Are unemployment insurance systems in Europe adapting to new risks arising from non-standard employment?

1
Are unemployment insurance systems in Europe
adapting to new risks arising from non-standard
employment?

• Janine Leschke, ETUI-REHS (research unit)
• QMSS conference 2007, Prague, 20-23 June 2007

2
Hypotheses

• Persons with non-standard employment contracts
  face greater risks to make transitions to
  unemployment or inactivity than persons with
  regular contracts
• Unemployment insurance systems that strongly
  incorporate welfare principles instead of
  insurance principles grant better coverage of
  non-standard workers.
• Insurance principle strong equivalence between
  contributions and benefits aims status
  maintenance, contributory justice Germany
  and Spain
• Welfare principle stronger degree of tax
  financing and thus weaker link between
  contributions and benefits (easier
  redistribution) aims preventing and mitigating
  poverty United Kingdom and Denmark

3
Part-time employment by gender, 2006 ( of
total employment of a given sex 15-64)
Source Eurostat LFS data, 2nd quarter 2006.
4
Fixed-term employment by age group, 2006 (
total number employees in age group)
Source Eurostat LFS data, 2nd quarter 2006.
5
The European Community Household Panel (ECHP)

• EU-15 countries
• 8 waves (for most countries 1994-2001)
• Sample of approximatelly 60 500 nationally
  representative households and about 130 000
  individuals (1994)
• Follow up data EU-SILC (from 2004 onwards)
• Advantages
• Country comparisons are possible (common
  questionnaires except for DE, UK, LUX, SE)
• Panel data (dynamic analysis is possible)
• Encompassing information on individuals
  (employed, unemployed, inactive) and households
• Disadvantages
• Sample size restrictions
• Some limitations in variables (f. ex. no
  distinction between unemployment insurance and
  unemployment assistance)

6
Short-term upward, downward, and sideward
transitions from and to full-time and part-time
employment for prime-age workers (25 to 55)
t t1 t1 t1 t1 t1
age 25 to 55 full-time part-time education unempl. inactivity
full-time DK 94.05 1.91 0.66 2.53 0.85
full-time DE 92.54 1.85 0.61 3.49 1.50
full-time SP 90.52 2.16 0.15 4.99 2.18
full-time UK 92.14 3.08 0.20 1.61 2.97

part-time DK 21.21 67.96 2.96 5.42 2.46
part-time DE 18.18 67.57 1.83 2.39 9.85
part-time SP 31.09 46.36 1.12 9.58 11.85
part-time UK 17.42 67.73 0.37 2.28 12.19

Source Own calculation based on European
Community Household Panel (ECHP), waves1994-2001.
7
Longer-term upward, downward, and sideward
transitions from and to full-time and part-time
employment for prime-age workers (25 to 55)
t age 25-55 t4 t4 t4 t4 t4
t age 25-55 full-time part-time education unempl. inactivity
full-time DK 90.75 3.23 1.22 2.89 1.91
full-time DE 86.02 2.71 0.79 5.18 5.30
full-time SP 87.87 2.19 0.05 4.67 5.22
full-time UK 85.62 4.67 0.25 2.78 6.68

part-time DK 37.67 48.34 2.75 6.69 4.56
part-time DE 29.48 51.04 0.92 4.83 13.73
part-time SP 42.46 34.01 0.33 8.66 14.54
part-time UK 28.35 49.76 0.49 3.91 17.58
Source Own calculation based on European
Community Household Panel (ECHP), waves 1994-2001.
8
Characteristics of unemployment insurance systems
that potentially restrict coverage of
non-standard workers

• Part-time workers
• Hours or wage thresholds (access)
• Qualifying period (access)
• Means-testing (access and level)
• Proportionality between benefits and former wages
  (level)
• Fixed-term workers
• Qualifying period (access)
• Benefit period depending on contribution time
  (duration)
• Means-testing (access and level)

9
Advantage of survey data
Registration at employment office by former
working hours
1-19 hours 20-29 hours 30 hours
Denmark 32 76 78
Germany 20 48 78
Spain 73 84 91
United Kingdom 28 28 54
Source Own calculation based on pooled and
weighted ECHP data Basis all unemployed
jobseekers.
10
Unemployment benefit receipt of former full-time
and part-time workers comparison of registered
unemployed (fond) and all unemployed jobseekers
(stripes)
Source Own calculation based on pooled and
weighted ECHP data.
11
Access of non-standard workers to unemployment
benefits (random effects logit model )
Dep. variable unemployment benefit receipt (no/yes) Denmark Germany Germany Spain United Kingdom
odds ratios odds ratios odds ratios odds ratios odds ratios odds ratios
last job part-time 0.32 0.32 0.47 0.80 0.29
REFERENCE reason for stopping last job obliged to stop REFERENCE reason for stopping last job obliged to stop REFERENCE reason for stopping last job obliged to stop REFERENCE reason for stopping last job obliged to stop REFERENCE reason for stopping last job obliged to stop REFERENCE reason for stopping last job obliged to stop
end of contract/temp. job 0.55 0.70 0.70 1.01 0.47
women 1.56 0.58 0.58 0.43 0.36

observations 892 2266 2266 4200 588
prob gt chi2 0.0000 0.0000 0.0000 0.0000 0.0000
rho .24 .48 .48 .45 .38
prob gt chibar2 0.004 0.000 0.000 0.000 0.001
Further variables included in models but not shown here age, household type, qualification, wage before unemployment, current household wage, occupation last job (except for Germany), length of unemployment, year dummies. Further variables included in models but not shown here age, household type, qualification, wage before unemployment, current household wage, occupation last job (except for Germany), length of unemployment, year dummies. Further variables included in models but not shown here age, household type, qualification, wage before unemployment, current household wage, occupation last job (except for Germany), length of unemployment, year dummies. Further variables included in models but not shown here age, household type, qualification, wage before unemployment, current household wage, occupation last job (except for Germany), length of unemployment, year dummies. Further variables included in models but not shown here age, household type, qualification, wage before unemployment, current household wage, occupation last job (except for Germany), length of unemployment, year dummies. Further variables included in models but not shown here age, household type, qualification, wage before unemployment, current household wage, occupation last job (except for Germany), length of unemployment, year dummies.
significant at 10 significant at 5 significant at 1 significant at 10 significant at 5 significant at 1 significant at 10 significant at 5 significant at 1 significant at 10 significant at 5 significant at 1 significant at 10 significant at 5 significant at 1 significant at 10 significant at 5 significant at 1
Source Own calculation based on ECHP data.
Basis all unemployed jobseekers.
12
Individual predicted probabilities of
unemployment benefit receipt for typical former
part-time and full-time workers
ideal type Denmark Germany Spain United Kingdom
predicted probability (95 confidence interval in brackets) predicted probability (95 confidence interval in brackets) predicted probability (95 confidence interval in brackets) predicted probability (95 confidence interval in brackets)
Former female part- time worker in couple without kids 0.62 (0.40-0.84) 0.14 (0.07-0.20) 0.23 (0.15-0.30) 0.11 (0.03-0.19)
Former female full- time worker in couple without kids 0.83 (0.71-0.95) 0.24 (0.16-0.32) 0.27 (0.20-0.34) 0.18 (0.06-0.30)
Age, former wage, current household income,
length of unemployment and qualification level
are set to their mean. Source Own calculation
based on ECHP data basis all unemployed
jobseekers.
13
Average monthly unemployment benefit level by
working time (in purchasing power parities)
Source Own calculation based on pooled and
weighted ECHP data. The figure refers to
unemployment benefit recipients with between 3
and 12 months of unemployment.
14
Net replacement rates at unemployment for former
part-time and full-time workers (approximation)
part-time (gt20 hours) full-time
Denmark 85 55
Germany 88 47
Spain 58 34
United Kingdom 73 22
Source ECHP data, own calculation. Average wage
and unemployment benefit level for people who
worked more than 20 hours.
15
Conclusions

• Non-standard workers are in all countries more
  likely than standard workers to become unemployed
  or inactive
• The expectation that unemployment insurance
  systems of the four countries strongly vary in
  their potential to cover non-standard workers did
  not prove true
• Differences in coverage rates are driven by the
  overall coverage levels rather than by the
  predominance of welfare or insurance principles
• There is evidence for redistribution concerning
  benefit levels

16
Conclusions

• Unemployment insurance reforms did not directly
  tackle deficient coverage of non-standard
  workers, nevertheless, they had some
  (unindended?) influences on the insurance
  situation of non-standard workers
• Hours thresholds have been lowered or abolished
  and some countries introduced favourable benefit
  regulations for workers who change from full-time
  to part-time work
• On the other hand, some of the reform measures
  that aimed at making the systems financially more
  viable worsened the situation of non-standard
  workers
• Abolishment of original unemployment assistance
  (Germany)
• Shortening of reference periods for contribution
  requirements (Germany and Spain)
• Extension of contribution requirements (Denmark)
• Cutting of the duration of non means-tested
  benefits (Germany and the UK)

17
Possible solutions for more inclusive benefit
systems

• Modifying the gendered division of labour
• Supporting high hours part-time employment and
  upward mobility
• Incentives for more equal distribution of
  household/care activities between men and women
• Modifying the design and functioning of
  unemployment benefits
• Basic or minimum insurance for all (but danger of
  further segmentation)
• Abolishing earnings- and hours thresholds and
  installing sufficiently long reference periods
  for contribution payments
• Individualisation of benefit receipt
• Modifications of financing mechanisms
• Encouraging mobility and employability through
  positive activation

• Fighting against the gendered division of labour
• Supporting high hours part-time employment and
  upward mobility (I)
• Incentives for more equal distribution of
  household/care activities between men and women
  (IV)
• Modifications in UI design (II)
• Basic or minimum insurance for all (but danger of
  further segmentation)
• Abolishing earnings- and hours thresholds
• Installing sufficiently long reference periods
  for contribution payments
• Individualisation of benefit receipt
• Modifications in financing mechanisms
• Encouraging labour mobility through positive
  activation

• Fighting against the gendered division of labour
• Supporting high hours part-time employment and
  upward mobility (I)
• Incentives for more equal distribution of
  household/care activities between men and women
  (IV)
• Modifications in UI design (II)
• Basic or minimum insurance for all (but danger of
  further segmentation)
• Abolishing earnings- and hours thresholds
• Installing sufficiently long reference periods
  for contribution payments
• Individualisation of benefit receipt
• Modifications in financing mechanisms
• Encouraging labour mobility through positive
  activation