Studying social policy in time

1
Studying social policy in time

• Giuliano Bonoli

2
Social policy in time

• At the micro-level longitudinal analysis
• Social processes
• Transitions between employment and non employment
• Social exclusion
• Event history analysis
• Child birth
• Disability recipiency
• But at the macro-level?

3
Three time-related processes

• Path dependency
• Example labour market policies
• Sequencing
• Example pension reform
• Relative timing
• Example Adaptation to new social risks

4
1. Path dependency

• In the early days of a policy, the room for
  manoeuvre is relatively large. Early decisions
  are relatively unconstrained
• Early decisions, however, put policies on course
  from which it is difficult to exit
• Illustration through Polya Urn processes in
  mathematics
• Illustration through the QWERTY keyboard layout
  in economic history

5
Why path dependency ? Increasing returns (P.
Pierson)

• Increasing returns
• Policy change requires collective action.
• Institutional density or stickiness
• Increasing returns in political authority
• Complexity of politics and policies

6
Example Labour market /industrial relations
policy in Britain and in Germany in the 1980s
(Steward Wood)

• PUZZLE
• UK Liberalisation, deregulation
• Germany status quo prevails in spite of attempts
  by the Kohl government in the same direction
• HYPOTHESES
• Employers preferences are different

7
Employer preferences in the UK and in Germany
shaped by EPL

• UK
• low employment protection allows employers to
  adjust quickly to demand fluctuations, and keep
  short term profits high
• Provides an incentive for workers to acquire
  transferable skills.
• Rigidities are detrimental to profitability

• Germany
• high employment protection generates a stable
  labour market which discourages free riding
• provides incentives for workers to invest in
  their own skills
• Collaboration with the unions is important for
  providing training

8
UK The Thatcher governments reduced the power of
the trade unions dramatically
9
Germany failed attempts to undermine union power
because of employers

• attempt to restrict access to unemployment
  benefits for workers indirectly affected by a
  strike.
• Attempt to undermine DGB monopoly in works
  councils

10
Conclusion

• Decisions taken decades earlier (interwar period
  for Germany) have shaped current employer
  preferences

11
2. Sequencing. Why do policy processes go through
sequences?

• Social learning (problem solving)
• Strategic learning by political actors
• Spill over effects (adjustment in the
  Netherlands).
• Staged processes with extension of the range of
  available options

12
3. Relative timing of key developments

• Key socio-economic developments are not
  synchronised across countries
• Their relative timing may produce interactions
  that can affect the patterns of
  opportunity/constraints for policy change

13
Old welfare states and new social risks

• Western welfare states were concieved and
  developed during the postwar years
• Socio-economic changes have resulted in the
  emergence of new social risks
• How are welfare states adapting to this?

14
it depends

• Nordic welfare states have been considerably more
  successful in adapting
• Continental European welfare states are lagging
  behind
• Timing of different developments explains
  divergence

15
Social change and new social risks

• Deindustrialsiation and the tertiarisation of
  employment
• Womens entry into labour markets
• Increased instability of family structures
• Destandardisation of employment

16
Old and new social risk policies

• Old
• Pensions
• Survivors ben.
• Short term unemployment ben.
• Sickness benefit
• Invalidity ben. and serv.

• New
• Long term unempl. ben./ALMP
• In work benefits
• Child care serv.
• Family ben.
• Parental leave
• Services for older people

17
It is justified to distinguish between two sorts
of social policies, because

• They constitute responses to different social
  transformations, and have different objectives
• The target groups of the two sets of policies are
  different
• Why not?

18
Spending on old and new social risk polices as a
of GDP, averages 1997-2001
Source OECD SOCX 2004
19
How do we explain divergence?

• Politics
• Institutional predisposition
• The relative timing of postindustrialisation,
  ageing and welfare state maturation

20
The timing of key postindustrial developments in
18 OECD countries
Source Based on OECD Statistical compendium
21
Relationship between the average benchmark year
and spending on new social risk polices, 1997-2001
Source Based on OECD Statistical compendium
22
Alternative explanations 1 the strength of the
left
Source OECD SOCX and Armingeon et al. CPDS
23
Alternative explanation 2 the strength of the
Christian democrats
Source OECD SOCX and Armingeon et al. CPDS
24
Alternative explanation 3 Catholicism
Source OECD SOCX and www.adherents.com
25
Competing explanations correlation matrix
Source see previous slides
26
Alternative explanation 4 catching upIncrease
in spending on new social risk polices in the
1990s and spending in 1987-1991
27
Conclusions

• Timing matters, but together with politics
• If timing matters, we should develop new
  techniques to introduce time base variables in
  policy making models
• Not good news for those who are hit by NSR in
  Continental European countries