Accuracy and reliability

Overall accuracy

The uncertainty in the development of the number of employees is estimated to be less than 1 per cent of the total number of full-time employees, where 1 per cent corresponds to approx. 20,000 full-time employees. As regards more detailed statistics in terms of industry and geographical distribution the uncertainty is much greater. Quality studies have so far only been made to a very limited extent.

Not only is there has been a significant quality improvement in employment statistics related to the transition to eIncome. Over time, the eIncome register have also been improved. Thus, a quality measure for the calculation of full-time employment is the proportion of hours paid in eIncome that have been imputed because they have either not been reported or because they have proved to be invalid.

In general, there is a tendency that more and more report information on hours paid to eIncome, which guarantee a better quality over time. Furthermore, the guidelines from the tax authorities on the reporting of hours paid to eIncome have becomed more clear, and the precision and knowledge of concepts is increased in the reporting over time so that, for example, the reporters become aware that unpaid absences is not be included in the reported hours paid.

Sampling error

Not relevant for these statistics.

Non-sampling error

The uncertainty in the development of the number of employees is estimated to be less than 1 per cent of the total number of full-time employees, where 1 per cent corresponds to approx. 20,000 full-time employees. As regards more detailed statistics in terms of industry and geographical distribution the uncertainty is much greater. Quality studies have so far only been carried out to a very limited extent.

Some reports to eIncome lack information on hours paid or the reported information has been found to be invalid. Therefore imputed (estimated) paid hours of work for these reports.

Quality management

Statistics Denmark follows the recommendations on organisation and management of quality given in the Code of Practice for European Statistics (CoP) and the implementation guidelines given in the Quality Assurance Framework of the European Statistical System (QAF). A Working Group on Quality and a central quality assurance function have been established to continuously carry through control of products and processes.

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.

Quality assessment

Data validation focuses on the areas that affect the statistics most. This means that there is a particular focus on the reporting of hours of paid work where defective reported wage hours are rectified. Also erroneous reporting of employees' attachment to the establishment (workplace) is rectified.

In the eIncome register information on hours paid in (employee) jobs is reported in each reference month. The quality of these data reflect, of course, how good the reports are. Generally, it is our feeling that it is high quality information. However, especially relating to information on unpaid hours of absence and hours of overtime there may be quality problems in the eIncome data. Lack of devaluation of the monthly paid hours of work as a result of unpaid absence leads to an overestimation of paid hours. Lack of registration of paid overtime would result in an underestimation of hours paid. Typically, the registration problem is worse about unpaid absences. There can be, especially for salaried who do not have paid absence, be a problem in that the hours are not always written down sufficiently during periods of absence. When reporters become aware that the hours must be changed when there is unpaid leave, and they suddenly begin to report these correctly, this affects the projection of paid hours of work over the year and the development from year to year. Unfortunately, this increased attention to the reported paid hours of work is not happening at the same time for everyone, why it can be difficult to detect these error alerts in the production system and adjust for this back in time. In connection with the publication of employment statistics for employees third quarter of 2012 in December 2012 data were adjusted for such error reports to eIncome for monthly paid employees in municipalities and regions back in time (for the period January 2008 to August / September 2011). This is done by using monthly reports on absences that employees are not paid for to the Earnings Statistics . The hourly reports to the Earnings Statistics include more detailed split than hour reports to eIncome.

On the main level assessed statements of full-time employees are judged to give a true picture of the level and development, but because of defects in the material, which has been reported, there is some uncertainty regarding the distribution on industry, sector and geography. Correct values of these variables depends on whether the employer report correct production numbers. For municipalities and regions, incomplete reporting of production unit numbers result in uncertainty regarding the branch location, while defects in the private sector especially raises uncertainty on the geographical distribution.

A constant full-time standard for all employees, namely 37 hours per week (160.33 hours per month). This is equivalent to full-time norm for most tenured functionaries. However, there may be groups who have another full-time norm, for example, hourly paid or newly hired salaried employees who have a full-time norm of less than 160.33 hours per month, as these groups typically will not receive wage when on temporary absence such as holiday. This has implications for the levels and the interpretation of the number of full-time employees. However, it is not possible based on eIncome material to divide the population according to various groups of employees with different full-time standards. One advantage of having the same full-time norm for all employees is that it is simple to convert the number of full-time employees to hours paid for, providing an indicator of trends in employment volume for employees. This enables users to easily develop alternative splits (with varying full-time standards) based on their needs and the groups they want to compare.

In the monthly statistics of employees and in the quarterly statistics of employees figures on the concept "number of persons with an employee job". In determining this concept, Statistics Denmark has taken into account that hourly workers and newly hired salaried employees do not have paid vacation. This is carried out by including employees who in a period of up to 45 days have not received wages but subsequently returned to the same employer also in the period in which they have not received salaries.

Under the old flex job scheme the employer in most cases pays full salary, though the number of working hours normally is significantly lower. The employer receives in return a government grant. Under the new scheme the employer pays, however, only for the agreed working hours while the person on flex job receive social assistance from the municipality. In order to improve comparability over time in the statistics, individuals included under both the old and the new flex job scheme, occure in the statistics with the number of hours corresponding to the agreed working hours. For the individuals on the old flex job scheme, where there is no information about the person's working hours, an average working hour is used.

Starting with the release of Q1 2016, more quality improvements are incorporated in the statistics, which are carried back to the statistics began in 2008. The majority of the quality improvements are already incorporated in the Labour Market Accounts, and will be incorporated in the new Working Time Accounts. For a number of groups the adjusted number of payed working hours are compared to previous publications of the statistic. Other quality improvements are also incorporated, including the geographic distribution. As a result of the changed number of hours of paid working hours, the level and evolution in the number of employees are changed compared to previous publications, while the total number of employee jobs are unaffected by the changes. A detailed description of the quality improvements can be read here https://www.dst.dk/ext/arbe/kvalitetsforbedringer.

Data revision - policy

Statistics Denmark revises published figures in accordance with the Revision Policy for Statistics Denmark. The common procedures and principles of the Revision Policy are for some statistics supplemented by a specific revision practice.

Data revision practice

Revision Policy: The preliminary analysis of the statistics are published approx. 45 days after the end of the quarter. The revised statement issued approx. 75 days after the reference quarter and the final statement 3 months later together with the revised statistics for the following quarter. No further revisions are made, as later changes have limited implications for the description of the short-term development. The final estimates of total employment is determined in the structural statistics Register Based Labour Force Statistics (RAS).