Register-based statistics on the Danish labour market - new possibilities in relation to external user needs

1
Register-based statistics on the Danish labour
market - new possibilities in relation to
external user needs
2
Employment statistics in Denmark

• Labour Force Survey - EU official source
• Register-based Labour Force Statistics - national
  official source
• socio-economic classification - ILO-guidelines
• Covers the period 1981-2007
• Annual register with reference point-in- time at
  the end of November and population at the start
  of the following year
• Very detailed statistics

3
Linking of employer - employee
Using information from the Salary Information
Register
Information on persons
Information on establish- ments
Job
- Period of employment - Job characteristics
- Firm identifier - Establishment identifier
PIN (Personal identification number)
Possibility of combination with other registers
at Statistics Denmark using unique identifiers
4
Registration of employment period

• Information for employee to Tax Authorities
• from Employer
• employed the whole year
• employed one period (from to)
• employed several periods
• employed on 30th of November
• (yes/no)

5
Access to data

• Statistics published in publications from
  Statistics Denmark
• Statbank Denmark
• External researchers scheme unidentifiable
  micro data administered by the Research Service
  Unit

6
Past requirements/developments

• Longitudinal labour market statistics/labour
  market dynamics - transitions between status
  observations end of November IDA database
• Further subdivisions of the socio-economic groups
  outside the labour force - subdivision of the
  group in the Register-based Labour Force
  Statistics
• Information on hours worked - Working Time
  Accounts
• Foreign workers/transfrontier commuting - nordic
  cooporation between statistial offices

7
IDA database - information over time
Establishment
Establishment
Jobs
Jobs
Person
Person
Time
8
Weaknesses of the register

• Only information on attachment to the labour
  market at the end of November
• The quality of the period indications is
  relatively low, which gives rise to uncertainties
  in the employment levels for persons at the
  fringe of the labour market
• The timeliness of the statistics is not
  particularly high

9
Information on the scope of hours worked
Low quality of information implying

• uncertain statistical data on employment in
• full-time equivalents
• uncertain distribution by full-time and part-time
  employment
• an element of uncertainty in the classification
• due to poor priority of main job vs. second job

10
New data source e-income

• Replaces the salary information register
• Monthly reports on all employee jobs, 12 reports
  instead of one report per year implis better
  information on duration of each job
• Information on hours worked
• Unique identification code for the establishment
  and person
• Information on hours worked
• Information for employees

11
Another new statistical system persons who are
not in ordinary employment

• Information on
• unemployed persons
• person in different labour market policy schemes
• sickness benefit
• early retirement pension
• Contains longitudinal data (integrated)
• which are subject to processing of overlaps from
  different data sources
• with start and end dates and
• information on hours

12
Integration of the two new data sources

• Would cover many of the socio-economic states
• employees
• unemployed
• persons temporarily outside the labour force
• early retirement from the labour market
• Could provide information in full-time
  equivalents
• Could ensure better quality in employment
  statistics
• qua data subject to processing of
• overlaps on the basis of several data sources

13
Identifying external user needs

• Investigating user needs concerning labour
• statistics. What we did

• labour market seminars
• individual meetings/dialogues with customers
• advisory research committees
• user survey (written and oral hearing)

14
Need for more timely statistic

• needs in relation to analyses, monitoring,
  forecasts, etc.
• Statistics Denmarks respons
• fast short-term statistics on the basis of
  e-income as the only source
• the decision as to compiling fast structural
  statistics await further examination,
• including examination of the supplementary
  sources on the populations attachment to the
  labour market
• ensuring the socio-economic classification, etc.
• Reality there is a high correlation between
• external funding
• and
• the possibility of faster structural employment
  statistics

15
Need for longitudinal statistics

• Need for more detailed transitions than from end
  november to end
• november
• gross flows
• transition from education to labour market
• impact measurement of the job activation efforts
• gross flows, e.g. between
• industries
• sectors
• job types and career paths (quality of work)
• geography
• retirement from the labour market
• job mobility (flexicurity)
• job creation/job destruction

16
Need for employment in full-time equivalents

• full-time equivalents defined on the basis of
  more exact hourly information
• possibility to relate full-time employed to other
  full-time equivalent measures (unemployment in
  full-time equivalents etc.)
• summable goals for full-time persons, which can
  be related to the mean population figure

17
Need for statistics on foreign work/transfrontier
commuting

• analyses of transfrontier commuting
• analyses of the use of foreign labour
• would imply more representative statistics with
  regard to
• regional distribution of employment frequencies,
• the number of workplaces in the border regions

18
The need for a more flexible classification

• possibility of varying the ILO classification
• which in relation to some users needs,
  attaches too great importance to employment
• flexible classification of persons who are in
  several states at the same time
• more advanced socio-economic classification
  (dynamic classification)

19
Experiences from the dialogue with the users

• Need for integration of data sources to develop
  longitudinal based data for the populations
  attachement to the labour market
• The work-sharing between the labour market
  researchers and Statistics Denmark is to be
  reconsidered
• who is to be responsible for which parts of the
  data processing and
• which consequences will this have on the quality
  of the primary statistical data
• Considerations with regard to the number of
  breaks in the data series
• Considerations with regard to an expedient use of
  the economic resources
• Continuous dialogue and hearing of the users are
  of importance
• NSI should be responsible for concepts, but there
  should be a dialogue with users when new concepts
  are developed