Modelling unemployment inflows and outflows in the Italian Labour Market

1
Modelling unemployment inflows and outflows in
the Italian Labour Market

• Baussola, M., Fabrizi, E., Mussida, C.
• Catholic University of the Sacred Heart
  (Piacenza)
• Monitoring Italy 2009 Measuring the progress of
  Italian society
• Isae, Rome, 3-4 June 2009

2
Labour mobility and unemployment in Italy

• Labour mobility has always been a controversial
  issue
• the debate originated in the 70s depicted the
  Italian labour market as typically affected by
  insider power
• the idea that Italian labour market was
  substantially tight was prevailing, but no
  accurate data were available to investigate this
  issue in depth

3

• New information available from the Italian
  Institute of Statistics (Istat) during the 80s
  help to build a dynamic picture of the labour
  market
• the Italian labour market was treated as a single
  and tight market with insufficent mobility
• such a pitfall in the analyses has been only
  recently recognised (Trivellato, 2005) by
  investigating Italian labour mobility in depth
• earlier studies (Baussola, 1985) emphasized how
  labour market flows were significant and how they
  affected the steady-state unemployment rate

4

• The 90s had been characterised by debate on the
  need of new labour market legislation, aimed to
  improve labour flexibility and increasing
  employment opportunities
• new regulations have been introduced the main
  interventions refer to Treus legislation package
  and, at the beginning of the new millenium, to
  Biagis law
• but aggregate comparable data on labour market
  flows for the decade 1993-2003 do not show any
  significant impact on aggregate labour mobility.

5
Stocks and Flows in the Labour Market,
longitudinal population 1993-1994 (thousands)
ue 448 (20.1)
ne 897 (5.6)
eu 392 (2)
en 1,113 (5.7)
un 673 (30.2)
nu 825 (5.2)
Transition probabilities in brackets Source
Authors calculation using weighted longitudinal
LFS micro data (Istat)
6
Stocks and Flows in the Labour Market,
longitudinal population 2002-2003 (thousands)
ue 427 (20.3)
ne 831 (5.6)
eu 270 (1.3)
en 701 (3.4)
un 542 (25.7)
nu 664 (4.5)
Transition probabilities in brackets Source
Authors calculation using weighted longitudinal
LFS micro data (Istat)
7
Stocks and Flows in the Labour Market, EU-SILC
longitudinal population 2004-2005 (thousands)
ue 519 (19.2)
ne 627 (3.5)
eu 564 (2.6)
en 1,345 (6.2)
un 586 (21.7)
nu 764 (4.2)
Transition probabilities in brackets Source
Authors calculation using weighted longitudinal
EU-SILC micro data
8
Stocks and Flows in the Labour Market,
longitudinal population 2006-2007 (thousands)
ue 395 (21)
ne 729 (5.6)
eu 300 (1.3)
en 751 (3.3)
un 401 (23.5)
nu 325 (2.5)
Transition probabilities in brackets Source
Authors calculation using weighted longitudinal
LFS micro data (Istat)
9
The determinants of labour market transitions a
multinomial logit approach

• Explanatory variables
• sex, age, agesq, married (unmarried), hhsize
• northwest, northeast, centre, (southislands)
• unempl2004, empl2004
• qu1, qu2, qu3, qu4, (qu5)
• comp, diploma, (degree)
• transition
• () base categories in brackets

10
Marginal Effects after MNL for outflows from U
11
(continue)
() dy/dx is for discrete change of dummy
variable from 0 to 1 () only vaiables relevan
for our investigatio are reported
12
Marginal Effects after MNL for outflows from E
13
(continue)
() dy/dx is for discrete change of dummy
variable from 0 to 1 () only vaiables relevan
for our investigatio are reported
14
Estimates and implications

• Successful exits from U (UE) are more strongly
  determined by geographical dummies, lagged U
  status, quintiles of income distribution,
  presence of labour market transitions within the
  twelve months. Also education plays a role
• Sex, age and its squared, and marital status are
  the main drivers for the exits from the labour
  force (UN)
• There is evidence of state dependence for UE the
  state of U acts as a trap for individual by
  causing a sort of lock-in into this condition and
  by alimenting the persistence
• The estimated coefs for quintiles of the
  equivalised disposable income are all negative
  and significant with respect to the fifth
  quintile (base)

15
(continue)

• Education plays a role relevance of tertiary
  education attainment title
• The presence of labour market transitions within
  the twelve months increases the probability of
  making other transitions in the consecutive
  period
• Being male reduces the odds of exiting the labour
  force relative to remaining U (UN) by 30
  compared to a female being male raises the
  likelihood of leaving U with success (UE) by 21
  percentage points in 2005-2006

16
Conclusions

• We have set up an empirical framework to
  analyzing labour market transitions and their
  determinants
• our findings suggest that human capital, income,
  geographical residence, experienced mobility and
  state dependence strongly determine UE
• these variables, with the exclusion of human
  capital proxy and the inclusion of the sex dummy,
  are also crucial in causing EU
• these results enable us to complete the puzzle of
  labour mobility in Italy and its effect on U and
  U duration
• an increase in this latter is inferred although
  labour mobility is not negligible, especially
  with respect to U stock

17
(continue)

• Labour mobility is encouraged, on tho one hand,
  by previous transitions in the labour market, and
  by variables reflecting personal characteristics
  or structural factors such as the geographical
  location
• Labour mobility is strongly affected and as
  concern the U outflow, is strongly reduced by
  previous U conditions
• This evidence suggests existence of hard core
  unemployment not affected by those changes in
  labour legislation introduced in the last decade
  this core U is therefore related to structural
  factors which implicitly suggest the need of
  policy interventions.