Statistical processing

Source data

The Danish Working time accounts (WTA) are based on a combination of census and survey data. The WTA are compiled on the basis of four primary data sources:

1. Labour Market Accounts see Documentations of statistics LMA
2. Structure of Earnings see Documentations of statistics SES
3. A-Income Statistics see Documentations of statistics AINCOME
4. Employment Statistics for Employees see Documentations of statistics BfL.

(1) With the Labour Market Accounts (LMA) monthly statements are available on employment, jobs, temporary absences and paid hours of work and for employees also compensation of employees throughout the calendar year for all the years covered by LMA data.

LMA form the basis of WTA on paid hours of work for self-employed and assisting spouses. In LMA these are calculated on the basis of hours paid for employees, but enumerated with how much more self-employed and assisting spouses are working according to labour force survey (LFS). Furthermore, WTA uses the latest developments in LMA to project jobs, employment and paid hours of work for self-employed and assisting spouses.

With LMA longitudinal data, it has become significantly easier to establish, whether leave has its origin in employment or unemployment. WTA uses information on leave from LMA covering all months of the year. Furthermore, continuing recent trends from LMA, WTA projects information on leave from employment (sickness and maternity) to months where no structural data exist.

Another huge quality improvement is that LMA can produce preliminary structural data for the reference year 2016 to be available already in August 2017.

(2) Structural Earning Statistics (SES) are used to convert paid hours of work from LMA to actual hours worked during the year in WTA.

Furthermore, data from the SES are used as help information to describe the distribution of hours worked over the months of the year in the WTA. Earning statistics are used for identifying jobs for workers paid by the hour, who are characterized by not being paid during absence. Therefore, the distribution of paid hours of work by hourly workers can represent the distribution of actual hours worked over the months of the year.

Furthermore, studies based on the labour force survey (LFS) shows that self-employed and assisting spouses do not have a significantly different distribution of hours worked over the year than employees. This information is in the WTA used for calculating the relative distribution of hours worked compared to hours paid for over the months of the year for all employed.

So although from eIncome (LMA and employment statistics for employees) only information on paid hours of work in the month are available, the WTA can thereby calculated how much this represents in hours worked per. month, based on the knowledge of how actual hours of work are distributed relative to paid hours of work over the months. Paid hours of work generally have a different distribution over the months of the year than actual hours worked due to the fact that absence is not evenly spread over the months of the year.

(3) Income statistics data (AINCOME) based on reports from the Danish Central Pension System (CPS) are used for adjusting compensation of employees in the WTA to include earnings of funded labour market pension.

(4) The Employment Statistics of Employees (BFL) contains monthly data on jobs, hours paid and compensation of employees throughout the year for employees. The information is used in the WTA to project compensation of employees, hours paid for, employment, primary and sideline (secondary, third etc.) jobs for employees during periods when there is no AMR data. Given that LMA include preliminary structural data, then the projection period is reduced so that the maximum length of projection is 15 months. The 15-month projection occurs in the calculation of the first quarter in June, while for example the calculation of second quarter figures in September will only be projected for six months. This increases the quality of the WTA statistics considerably.

In deciding which data sources to apply in compiling the WTA, attention is centred on the major advantages provided by each individual statistics. For example, LMA are used to ensure complete coverage in the calculation of employment, number of jobs, aggregate payroll costs and paid hours of work. This includes personal interviews used for obtaining information on groups that are not covered by the administrative registers. Information from the wage and salary system of the enterprises is used to convert paid hours of work into hours worked during the year.

The Working Time Accounts are exclusively based on existing data sources, which are subsequently converted to the concepts used in the WTA. The WTA is flexible in its choice of primary sources, which can be replaced by other sources, if these have proved to be more accurate. The choice of primary source decides the amount of data editing necessary. When it comes to integrating all the sources, however, all the concepts are consistent in conforming to international standards and every variable fulfils the requirement of the system for the WTA.

Data in WTA are summarized (aggregated) prior to integration and projected so that the output data alone are broken down by socioeconomic status (whether you are an employee, self-employed or assisting spouse), industries, sectors, gender and amount of work.

Frequency of data collection

The Working Time Accounts (WTA) make use of already existing statistics in Statistics Denmark when new information is available. Typically, the ATR obtains information from the Structure of Earnings Statistics (LON) once a year (end of October). From the Labour Market Accounts (AMR) once a year (end of October) and the Income Statistics (AINDK) once a year (end of October) Employment of employees (BFL) once a month. BFL is used for the forecasting of the structural year

Data collection

The information is retrieved from these existing sources in Statistics Denmark when new data are available.

Data validation

Data are already checked for errors in the primary statistics. In the Danish Working Time Accounts (WTA) further checks, troubleshooting and debugging are carried out. This is partly based on the information from the producers of the input sources, partly systematic (mostly figurative) controls the internal consistency between variables and over time, and by comparisons with other published statistics. Finally developments are systematically being discussed with stakeholders from other statistics in Statistics Denmark.

Data compilation

The starting point for the calculation of the average employment and the average number of jobs is the status information on the number of persons employed and the number of jobs each day of the month according to the Labour Market Accounts (LMA). When calculating the average number of persons employed or the average number of jobs in the quarter or year, this is done as an average of the 3 months in the quarter (12 months in the year), hours worked and payroll in the quarter (year) are calculated as the sum of hours worked and payroll in the 3 months in the quarter (12 months in the year). The Working Time Accounts (WTA) are statistics based on several input sources. Revisions are continuously made due to new data input sources, data breaks in existing input sources, new industry formats, new sector codes, new or changed user needs (national and international), etc.

The series are seasonally adjusted. In line with international guidelines, employment and jobs are seasonally adjusted, but these series are not trade-day adjusted. Hours worked and payroll for employees are both seasonally and working day adjusted.

In delivery to the EU statistics are trade day adjusted before the index is calculated. In Council Regulation (EC) No 2019/2152 of 27 November 2019 concerning European Business Statistics, only indexed data are provided.

The labour market and hence the Working Time Account (WTA) have been strongly affected by the shutdown of society from mid-March 2020 as a result of COVID-19 and by the measures put in place to mitigate the effects of the shutdown. The exceptional circumstances from the latter part of the first quarter of 2020 until the start of 2022 mean that the statement of hours worked is subject to greater uncertainty than usual. Similarly, the new holiday law which came into force on 1 September 2020 has caused a change in the pattern of paid hours, particularly for newly recruited officials who, unlike in the past, are entitled to paid holiday from the start of their employment.

Adjustment

The data are seasonally adjusted, but otherwise no corrections of the data are carried out beyond what has already been described during data validation and data processing.

In the national statistics, the following series are seasonally adjusted: - employment (not working day adjusted, seasonally adjusted) - jobs (not trade day adjusted, seasonally adjusted) - hours worked (working day adjusted and seasonally adjusted) - total wages and salaries of employees (working day adjusted and seasonally adjusted)

All variable types (number of jobs, employment, hours worked and payroll) are seasonally adjusted (i.e. corrected for 'fixed calendar effects'). Of the 'moving calendar effects', all variables are corrected for any Easter effects. Furthermore, hours worked and payroll are corrected for any trading day effects (including any leap year effects).

The calculations of the latest seasonally adjusted figures in particular are subject to increased uncertainty. In particular, the COVID-19 crisis has triggered large and unusual fluctuations in the non-seasonally adjusted hours worked figures, and the result of the seasonal adjustment should therefore be used with caution. When the seasonally adjusted values are subject to greater uncertainties than usual, this could also lead to greater ex-post revisions than usual.

Seasonal adjustment will itself lead to revisions in previously published seasonally adjusted figures - even if there are no changes in the historical non-seasonally adjusted figures. For the national publication, the seasonal adjustment is made on monthly data broken down into 19 industries (standard db19 grouping), each of these further broken down into two sectors (public administration and services vs enterprises and organisations). The series are further divided into 3 socio-economic groups (employees, self-employed and assisting spouses). Where seasonally adjusted values are not available, actual values are used instead. When seasonally adjusted data are added to the database, the main figures are calculated by indirect seasonal adjustment as the sum of the seasonally adjusted breakdowns for the sub-groups (Socio-economic status 3 * sector2 * db19).

For the purposes of Eurostat's STS Regulation, seasonal adjustment is not applied, but only trade-day adjustment, and is provided as index calculations. The following data are provided 1. employment (index calculation only) 2. hours worked (working day adjusted and indexed) 3. hours worked (indexed only) 4. compensation of employees (working day adjusted and indexed) 5. compensation of employees (indexed only)