LIAB Cardiff

1
The Linked-Employer-Employee-Data of the
Institute of Employment Research, LIAB
annual conference of Business Data Linking, BDL,
September 7th to 9th 2005 in Cardiff/UK

• Dr. Holger Alda
• Institute of Employment Research
• Data Research Center
• Nuremberg/ Germany

e-mail holger.alda_at_iab.de
2
Topics

• data sources for LIAB
• LIAB data models
• linking persons and plants to LIAB-Versions
• outlook
• how to get access to LIAB data
• utilities for LIAB data
• summary

3
data sources for LIAB the demand side of the
labor market

• a) the employment statistics register (data for
  persons)
• base is the integrated notification procedure
  for the health, pension and
• unemployment insurance (? freelancers, self
  employed persons and civil
• 
  servants are not included)
• notifications have to be submitted to the social
  security agencies at least
• once a year
• notifications include for example information
  about

• employees entries and exits
• gender, qualification and current occupation
  (3-digit)
• gross daily precise wages (including all bonus
  payments)
• an unique plant identifier and the industry code
  
• since 1999 regional information

each person has an unique identifier (social
insurance number)
4
data sources for LIAB the demand side of the
labor market

• individual data on employment are added by
  unemployment notifications
• - the federal unemployment register contains for
  example information about

• start and end of each benefit notification
• reason for submitting the notification
• type and amount of the benefit
• age, gender, nationality, childs (yes/no),
  married (yes/no)

each person has an unique identifier (social
insurance number)

• both kind of data can be integrated single
  notifications are
• appended to the (un-)employment histories of
  persons

5
data sources for LIAB the supply side of the
labor market
b) the IAB-Establishment-Panel

• is a survey conducted since 1993 on an annual
  basis
• is drawn from a stratified sample following the
  principle of
• optimum stratification (ten size classes and 20
  industries
• (until 1999 16 industries)) from single plants
  (not firms) with
• an unique plant identifier
• to correct for panel attrition and newly founded
  units, the
• samples are augmented regularly ? unbalanced
  panel
• reference point for questions is in regular the
  30th of June
• examples for number of observations
• 1993 4200 1996 8500 2001
  15000

6
data sources for LIAB the demand side of the
labor market
the IAB-Establishment-Panel provides information
about plants
- total employment (also for different skills) -
standard and overtime hours - wage recognition -
output - exports - investment - urbanicity -
ownership - technology (subjective measure) -
organisational change and in-vocational
training - profitability - age of plants and
whether parent is a single firm
and has an unique plant identifier (like in the
administrative data).
7
LIAB data models
8
linking persons and plants to LIAB versions
at present the Data Research Center (DRC) of the
IAB at the Federal Employment Service provides
the following LIAB versions

• LIAB cross-sectional model, version 1
• all employment notifications covering 30th of
  June in plants being part of
• the IAB-Establishment Panel
• between 4,000 and 16,000 plants in East and West
  Germany
• covering 2.5 million employees
• size of yearly individual datasets varies from
  100 to 150 MB
• yearly updates (current 2002)

9
linking persons and plants to LIAB versions

• LIAB longitudinal model, version 1
• all notifications of employees, who work at
  least one day in a period
• of 1996 to 2001 in plants, who have valid
  interviews in the IAB-
• Establishment-Panel in the years 1999 to 2001
  number of observations
• is restricted to 2100 each in East and West
  Germany
• - working histories of such persons are
  applicable from 1990 to 2002
• including times of unemployment ( 1 mill
  (West), East 650 tsd)

• LIAB longitudinal model, version 2
• like version 1 in addition all plants with
  interviews from 2000 to 2002,
• number of observations 5500 in West and 4500
  in East Germany
• ( 1.8 (1.05) mil persons with 28 (18) mil
  notifications)

• trade-off between number of observations
  (plants) and
• technical restrictions

10
summary LIAB-Data in the RDC
working histories persons version 1
persons version 1
t
11
outlook combining the advantages of both LIAB
models
synthetical and original variables of employment
and unemploy-ment notifications for five years
employment notifications for ten years
IAB-Establishment- panel 2000 (plants)
12

• synthetical variables are for example
• daily precise job durations (left censored at
  ten years)
• proportion of times of employment to times of
  employment plus
• unemployment in the last three and five years
• number of unemployment events in the last five
  years
• number of employers in the last five years

• original variables are for example
• last wage by the former employer (date of exit)
• coming to the employer from unemployment,
  employment or otherwise
• ending date of the last notification before
  working for current employer
• next event after exit (unemployment, other
  employer, otherwise)

13
How to get access to LIAB-Data

• authorisation of a research application must be
  in the social services fields
• according to 75 of the Social Code volume X
  (disclosure of social data
• for research and planning)
• applications can be directly submitted to the
  RDC
• application form is linked to the guidelines of
  the German Federal Data
• Protection Act after the notice of approval
  the RDC and the user make an
• agreement
• problem yet we have no english documentation,
  application and
• agreement forms
• external researchers can gain access to slightly
  anonymised data (only) by
• visiting the RDC in Nuremberg (once created the
  dataset for analyses, we
• can arrange controlled remote data access)
• - Software STATA, SPSS, SAS, TDA (others maybe,
  on request)

14
utilities for LIAB data
are all applicable via the internet including

• the documentation of the IAB employment sample
• (data are very similar to LIAB Data for
  persons)
• the documentations for the IAB-Establishment-Pane
  l
• LIAB data reports (several)
• a special LIAB data documentation including code
  plans and a codebook
• of the original data
• anonymised test data (partly invented) for the
  both LIAB data
• models
• STATA/SPSS example codes
• a panel dataset for all the waves of the
  IAB-Establishment-Panel,
• where 30 of the most important plant
  characteristics are
• harmonised

15
summary

• German LIAB data are created by merging the data
  of the IAB-Establishment-
• Panel with the employee and benefit recipient
  history of the IAB
• there are technical restrictions meaning there
  is a trade-off between the
• universe of applicable information about
  persons and the number of plants
• covered by a specific LIAB version
• future activities will make this problem less
  important
• LIAB-Data can be accessed by visiting the RDC
• LIAB-Versions allow to process the data using
  standard software (such as
• STATA or SPSS)
• there are several utilities existing for LIAB
  data making the work with the data
• more easier (all applicable via the internet)
• - remaining questions answer the LIAB literature
  and/or the RDC

16
(No Transcript)