Accuracy and reliability

Overall accuracy

The overall accuracy of the statistics is considered high as the delimitation of the population is effected on the basis of Statistics Denmark's Central Business Register comprising the industry codes stated. Furthermore, the statistics are compiled in accordance with legislation and the reminder procedure is intensive, which results, as expected, in a response rate of over 95 pct.

For the business units with more than 100 employees which have not reported data, an imputation is conducted. The imputation rate is usually less than 2.5 per cent.

The number of job vacancies tends to decline between 0.5-1.5 per cent while the job vacancy rates often are stable and never have declined more than 0.1 percentage points, before the data obtain status of final the same quarter a year after the first publication.

Sampling error

As with all other sample-based statistics, there are some sample errors associated with the estimates. The information on the statistical uncertainty is used for assessing to which extent a change, e.g. in the number of job vacancies, reflects some degree of uncertainty or the increases or falls are real. The margins of sample errors are influenced by the sample size, and the statistical uncertainty is reduced by 50 per cent when the sample size is increased fourfold. As is the case in other EU Member States, the variation coefficient, which is the standard deviation in relation to the estimate, is used in calculating the sample errors.

For the total number of occupied posts the variation coefficient were 0.7 percent, while for the total number of job vacancies the variation coefficient were 2.7 percent. The coefficient of variation for a large part of the NACE sections and size classes are relatively high. The reason why the latter variation coefficient is relatively high is not due to the raising of figures or the design and size of the sample, but due to the great variations between the number of job vacancies reported and the many data reported concerning zero vacancies in each stratum.

The variation coefficient CV for job vacancies for the 4th quarter of 2023:

• Manufacturing, mining and quarrying and utility services (B+C+D+E): 4.4 percent.
• Construction (F): 10.4 percent.
• Trade and transport etc. (G+H+I): 4.7 percent.
• Information and communication (J): 10.2 percent.
• Financial, insurance and real estate (K+L): 8.5 percent.
• Other business services (M+N): 6.4 percent.
• Total (B-N): 2.7 percent.

Non-sampling error

In the Job Vacancy Statistics the sampling frame is the Central Business Register, which covers all registered businesses in Denmark. The register is currently updated on a daily basis and is expected to have full coverage.

Every quarter a new sample is extracted from the Central Business Register and at that moment there are very few differences between the register and the sample. As times goes the register will be updated (some business units are no longer statistically active and new business unit have entered the register). Before grossing-up we update the sample with information on business units no longer statistically active. Unweighted is the over coverage below 1 per cent.

Errors caused by (partial) non-response, incorrect submitted data and misunderstandings are sought to be minimised by repeated reminders (response-chasing) and a thorough error-checking of the submitted data.

Statistics Denmark assumes that the number of job vacancies reported by the business units, in some degree, is a variable with non-negligible measurement error. The main source of non-negligible measurement error in connection with job vacancies reports is uncertainty among the respondents regarding how to measure job vacancies and which units they have to report data for (business unit or the entire business). Furthermore, we have some business units called ‘fictitious workplaces’, i.e. business units in which employees without a physical workplace are grouped, e.g. salesmen and sailors. Reporting data for these fictitious workplaces can sometimes be difficult. Statistics Denmark is aware of the problem with these measurement errors. Most of the measurement errors are captured in micro validation where corrections to the reported data are made.

It is mandatory to submit data electronically, either by the public reporting platform Virk or by the telephone entry system, and it is not possible to submit a form with blank fields. Partial non-response is therefore zero.

The response rate is high due to the fact that the businesses are legally obliged to report to the statistics. Un-weighted the unit non-response rate is below 5 per cent.

Even though the statistics are published quarterly, data are collected monthly, and consequently the quarterly sample is equally distributed over the three months of the quarter. The variation among the workplaces in each individual stratum is equally distributed, and each individual workplace is only selected once every quarter. The aim is to fill job vacancies throughout the quarter and avoid variation due to e.g. holiday periods.

The Job Vacancy Statistics only imputes unit non-response for business units with more than 100 employees, and the imputation is done just before the enumeration procedure.

The method is ratio imputation from the program Banff. As a first choice, the imputation is done in relation to business units within the same business and as a second choice to business units within the same stratum. The number of job vacancies (and occupied posts) is estimated by using the generalised regression estimator with the number of registered employees (through the Central Business Register) in each business unit as auxiliary information.

In the enumeration process the model-assisted estimation from the program CLAN is used.

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

As with all other sample-based statistics, there are some sample errors associated with the estimates. The information on the statistical uncertainty is used for assessing to which extent a change in, e.g. the number of job vacancies reflects some degree of uncertainty or the increases or falls are real. The margins of sample errors are influenced by the sample size, and the statistical uncertainty is reduced by 50 per cent when the sample size is increased fourfold.

As is the case in other EU Member States, the variation coefficient, which is the standard deviation in relation to the estimate, is used in calculating the sample errors.

Every quarter a sample is selected of approximately 7,000 workplaces with at least one employee at the time of selection. All workplaces with more than 100 employees are selected in the sample and the remaining part of the sample is based on random selection. The sample is stratified in accordance with 13 industries and 6 size groups. In connection with the selection of the sample, the workplaces are classified into the 6 size groups on the basis of the number of employees who are registered at the workplace according to Statistics Denmark's Central Business Register.

When the statistics are published the workplaces are re-classified, implying that they are classified to the size groups which correspond with the data reported on the number of employees. The reason for this re-classification is that register-based information is frequently a couple of years old. On the basis of new employment figures from Statistics Denmark's Central Business Register, the sample is optimised once every year with regard to movements, i.e. small workplaces which are now employing more than 100 persons are compulsorily selected, and large workplaces which are now employing less than 100 persons are released.

1/12 of the workplaces employing less than 100 persons are replaced every quarter. In this way, it is ensured that the sample size is maintained at the same level and that the sample gives an up-to-date picture of the number of occupied posts and job vacancies. This implies that a workplace employing less than 100 persons participates in 12 consecutive quarters and is subsequently exempted from participating over a long period of time. Workplaces employing more than 100 persons are compulsorily selected and consequently always participate in the sample. This is in compliance with Statistics Denmark's data supplier policy.

The businesses are legally obliged to report data for the statistics and failure to report data will result in a police notification, cf. the Act on Statistics Denmark. The response rate is normally over 95 per cent.

The data are continuously subjected to data editing and correction. In order to ensure that data are reported for the individual workplace, the number of employees reported is validated with the register-based information concerning the number of employees from Statistics Denmark's Central Business Register, and if there are major differences in the data reported, compared to the data stored in the Central Business Register, we contact the workplace for the purpose of verifying the data reported.

The data on job vacancies are also subjected to editing in relation to the number of employees reported. If there are any major differences in the data reported on the number of job vacancies, compared to the number of employees that is expected for a workplace of the size stated, we contact the workplace. In connection with subjecting the data to editing and correction, we contact about 10 pct of the workplaces. It must be assumed that not all errors are detected, and consequently some sample errors remain.

For the compulsorily selected workplaces which have not reported data or have been exempted from reporting, imputation is conducted, in cases where it is possible, on the basis of the data reported from the same CVR no. over the same period. We make efforts to ensure that all compulsorily selected workplaces have reported data, so that the imputation is only conducted for those workplaces which for one reason or another have been exempted from reporting. The program CLAN is used in connection with the raising of figures.

For the total number of occupied posts the variation coefficient is under 1 per cent, while for the total number of job vacancies the variation coefficient is 3-5 per cent. The coefficient of variation for a large part of the NACE sections and size classes are relatively high. The reason why the latter variation coefficient is relatively high is not due to the raising of figures or the design and size of the sample, but due to the great variations between the number of job vacancies reported and the many data reported concerning zero vacancies in each stratum.

Every time we publish a new quarter we also publish revised data from the previous four quarters. The main reason for this is that we receive data from business units which do not report data punctually. As a consequence, on-going error correction of reported data takes place.

The number of job vacancies tends to decline between 0.5-1.5 per cent while the job vacancy rates often are stable and never have declined more than 0.1 percentage points before the data obtain the status of final.

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

Every time we publish a new quarter we also publish revised data from the previous four quarters. The main reason for this is that we receive data from business units which do not report data punctually. As a consequence, on-going error correction of reported data takes place.

The number of job vacancies tends to decline between 0.5-1.5 per cent while the job vacancy rates often are stable and never have declined more than 0.1 percentage points before the data obtain the status of final.