IAB Institute for Employment Research Markus Promberger Tel': 49911 179 31 39 email: Markus'Promberg

1
IABInstitute for Employment ResearchMarkus
PrombergerTel. 49/911/ 179 31 39e-mail
Markus.Promberger_at_iab.dePanel Analysis What
added value at what price?Experiences from the
IAB establishment panel
Dr. Markus Promberger
2
Panel Analysis What added value at what price?
Dr. Markus Promberger

• 1. The IAB and the Federal Employment Agency
• 2. The IAB Establishment Panel
• 3. Panel analysis options Interests and use
• 4. Conclusions Added value and price of keeping
  up a panel survey

3
Panel Analysis What added value at what
price?1. The IAB and the Federal Employment
Agency
Dr. Markus Promberger

• The IAB Research institute of the Federal
  Employment Agency,
• Established in 1967 by labour market legislation
• Presently 160 employees (110 scientists)
• Legal task Exploring and explaining the causes
  and structures of unemployment, accompanying and
  evaluating measures to avoid or decrease
  unemployment.

4
Panel Analysis What added value at what price?
1. The IAB
Dr. Markus Promberger

• Sections for
• 1 labour market (LM) and social policy analysis
• 2 International comparison and European
  Integration
• 3 Growth, demography and the LM
• 4 Business cycles, working hours and LM
• 5 Regional LMs
• 6 Establishments and employment (!)
• 7 Education, Occupation and Life Cycle Analysis
• Cross-sectional working groups on
• Temporary agency work
• Linked employer-employee datasets
• Re-thinking occupational research

5
Panel Analysis What added value at what
price?1. The IAB
Dr. Markus Promberger

• The Federal Employment Agency
• Re-established 1952
• Ancestors in the 1920s
• funded from the obligatory contributions of
  employers and employees to unemployment
  insurance
• self-ruled public body
• publicly controlled by a board of stakeholders
  from unions, employers federations
  and government bodies

6
Dr. Markus Promberger
Main tasks of the Federal Employment Agency
1
Paying subsidies for unemployed persons,
Keeping up Job placement services Fighting
unemployment by active labour market
policies Controlling labour migration, agency
work, private job placement services, illegal
labour Keeping up labour market statistics
2
3
4
5
7
Panel Analysis What added value at what
price?2. The IAB Establishment Panel
Dr. Markus Promberger

• structure and history
• Project of IAB section 6
• Started in 1993, limited to Western Germany
• extended to Eastern Germany in 1996
• extended by additional samples for the federal
  states in 1999, enabling statewise analysis
• covering all sectors, including public and NPOs
• covering establishments with 1 or more employee
  liable to social insurance
• stratified sample, due to size, sector and
  region
• standardized face-to-face interviews, based on a
  paper questionnaire

8
(No Transcript)
9
Panel Analysis What added value at what
price?2. The IAB Establishment Panel
Dr. Markus Promberger

• response and return rates, averages
• 60-65 for establishments being questioned for
  the first time
• 80-85 per year for panel repeaters
• 25 for mailed questionnaires (400 cases per
  year)
• still more than 1000 surviving establishments
  from the 1993 wave in 2003

10
Panel Analysis What added value at what price?
Dr. Markus Promberger

• 2. The IAB Establishment Panel
• funding structures
• The mere fieldwork requires a budget of 0,75
  million , contributions by
• IAB research budget (40 of the panel budget)
• Federal states (Governments, state branches of
  Federal Employment Agency, 60)
• Annual and biennial contracts between funding
  parties
• IAB staff costs not included in budget, they are
  covered by the IAB staff budget (60) and by
  external research funding (40)

11
Panel Analysis What added value at what price?
Dr. Markus Promberger

• 2. The IAB Establishment Panel
• Stakeholders
• IAB
• Federal Labour Administration (BA)
• Federal State officials
• Regional offices of the BA
• Cooperating institutes for regional analysis
• additional supporters
• Unions, Employers federations
• Federal Ministery of Labour
• Scientific Advisory Board of the IAB
• Scientific Community

12
Panel Analysis What added value at what price?
Dr. Markus Promberger

• 2. The IAB Establishment Panel
• analytical and publishing work done by
• IAB staff, most from FB6, funded by IAB, BA and
  external research funds, working for scientific
  and/or policy advising purposes
• cooperating institutes, additionally funded by
  federal states, working for the special needs of
  policy makers on state level
• external scientists from all over the world,
  working for scientific purposes

13
Panel Analysis What added value at what price?
Dr. Markus Promberger

• 3. Panel analysis options Interests and use
• policy makers and cooperating regional
  institutes, labour administration
• simple or repeated cross-sectional analysis
  (time series), imputation from recent waves
• external scientists
• Panel and cross-sectional analysis, economists
  mostly interested in panel analysis,
  sociologists mostly in cross-sectional or
  time- series analysis
• IAB staff Both

14
Panel Analysis What added value at what price?
Dr. Markus Promberger

• 3. Panel analysis options Interests and use
• Panel analysis has to be distinguished from
  repeated cross sectional analysis Panel
  analysis observes the same single entities at
  different times, repeated cross-sectional
  analysis observes the same aggregate at
  different times, whether it consists of
  different single entities or not.

15
Panel Analysis What added value at what price?
Dr. Markus Promberger
Direct use of panel analysis options
Low-intensity High-intensity

• Imputation of additional variables from recent
  waves, e.g. measuring employment growth needs
  employment information from two years. Or
  time-accounts were only questioned in 2001, so we
  write them into the 2003 file, if the
  establishment was participating in both years.

• Explicitely longitudinal research design
  Observing single establishments behaviour through
  time for longitudinal analysis required by
  special problems (e.g. effects of changes)

16
Panel Analysis What added value at what price?
Dr. Markus Promberger
Direct use of panel analysis options
Low-intensity High-intensity

• Imputation of additional variables from recent
  waves
• Very common! Crucial part of any sophisticated
  cross-sectional analysis

• longitudinal research design
• not much common outside scientific community
• econometrics

17
Panel Analysis What added value at what
price?4. Conclusions
Dr. Markus Promberger

• Price
• Sampling costs may be higher, esp. for small
  establishments
• Panel surveys usually require high
  response/return rates
• This can be achieved only by expensive
  interviewing methods
• Price difference caused by the panel character
  is about 10-30 of the IAB-panels fieldwork
  budget, cannot be specified more exactly

18
Panel Analysis What added value at what
price?4. Conclusions
Dr. Markus Promberger

• Added Value
• Imputing variables from recent waves allows
  keeping the questionnaire shorter and therefore
  more effective while keeping up a broad scope
  of themes, even if analysis is mostly
  cross- sectional
• Sampling costs maybe lower, esp. for huge est.
• Longitudinal analysis is a crucial tool for
  several problems Evaluation, impact research,
  effects of changes. Growing field!
• Longitudinal analysis brings us more close to
  finding causal relations temporal sequence
  is/seems to be more causal than mere covariation
  
• Panel structure of the survey is crucial for
  scientific reputation or else it is just
  another opinion poll which may cease next year.
• Funders at least in Germany are not
  constantly interested in funding just another
  poll but aim at an instrument of high
  scientific reputation and value.

19
Panel Analysis What added value at what price?
Dr. Markus Promberger
RISKS OF KEEPING UP PANEL OPTIONS
RISKS OF NOT HAVING PANEL OPTIONS

• Loss of scientific interest and reputation?
• Loss of interests from funders?
• Loss of analytical options!
• Disconnecting from scientific developments What
  scientists develop today may be an instrument for
  policy analysis tomorrow!

• Higher costs due to more expensive methods
• Underexhaustion of analytical options, therefore
  problems to legitimize high costs