Accuracy and reliability

Overall accuracy

The register-based labour force statistic is the data foundation for the commuting statistic. I terms of sources of uncertainty it is especially the following that is relevant to the commuting statistic:

• Determination of the primary attachment to the labour market for a person with more than one job at the end of November: In these cases the primary job is determined by the number of hours worked. The job with the most hours is defined as the primary. For self-employed and assisting spouses the information about working hours is always imputed, and in some cases that goes for employees as well. For employees the imputed part is though decreased from 14 pct. in 2008 to barely 4 pct. in 2013. It is the workplace address for the primary job and the address of residence that is the foundation of the calculation of commuting distance. The imputation can in some cases mean that the wrong job is pointed out as the primary, and therefore also the wrong address that the calculation is done from.

• Employed distributed by industry: There are some uncertainties concerning which workplace the specific employee job is situated, because a number of reports from the employers are insufficient. The insufficient reports causes a necessary search for and correction of errors regarding the workplaces, but even though the errors are corrected in the best possible way there will continuously be uncertainties about parts of the information about workplaces - especially on a detailed level. The uncertainties are especially linked to the detailed distributions of employment on industries in the public sector, and particularly on the municipal area. In that area the distributions on industries within "residential care activities" (industry 87) and "social work activities without accommodation" (industry 88) are uncertain on a more detailed level, but also between the two industry. Detailed distributions on "Education" (industry 85) are also combined with uncertainties.

Before 2008 the data sources for RAS was more uncertain. Especially the data source for information about employee jobs: the central register for information notes from the tax office (in Danish: det centrale oplysningsseddelregister). The employer should on the note indicate whether the employee has been employed at the end of November. This information was exclusively used for statistical purposes, and Statistics Denmark did not have the chance to control whether the reports where correct or not. The central register did not have information about wage hours. As a replacement an extent of working hours where calculated on the background of the yearly labour market supply (ATP) and the length of the employment, though the length of the employment was considerably uncertain. The calculated extent of working hours was used at the classification - that means when the primary connection to the labour market, for a person with more than one job, is specified. From 2008 the information about working hours are significantly more precise, which has also improved the precision of the classification significantly.

Sampling error

Not relevant for this statistic.

Non-sampling error

Inadequate or conflicting reports are corrected. Among others the information about working hours is imputed for persons with no information reported. The share of imputed working hours for employees is decreased from 14 pct. to barely 4 pct. in 2013 and therefore represents a still smaller uncertainty.

If a person has more than one job at the time of reference, the primary job is determined as the job with the highest assumed number of working hours.

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

RAS - the data foundation of the commuting statistic - is a total count of the people resident in Denmark and the data foundation is the Labour Market Account (LMA). In LMA various data sources are integrated, corrected and harmonized in one joined system. Since RAS is a total count of the population there is not the same uncertainties as with statistics based on surveys. Read more about the quality assessment of RAS in the documentation.

Regarding the commuting statistic following elements in RAS are especially relevant:

• Determination of the primary attachment to the labour market for a person with more than one job at the end of November: In these cases the primary job is determined by the number of hours worked. The job with the most hours is defined as the primary. For self-employed and assisting spouses the information about working hours is always imputed, and in some cases that goes for employees as well. For employees the imputed part is though decreased from 14 pct. in 2008 to barely 4 pct. in 2013. The imputation can in some cases mean that the wrong job is pointed out as the primary. The commuting distance is based on the workplace address of a persons primary job.

• Number of self-employed and assisting spouses: The period of jobs for self-employed and assisting spouses is more uncertain than for employees. A series of information is used to validate whether the person actually is active in a job as self-employed/assisting spouse. There will however be a more significant uncertainty about the number of self-employed and assisting spouses than with employees.

• Employed distributed by industry: There are some uncertainties concerning which workplace the specific employee job is situated, because a number of reports from the employers are insufficient. The insufficient reports causes a necessary search for and correction of errors regarding the workplaces, but even though the errors are corrected in the best possible way there will continuously be uncertainties about parts of the information about workplaces - especially on a detailed level. The uncertainties are especially linked to the detailed distributions of employment on industries in the public sector, and particularly on the municipal area. In that area the distributions on industries within "residential care activities" (industry 87) and "social work activities without accommodation" (industry 88) are uncertain on a more detailed level, but also between the two industry. Detailed distributions on "Education" (industry 85) are also combined with uncertainties.

• Socioeconomic status: The division of employees into level of skills (top managers, employees at upper levels, employees at medium levels, employees at basic levels, other employees and employees not further specified) is done by means of information about work function, which is reported to the wage statistic or otherwise imputed. That happens when the employees work in smaller firms in the private sector (less than 10 full-time employed), which are not obliged to rapport to Statistic Denmark’s wage statistic. In these cases the quality of information about the skill level is of lower quality.

When it comes to the calculated commuting distance it is also substantially to be aware that it reflects an ideal situation where every employed person is assumed to be travelling from residence to workplace at the shortest possible route and by car. The distance is therefore a measure of how far a person must travel to get to work, and not an expression for the real traffic pattern. Accessibility, preferences and other habits are not taken into account. Neither is public transportation which will often follow another route than the shortest. And finally is also happens that the workplace address refers to a head office while the real workplace is placed differently.

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

When the commuting statistic is published data is considered final. However data can be revised as a result of changes in methods or new information available. Since the data foundation for RAS changes to the labour market account (LMA) following changes have been made:

1. In connection to the publication of data for November 2013 data from 2008-2012.
2. At the publication in April 2016 the data is again revised back to 2008, because more updated information about the population is available at the time. Beside these changes smaller corrections of industries, work places and sector codes for employees are also made.

No further revisions are planned.