0 Administrative Information about the Statistical Product


0.1 Name
Register-based labour force statistics (RAS statistics)

0.2 Subject Area
Labour market

0.3 Responsible Authority, Office, Person, etc.
Labour market
Katja Stage, tel. + 45 39 17 34 77, e-mail: ksg@dst.dk
Thomas Thorsen, tel. + 45 39 17 30 48, e-mail: tst@dst.dk
Pernille Stender, tel. + 45 39 17 34 04, e-mail: psd@dst.dk

0.4 Purpose and History
The purpose of the RAS statistics is to provide a description of the Danish population's attachment to the labour market at a given moment en time (at the end of November). RAS statistics follows the international guidelines set by the International Labour Organisation (ILO). The guidelines are directed at Surveys (Labour Force Surveys) where information about attachment to the labour market is given by respondents. In RAS the guidelines from ILO have been adjusted to the fit the possibilities available when using material based on registers.

The first version of the RAS statistics includes the population resident in Denmark as of the 1 January 1981 and its attachment to the labour market at the end of November 1980. The statistics have been compiled once every year over the period 1981 to 2006.

0.5 Users and Application
RAS statistics are widely used by municipalities, counties, government departments, the news media and private enterprises. RAS is primarily used to analyse labour market structures because it contains very detailed information.

Information from RAS is also widely used in other statistical registers and databases maintained by Statistics Denmark. RAS is used in Working Time Account (WTA) and WTA is used in National accounts and balance of payments.

0.6 Sources
Registers operated by:
The Central Customs and Tax Administration, the Salary Information Register (COR)
Central Pension Register (CPS).
Data reports to Statistics Denmark from local government agencies regarding their salary payments
Extracts from the Central Business Register, the statistics on labour market policy measures (AMFORA),
The Register for Education of the Population (BUE),
The Central Register of Labour Market Statistics (CRAM),
The Population Statistics Register, the Employment Classification Module (AKM),
The statistical register on benefits during sickness or in connection with childbirths and the statistical register on benefits under the Act on Active Social Policies.

0.7 Legal Authority to Collect Data
This is not relevant for this statistics.

0.8 Response burden
The RAS statistics are exclusviley compiled on the basis of administrative and statistical registers, and there is therefore no response burden.

0.9 EU Regulation
No EU regulation.


1 Contents


1.1 Description of Contents
The register-based labour force statistics (RAS) are annual status observations of the population's attachment to the labour market at a given point-in-time at the end of November.

The RAS statistics the population is divided into 3 main groups according to the guidelines laid down by the ILO. The 3 groups are persons in employment, unemployed and persons outside the labour force. Employed persons are either employees, self-employed or assisting spouses. Persons outside the labour force are divided into 4 groups. Persons temporarily outside the labour force, retired from the labour force, Pensioners and others outside the labour force.

In some cases a person can be part several different groups at the time of reference (they are a part of more than one gross stock). To determine the most important attachment to the labour market an order of priority is used. The order of priority is based on:

- The principles for the population's attachment to the labour market as given from ILO (International Labour Organisation)- The quality of the sources- The extent of the persons participation in the different activitiesIn 2003 there is a break in the timeseries of the register based labour force statistics (RAS). Changes on the labour market and better sources resulted in a need for alterations in the statistics. The alterations resulted in e new priority:

Before 2003:
1. Unemployed
2. Early retirement pensioners.
3. Employers
4. Employees with jobs at full-time equivalence
5. VAT-registered persons
6. Self-employed insured against unemployment.
7. Employees with jobs at part-time equivalence, or duration of employment not stated.
8. Other self-employed
9. Assisting spouses.
10. Recipients of early retirement pay
11. Participants in labour market policy schemes.
12. Recipients of benefits during sickness or childbirth
13. Recipients of rehabiliation or cash benefits
14. Children and young people
15. Recipients of old age pension or early retirement pay
16. Recipients of civil servants' earned pensions
17. Persons in education
18. Others outside the labour force.

From 2003:
1. Unemployed
2. Employers
3. Employees with jobs at full-time equivalence
4. VAT-registered persons with activity above the triviality limit
5. Employees with jobs at part-time equivalence, or duration of employment not stated.
6. VAT-registered persons with activity belove the triviality limit
7. Early retirement pensioners.
8. Self-employed insured against unemployment.
9. Other self-employed
10. Assisting spouses.
11. Participants in labour market policy schemes.
12. Recipients of rehabilitation payment.
13. Recipients of unemployment payment
14. Recipients of benefits during sickness or childbirth
15. Recipients of cash benefits or introduction payment
16. Children and young people
17. Recipients of old age pension or early retirement pay
18. Recipients of civil servants' earned pensions
19. Persons in education
20. Other recipients of labour market policies payment
21. Others outside the labour force.

From 2004 the order of priority between Recipients of civil servants' earned pensions and Persons in education are switched.

RAS also contains addresses of workplace and home addresses. The information is used to create statistics of commuting.

1.2 Statistical Concepts
The population in the RAS statistics are persons resident in Denmark on January 1. The populations attachment to the labour market at the end of November the year before is measured.

The RAS statistics contain a number of variables. The most important variable is socio-economic status, which is defined for each person in the population. The population is divided into following groups:

- Self-employed
- Assisting spouses
- Top managers
- Employees upper level
- Employees medium level
- Employees basic level
- Other employees
- Employees not specified- Unemployed
- Recipients of pension benefits payable between early retirement and normal retirement pension.
- Participants in labour market policy without pay
- Participants in educational activities
- Participants in specially adapted projects
- Participants in integration education
- Recipients of unemployment payment
- Participants in rehabilitation programs
- Recipients of benefits during childbirth
- Recipients of benefits during sickness
- Recipients of cash benefits
- Recipients of introduction payment
- Children and young people
- Recipients of old age pension
- Recipients of early retirement pay,
- Persons in education
- Recipients of other pensions
- Other recipients of labour market policies payment
- Others outside the labour force.

Special notes regarding socioeconomic status for employees
Employees are divided into levels of skills based on DISCO-88. The level of skills is achieved through education or learning in praxis. The employees are divided into six groups

1. Top managers in enterprises, organizations and the public sector
2. Employees doing work that requires upper level skills.
3. Employees doing work that requires medium level skills.
4. Employees doing work that requires basic level skills.
5. Other employees
6. Employees not specified

Group 1: Includes managerial work at the upper administrative level in enterprises and organizations- regardless of size or skill requirements. The work consists of decision making, planning and controlling etc.

Group 2: Includes work that requires upper level skills. The group includes physicists, building engineers, architects, doctors, midwives, teachers, lawyers, accountants, librarians, journalists, actors, musicians, priests.

Group 3: Includes work that requires medium level skills. The work consists of technical work in production and transport, as well as assistance in trade and administration. Also in consists of education and caretaking in kindergartens and residential homes for children and young persons together with nursing and assisting work in the health sector e.g. laboratory technician, computer programmer, photographer, nurse, kindergarten teacher, real estate agent, police officer.

Group 4: Includes work that requires basic level skills. The work consists of office work, customer service, surveillance, rescue work, farm work, gardening, crafts oriented work and operating different types of stationary machinery.

Group 5: this is a residual group. It contains work that does not belong to one of the other groups e.g. cleaning, delivering services, guard work, packing and transport work without the use of machinery.

Group 6: Contains persons where skill level cannot be determined, due to missing information in the statistic registers.

The period 1997-2003
The information regarding the level of skills for employees working in the State, Counties and Municipalities is collected from Earnings Statistics on central and local government employees. Information on employees in the private sector comes either from E Earnings Statistics for the private sector, or from the unemployment statistics register, or from the education classification module. The information from Earnings Statistics relates to the persons most important job during the year, this might not be the same job, in which the person is employed at the end of November.

The period 2004
From 2004 the level of skills is for 67 pct. of the employed directly connected to the job they are managing at the end of November. Therefore there is a much better coherence between the level of skill and variables of industry and sector. For some part of the employees, information about earnings is still missing. For these persons other sources such as information about education and line of industry, and membership of unemployment insurance fund are used. Today people can choice more freely between unemployment insurance funds. This makes it harder to establish the level of skill, when the only source is membership of such a fund. As a consequence the group: Employees not specified has grown.

About the level of skills
As a consequence of validating information about earnings, the number of reports from the private sector fluctuates across years. Therefore it is advised to be very careful, when using information about socioeconomic status for employees. Analysis of the development in number of employees at a specific skill level must always take into account the development of the employees not specified.

Information about the person originates mostly from other registers in system of person statistics. The following list includes the most important variables:

- Demographic information (e.g. sex, age, county of living)
- Education (completed education, ongoing education)
- Ethnicity (country of birth)
- Income
- Information about participation in Labour market policy measures
- Unemployment
- Extent of working time (full-time/part-time)

Background information on workplace and person
The statistics includes background information about the person and the place of work. The most important information about the workplace is:

-Industry (of the workplace)
-Municipality (of the workplace)
-sector (of the workplace))


2 Time


2.1 Reference Period
Register based-labour force statistics is dated after reference time for the population which is 1st of January. The time of reference for the populations attachment to the labour market, is the last working day in November the year before.

2.2 Date of Publication
The Statistics is scheduled to be published within 16 months after the end of the reference year. Validation and joining of data from the tax system and other registers takes almost 9 month and is the main reason for the late publishing.

2.3 Punctuality
Normally RAS-statistics is published 16 month after the time of reference.

2.4 Frequency
Annual statistics.


3 Accuracy


3.1 Overall accuracy
The data are generally considered to have a high quality.

3.2 Sources of inaccuracy
1.The population of employees is obtained on the basis of data recorded in the Salary Information Register (COR). The delimitation of the population thus depends on the extent to which business enterprises have given correct answers on the form that they are obliged to fill in. From 2003 information from Earnings statistics on the public sector is used as a supplement source to decide whether the job is active ultimo November.

2. In a number of cases data registrations may occur in different registers, which are mutually exclusive. In these cases the most likely registration is chosen after an examination of the registrations. In some cases a secondary job or activity may mistakenly be registered as the main activity/job.

3. The division of employees into level of skills ( top managers, employees in upper levels, employees in medium levels, employees in basic levels, other employees, employees, not further specified), is made on the basis of different data sources of which some have a high quality and others have a slightly lower quality. Information from the statistics on earnings is generally considered to have a high quality, and this statistical source is used wherever possible. Enterprises with less than 10 persons employed do not report data to the statistics on earnings, and consequently the quality of the information about socio-economic status for persons who are employed in these enterprises is lower.

3.3 Measures on accuracy
It is not possible to calculate the margins of statistial error.


4 Comparability


4.1 Comparability over Time
Comparable data exists from 1981 (nov. 1980) to 2002 (nov. 2001). In 2003 (nov. 2002) theres a break in the timeseries. The number of employed persons drops 17.596 as a consequence of the change in method. In connection with the break Statistics Denmark sent out following press release.

Changes in the labour market have created a need for alterations in the registerbased labour force Statistics (RAS). At the same time better sources have made the alterations possible. Therefore Statistics Denmark was able to publish an improved registerbased labour force Statistics. The alterations in the statistics concern the calculation of employed persons and persons outside the work force.

In 1993, the DB93 replaced the DSE77 as the official Danish classification of economic activities, and a break in the time series occurred. However, 1992 was an overlapping year for which both sets of activity codes are available. I 2003 the DB93 classification is replaced by DB03.

In 1997, the division of employees into salaried employees, skilled wage-earners and unskilled wage-earners was replaced by a division based on DISCO-88, which is the Danish version of the ISCO-88 nomenclature.In 1998, a new classification of persons outside the labour force was introduced. In the new classification, participants in labour market policy schemes, recipients of sickness or maternity benefits, recipients of cash benefits, recipients of rehabilitation benefits and recipients of pensions from civil servants' funds are introduced as separate groups. The new classification causes a substantial decrease in the group; others outside the labour force. For most of the subgroups within the group "persons outside the work force" it is not possible to compare the groups before and after the new classification. This however doesnt apply to recipients of pension benefits payable between early.

In 2003 there were changes in the classification of the group "persons outside the work force". From 2003 recipients of unemployment payment are a part of the group. Recipients of benefits during childbirth and sickness are no longer required to come from unemployment. Recipients of economic benefits in labour market policies included in the statistics, but at a low priority.

From 2004 the group "persons in education" range higher than "recipients of other pensions" in the order of priority. This has relevance for young people in education who receive other pensions from a deceased parent. From 2004 these person at categorised as persons in education.

From 2005 the statistics of employment in businesses (EBS) is also source to information about self employed person who only have to pay tax. The main part of this group of self employed persons would also have been categorized as self employed persons before. ESB is also the source of information about industry from 2005. As a consequence some people will be categorized in another industry. Generally theres an increase in the number of self employed persons working in the health sector

4.2 Comparability with other Statistics
The number of people employed and unemployed in RAS deviates from the numbers in other statistics.

RAS and the Labour Force Survey (LFS)
The purpose of the Labour Force Survey (LFS) is to collect information regarding the populations attachment to the labour market. LFS is a survey while RAS is a registerbased statistics. The unit of enumeration - persons - is the same in both statistics and they are compiled according to the guidelines laid down by the ILO. Similar definitions of employment are therefore used as starting point.

However, the reference period used in the two statistics differs, as the RAS statistics measure employment in the last week of November, while the LFS collects information on all weekdays over the quarter. The figure obtained from the LFS is therefore an average of the entire quarter.

The greatest difference between the two statistics is the way in which they are produced. Employed persons in the RAS statistics are classified according to register-based information from the Danish tax authorities. In the LFS people are classified as persons in employment, if it has been stated by the persons themselves that they have been employed for at least one hour during the week prior to the week in which the interview took place.

Taking this fundamental difference into account, there is a high degree of consistency between the statistics regarding the total number employed persons, while there is a greater difference in the number in the three socio-economic groups (self-employed, assisting spouses and employees).


RAS statistics and the statistics of employment in businesses (EBS)
In the EBS statistics the unit of enumeration is the workplace and job. The RAS statistics and the EBS statistics are coordinated statistics and are based on the same primary statistical data. However, some major conceptual differences should be noted. The RAS statistics consider only primary jobs, whereas the EBS statistics cover secondary jobs as well. Conversely, the RAS statistics are more comprehensive than the EBS as regards certain groups of self-employed persons and as regards persons on leave from their jobs at the end of November.

RAS and the ATP statistics
The employment in the RAS and ATP statistics differs. The reasons for this difference is the use of different units of enumeration and populations in the two statistics. The RAS statistics incorporate all employees in employment, while the ATP statistics only provide data on full-time equivalents for wage-earners aged 16-66.

RAS and the National Accounts statistics
Employment in the national accounts is compiled in accordance with the guidelines laid down for the System of National Accounts. Employment is calculated as a simple average of employment in two consecutive years, and persons below 15 are not calculated as employed. A number of further corrections in employment are conducted in the national accounts. This implies that employment is, e.g adjusted for persons on labour market and maternity leave and adjustments are also made with respect to undeclared work.

RAS and the Register-based unemployment statistics (CRAM)
Unemployed persons in the RAS statistics are persons who are fully unemployed in the reference week. In the register-based unemployment statistics, unemployment is calculated as full-time persons, which implies that a person who is not unemployed throughout the whole week will still contribute to the calculation.

RAS og the Annual Working Time Account (WTA)
the number of emplyoed persons in RAS is used in WTAto calculate the number of employed people during the year. People under the age of 15 are not a part of the WTA. For further information look at the declaration of content for ATW.

4.3 Coherence between provisional and final statistics
Provisional results are not compiled.


5 Accessibility


5.1 Forms of dissemination
The statistics are published in
Nyt fra Danmarks Statistik (News from Statistics Denmark)
Arbejdsmarked (Labour market) appearing in the series
Statistiske Efterretninger (Statistical News).

Statistics are also available on the Internet from Statbank Denmark. Summary statistics are given in the Statistical Yearbook.

5.2 Basic material: Storage and usability
The RAS statistics are individually based statistics and information is available for the period 1981-2006.

5.3 Documentation
The variables are documented in TIMES.

www.dst.dk/times

5.4 Other Information
For more information (in Danish) regarding sources, methods and concepts, see the chapter on methodology in: Arbejdsmarked 2004:23 (Labour market 2004:23) appearing in the series Statistiske Efterretninger (Statistical News).


Supplementary documentation
No supplementary documentation is available