The local units are identified by a 10-digit local unit code number. The Central Business Register is continuously updated, which implies that there are only minor problems in achieving close to complete coverage. Information on firms and economic activity of the local units, geographical location and legal ownership is collected from the Central Business Register. Furthermore, the register contains information on total employment measured in full time units.
The reference area is Denmark.
Not Applicable
Despite being a census the accuracy is affected by different types and degrees of errors such as coverage-, measurement- and non-response errors.
The margins of statistical errors are especially linked to hours of work. Data reported on paid absence can be subject to inaccuracies. In addition to this, there may be errors in the periodic delimitation, which are essential to the compilation of hours worked as well as the agreed working time. However, efforts are continuously made to improve the data quality through feedback to enterprises and through update and improvement of the production systems.
In addition, there is some inaccuracy linked to the variables on the number of employees in the labour cost survey. This is primarily because the underlying surveys are intended to capture trends in earnings and labour costs and not the size or the structure of the labour force. The variables on the number of employees is calculated based on how many records we receive from the enterprises, which means that improvements in the pay-roll systems of the enterprises can have implications on the figures. It is reckommended that other sources are used for reliable information on the actual size of the labour force in Denmark.
Not Applicable
Total labour costs is based on data transmissions from business enterprises covering 58 percent of the target population. Due to the stratification of the sample, there are large differences between the degrees of coverage within the groups of stratification. Consequently, the number of jobs in the statistics has been raised. This is conducted by classifying the business enterprises by industry and number of employees. Subsequently, the degree of coverage is ascertained for each industry group, which forms the basis for calculating the raising factor allocated to each enterprise in the group in question.
Total labour costs are compiled on the basis of the labour cost survey, which comprises two sub-surveys, the annual structure of earnings survey and the survey of other labour costs for the private sector. The primary data for the total labour costs are collected in collaboration with the Danish Employers' Confederation and the Danish Employers' Association of the Financial Sector, which collect information from their affiliate business enterprises and the information, is then made available to Statistics Denmark. Statistics Denmark collects information from non-affiliate business enterprises and organisations. When possible the information is collected from ATP.
Not Applicable
There are normally no delays in the national publishing of the statistics or in the delivering of data on labour costs to Eurostat.
The regulations ask for information on the number of local units in the universe (E1) and the number of units in the sample (E2). The Danish figures show the exactly same number in the variables E1 and E2. This is not a picture of how things are in our register, but a picture of how things should be, and consequently a manipulation of the variable E2.
Following several talks with Eurostat concerning the LCS 2000, it was decided that this was the most appropriate thing to do. The problem is as follows:
Statistics Denmark receives an identification number (CVR number) from all enterprises and an identification number for each local unit associated with the enterprise. These two numbers are used in identifying the enterprise in the Statistical Business Register, where NACE, number of full-time employees, address, etc. are linked. In a few instances, it is impossible to identify the local unit in the Statistical Business Register, and the persons linked to the non-identifiable local unit(s) are subsequently allocated with the enterprise but without a local unit.
If Statistics Denmark were to report the data that are actually recorded in our register on earnings, E2 would be lower than E1. The problem is that this would not reflect the reality, but only reflect some internal technical problems involved in matching two registers.
In general it is possible to compare the LCS 2012 with LCS 2008. However a few small changes have been made regarding the placement over certain variables from the Danish Other Labour Cost survey.
Reimbursements for expenses regarding paid absence, maternity leave etc. has been moved from D5 (Subsidies received by the employer) to D1221 (guaranteed remuneration in the event of sickness).
Contributions/provisions for company pension schemes (benefit plans) has been moved from D3 (Other expenditures paid by the employer) to D1222 (Employers’ imputed social contributions for pensions and health care).
Since 1995, changes in the methods or variable definitions to be reported from the respondents have not been made. In contrast, there have been major improvements in the coverage and quality. As mentioned earlier, Statistics Denmark collects data every year for the national SES and LCS. This implies that the respondents are familiar with the way in which to report data, both with regard to regularity of the survey, but also due to a relatively efficient way of contacting the enterprises in cases of insufficient data reports.
