Comparability

Comparability - geographical

The statistics are worked out according to international guidelines, European System of Accounts (ESA 2010) and International Labour Organisation (ILO 1988: Current International Recommendations on Labour Statistics), the latest of which is reviewed in 2013 (Resolution concerning statistics of work, employment and labour underutilization, 19th International Conference of Labour Statisticians).

The population are persons affiliated to Danish registered companies, which is consistent with ESA2010 boundaries. ESA2010 includes working in resident companies (see ESA 2010 paragraphs 2.04 to 2.11) (colloquially called 'the day-time population'). ILO guidelines include the resident population (colloquially called 'the night-time population').

Comparability over time

From the fourth quarter of 2019 and onwards, the payroll is incorporated so that accrued holiday pay is recorded at the time of earning and not, as previously, at the time of payment.

Otherwise, there is no data breach in the current ATR time series. But compared to historical versions / releases of WTA, there have been the following data breach:

As from the publication in September 2016 of the WTA for the second quarter of 2016 and the annual WTA for 2008-2015 the working time accounts system surpasses to use longitudinal data from the new Labour Market Accounts (LMA). WTA thus builds on structural data covering all months of the year. Further, the period of projection using short term statistics has been significantly reduced, since the working time accounts in September 2016 make use of a preliminary version of LMA for the year 2015 and the structure on earning statistics for 2015 has been included as well. In February 2017, the preliminary data from LMA for 2015 is replaced with final data for 2015. There is therefore a significant improvement in quality of the WTA. The revision of the WTA on transition to the LMA has given rise to minor revisions in the levels of employment, jobs, hours worked and compensation of employees:

Size of the WTA revision upon transition to LMA (pct.):

In December 2012 the Working Time Accounts (WTA) were adjusted, implying that new data sources (primarily based on eIncome) are used for the compilation. As changes have been made to the population, concepts, sources as well as methods, this has resulted in revised levels and revised developments throughout the year. See Break in WTA on transition to eIncome.

With eIncome the target population was expanded to include people who work in Danish companies but live abroad:

Furthermore, there were breaks in the WTA as a result of revisions to the classifications of sectors and industries, see [Break in WTA due to changed classifications since 2008]https://www.dst.dk/ext/arbe/WTAbreak).

Earlier versions of the Working Time Accounts (WTA) have been used to write-back the current series to 1995. However, these back-written series are not used for separate disclosure, as the quality is not good enough at the detailed level published in the WTA.

But the series are used to write-back the National accounts employment data series, see for example:

• Quarterly national accounts, industries: NKBB10: Employment (10a3-grouping) by socioeconomic status, industry and seasonal adjustment.
• Annual national accounts, industries: NABB69: Employment (69-grouping) by socioeconomic status and industry.

In addition, the back-written series are used in deliveries to the Eurostat Short Term business Statistics (STS), but on the condition that they may not be published separately but solely together with series from other countries.

Coherence - cross domain

A fundamental principle of the Working Time Accounts (WTA) is to document the coherence between statistics utilized in the WTA and to document coherence between the primary statistical data and the WTA.

The WTA are worked out according to international guidelines.

However, since the WTA is primarily a register-based statistic, it does not include information on unpaid overtime and illegal (including undeclared) hours of work. In the Labor Force Surveys (LFS) and in the National Accounts, the hourly concept includes unpaid overtime and illegal (including black) work, as explained by respondents in LFS.

Transitional tables between the WTA and the RAS: Register-based Labour Force Statistics (employment) and the ERE: Establishment-related Employment Statistics (jobs and compensation of employees) is published in Statistical Reports (Statistiske Efterretninger in Danish only) for the annual WTA. A description of the transitional tables between the WTA and the National Accounts (employment, hours and compensation of employees) can be found in the publications on the National Accounts.

The basis of the number of jobs in the Working Time Account is end of November statements of the number of jobs in the Establishment-related Employment Statistics (ERE statistics). Unlike in the ERE statistics WTA number of jobs also include jobs in business statistics below the activity limit used in business statistics and also include jobs for people in eIncome not resident in Denmark.

Employment includes persons who are temporarily absent due to leave, but who have an affiliation to a workplace in the form of a job to return to. The transformation from job to employment include deduction of the persons' sideline (secondary, third etc.) jobs at the end of November and addition of the number of people who are either on sick leave, maternity leave or childcare leave from employment. Employment in The Register of Employment Statistics end of November (representing average employment per day in November in the WTA) includes, in addition to RAS employment also employment for persons not resident in Denmark. The Working Time Accounts employment for self-employed and assisting spouses is defined in the same manner as in the register-based labour force statistics (RAS), where self-employed consists of the following three groups: employers, VAT payers and other self-employed.

To get the average number of employees in Q4 an average of employment in October, November and December is used. The average employment during the year is calculated as the average number of employees in the 12 months of the year, or an average of employment in the four quarters of the calendar year. Similarly, the number of jobs in the year is calculated as the average number of jobs in each of the 12 months, where the number of jobs end of November in the Register of Employment Statistics represents the average number of jobs in the month of November in the WTA.

The concept of earnings in the WTA is the same as that used in the ERE statistics, but the population is slightly different because the ERE statistics operates with an activity limit (in accordance with international guidelines for business statistics), which is not used in labour market statistics (or economic statistics).

The revision in December 2012 implied a change in the division of labour between national accounts and work accounts so that the working time account adjusts what can be adjusted on job level, while national accounts make further adjustments on a more aggregate level. This change of labour division leads to greater differences between the working time accounts and national accounts figures. On the other hand, it implies that adjustments are made to the extent possible where the greatest expertise is. Finally, the changes in labour division also imply fewer revisions of the Working Time Accounts, which is hereby no longer dependent on the final data in the national accounts.

The new WTA also includes jobs for employees who do not live in Denmark, if they have jobs in companies in Denmark or on Danish ships. The WTA does not include employees of foreign companies working in Denmark the rules on hiring out of labour.

The compilation of Working Time Accounts is based on the idea that the figures are comparable over time to the highest possible degree. The sources will continuously be improved and replaced by other sources if these have proved to be more accurate. New sources will always be adapted to the concepts of the Working Time Accounts System. This implies that adjustments of existing sources cannot immediately be seen as changes of variables and concepts in the Working Time Accounts Statistics, although adjustments of the level of the specific variable may be made according to the new and improved information.

There may be several reasons for the lack of data comparability between the different statistics: - Differences in compilation methods - Differences in the population - Differences in the definitions - Differences in the margins of statistical errors - Differences in the time of publication.

Coherence - internal

The big improvement in relation to internal consistency between variables in the WTA is due to a system that primarily build on a single source, namely eIncome. Previously, data from many different statistics were used. See note on data breach in ATR on transition to eIndkomst.

In the transition to LMA the internal consistency increased further, mainly because there are now significantly better information of leave of absence from sickness and maternity as well as jobs and employment for self-employed and assisting spouses during the year, and that all this information is processed and overlap treated in the same system, namely the LMA. Furthermore, the internal consistency increased because the period projected using short term statistics has been reduced significantly.