THE DETERMINANT OF YOUTH DISADVANTAGE. A PANEL DATA ANALYSIS.
Francesco Pastore Seconda Universit di Napoli, IZA of Bonn and
Italian Association of Labor Economists Email:
[email protected]@unina2.it Luca
Giuliani Seconda Universit di Napoli Email:
[email protected]
Slide 2
Aims of the presentation To give you some information on key
stylised facts To provide a frame of mind to understand the youth
unemployment problem, its causes and, in part, its consequences To
provide criteria to Define the objective And assess the ex ante
effectiveness of youth employment policy To understand the
evolution of the debate on this issue: From educational system to
determinant of youth disadvantage Keywords: youth experience gap;
SWT; flexicurity;
Slide 3
OUTLINE Youth experience gap Two policy approaches: The
liberalist view The interventionist view Aim of presentation
Methodology and data Descriptive analysis Static and dinamic panel
data estimations Conclusion
Slide 4
Understanding the nature of youth unemployment from stylised
facts The flow in and out of unemployment is higher for young
people (Clark and Summers, 1982) because: Young people are in
search for their best job-worker match And sometimes go back to
education and training Especially low-skill young people Employers
are also in search for their best job match Consequences of job
shopping: Shorter average unemployment duration, but For the
low-skill: High risk of falling into a chain of low pay temporary
or part-time work Two paths for high-skill and low-skill young
people)
Slide 5
The youth experience gap Young people have low human capital:
Despite increasing educational attainment, they lack the other two
components of human capital: generic and job-specific work
experience; To fill in the youth experience gap, young people move
in and out of employment in search for the best job-worker match;
This search requires the least skilled to go back to education
and/or training schemes; Youth unemployment is temporary;
Slide 6
The liberalist (optimistic) approach Why to bother? High YU is
the consequence of a search for the best match by young people and
employers The only thing to do is: make the market more flexible to
increase the chances for young people to find a good job The market
will provide training via temporary work, in the case of countries
that implemented two-tier reforms Lower entry wages would be the
solution to the lower degree of work experience of young
people
Slide 7
Three main objections to the liberalist approach and to
temporary work A. Beckers (1962) hypothesis of market failure for
job specific training B. Heckman and Borjas (1980) hypothesis of
omitted heterogeneity C. Empirically, it amounts to ask: Temporary
work arrangements: Are they stepping stones or dead end jobs?
Slide 8
Arguments against labour market flexibility and temporary works
PROS It is a stepping-stone for young people to find their best
job-woker match Employers have a buffer stock of workers in
expansionary times to be possibly fired in downturns Employers pay
low wages for low productivity; Employers try young people; No
compansation mechanisms but special intervention is needed only for
particularly weak young people; Cons Sometimes they become low-pay
traps Low pay jobs become permanent And cause precariuosness of
labour market experiences (so-called training trap); Increasing
power of insiders Therefore, the need for constraints to the use of
temporary work; High cost for households;
Slide 9
Evidence on temporary work as stepping stone In almost all
countries the net impact on employability is positive Spain and
some USA programmes are exceptions Ichino et al. (2005, 2008) find
a net positive effect of 19 in Tuscany and 11% in Sicily (here only
weakly significant) a gross effect of 31 and 23%, respectively
Berton, Devicienti and Pacelli (2008) find also evidence of a trap:
increased probability of temporary work to remain such
Slide 10
Arguments against labour market flexibility: why temporary work
may be a trap? Becker (1962) Temporary contracts reduce the gap in
generic, not in job specific work experience; Short-time horizons
work as a disincentive for employers and employees; Bentolilla and
Dolado (1994) find for Spain an increase in the youth experience
gap as a consequence of temporary work (see also Acemoglu 2002)
Heckman and Borjas (1980) duration dependence in an individuals
unemployment experience often is an artefact of statistical data;
Once controlling for unobserved heterogeneity, duration dependence
disappears It is not a consequence of low labour turnover, but of
low skills and motivation; Education and ALMP are the
solutions.
Slide 11
Fields of intervention: different policy mix The educational
system: Rigid versus flexible systems Sequential versus dual
systems The welfare system Pro-active schemes versus passive income
support Targeting and scale of expenditure State- versus
family-based welfare systems Fiscal incentives
Slide 12
The interventionist view: Flexicurity Flexicurity, meaning
Passive income support during unemployment spells and social
security provisions to increase the cost of temporary work for
firms and reduce the Pro-active schemes to increase employability
Employment rather than job stability
Slide 13
Objection: The training trap Also ALMP may become a trap: young
people move from a programme to the next to get the linked benefits
Van ours (2004) find a looking-in effect for training programmes
participants in Slovakia Wunsch and Lechner (2008) find a similar
effect in Germany Caroleo and Pastore (2005) find evidence of a
training trap for some young people involved in training programmes
in ALMP in Italy
Slide 14
Interventionist view: Reforms of the education and training
systems Increase the quality of education, trhough: Evaluation
mechanisms Incentives for quality increasing intevention Introduce
a duality principle: Training should be provided together with
general education Favour smooth STW transitions through job
placement: Dual system in Germany Jisseki Kankei in Japan Job
placement services in Anglo-Saxon countries
Slide 15
EU SWT models la Esping-Andersen? The German system: The dual
educational system The Scandinavian system: ALMP on a large scale
The Anglosaxon system: High quality of education and labour market
flexibility The South Mediteranean System: The family and temporary
work The new member states: Building a modern welfare system
(????)
Slide 16
The Aim The red line of this paper is using econometric
analysis to empirically test whether after controlling for some
variables, such as: GDP, youth population, secondary & tertiary
education attainment, ALMP & PLMP, employment protection index
There remains part of the YUR and RD differential across-countries
that is still unexplained and that can be caught by School-to-Work
transition regimes ==
Slide 17
DATA: SWT regimes and countries SWTR is a set of 5 dummy
variables that represent school-to-work transition regimes:
North-European System: 1 if Estonia and Sweden, 0 otherwise;
Dual-Educational System: 1 if Belgium, Germany, Austria,
Netherlands, Denmark, France, Slovenia, Luxemburg, 0 otherwise;
Anglo-Saxon system: 1 if United Kingdome and Ireland; 0 otherwise;
South European System: 1 if Greece, Italy, Portugal, Spain; 0
otherwise; New Member State System: 1 if Poland, Slovakia, Hungary,
Estonia and Czech Republic; 0 otherwise;
Slide 18
ModelVariableDescriptionUnit of Measurement Yl_yur1524 Youth
unemployment rate (15- 24) Percentage, log X dl_gdpGrowth of GDP
US$ current prices, difference of log l_gdpGDPUS$ current price,
log l_yupopYouth population (ylf/tlf)Thousand of persons, log
l_edu2Secondary educationPercentage, log l_edu3Tertiary
educationPercentage, log l_epiEmployment protection indexIndex of
costs, logs l_almpActive labour market policies Public expenditure
as a percentage of GDP, log l_plmpPassive labour market policies
Public expenditure as a percentage of GDP, log D D_NECountry dummy1
or 0, binary D_CECountry dummy1 or 0, binary D_ASCountry dummy1 or
0, binary D_SECountry dummy1 or 0, binary D_NMSCountry dummy1 or 0,
binary CONTROL VARIABLE
Slide 19
Statistic Panel Data Fixed Effect Model (FE) - Is the dependent
(endogenous) variable, - is a time invariant individual effect - it
measures the effect of all the factors that are specific to
individual i but constant over time, - is a row vector of
observations on K explanatory STRONGLY EXOGENOUS factors for each i
at time t, not including the constant term. It means that is not
correlated with present or past. If it does not hold you will use
dynamic panel.
Slide 20
Fixed Effect Model (FE) The FE model can be divided in to two
parts: take in mind that the Within Estimator for fixed effect
model could be written as: Where is the mean of the observations on
the outcome for individual I, and, is the mean row of the
observations on the explanatory variable x for individual i. The
Between Estimator could be written as:
Slide 21
Random Effect Model (RE) The random effects model is an
alternative to the Fixed effects model. The estimation equation is
the same: contrary to the Fixed effects: the random effects are
assumed not to be estimable-in contrast with Fixed Effect that can
be estimated-; they measure our individual specific ignorance which
should be treated similarly to our general ignorance.
Slide 22
It come out that the transformation for each individual-time
observation is: Where is : =
Slide 23
RANDOM EFFECT= OLS RANDOM EFFECT = FIXED EFFECT
Slide 24
Hauseman Test The natural question that arises after
introduction of RE and FE models is: Which one should we use?. True
Consistent More Efficient Inconsistent [1] [1] Remember that the
null hypothesis is that RE Model is the correct one (p-value ha sto
be smaller than 0.05)
Slide 25
Dynamic Panel Data In order to measure the persistence of the
results in long-run and short-run a lagged variable should be
introduced in the previous model.
Slide 26
GMM We can apply directly to this model IV. Assume the
existence of a T*r matrix for instrument, where r K is the number
of instruments, that satisfy the r moment conditions: The GMM
estimator based on these moment conditions minimizes the associated
quadratic form:
Slide 27
GMM estimator Where denotes an r x r weighting matrix. Given
some algebra gives the Panel GMM estimator: The essential condition
for the existence of this estimator is, once again, :
Slide 28
One-Step Panel GMM The one-step GMM or two-stage leaste-square
estimator uses weighting matrix : leading to: This estimation is
called one-step GMM because given the data it can be directly
computed using the equation above.
Slide 29
Two-Step Panel GMM The two-step GMM is based on the
unconditional moment of using weighting matrix, where, where is
consistent S defined as: Using you have the two-step GMM estimator
:.
Slide 30
The Arellano-Bond estimator The microeconomics literature
refers to the resulting GMM estimator as the Arellano-Bond
estimator. The estimator is: Lags of or can additionally be used as
instruments, and fore moderate or large T there may be a maximum
lag of that is used as an instrument, such as not more than..
Slide 31
Constructing the Panel Data The data bank includes 21 countries
observed over a period of 10 years, by 2000 till 2011. The number
of variable used was around 97. Hence, the Panel was composed by
231 observation. VariableUnitNameSOURCE EPI_CIndices of costs
Employment Protection Index_Collective Labour>Employment
Protection> Strictness of employment protection collective
dismissals (additional restrictions) EPI_I Indices of costs
Employment Protection Index_Individuals Labour>Employment
Protection>Strictness of employment protection individual
dismissals (regular contracts) LTIR Long Term Interest rate General
Statistics > key short-term Economic indicator > Long Term
Interest Rate
Slide 32
AI Annual Inflation Prices and Purchasing power>prices and
prices indices > consumer price (MEI)>consumer prices- Annual
inflation RIR index (where the year 2005 is the base year) Real
Interest Rate Finance>Monthly financial statistics>monthly
monetary and financial statistics(MEI)> interest rates GDP US $,
current prices, current PPPs, millions real GDP (98- 2012) National
Account> Annual national account>Main aggregate> gdp>
Gross domestic product (GDP) MetaData : GDP, US $, current prices,
current PPPs, millions EMPL Thousands of persons Empoyed (98- 2012)
Labour>LFS>Short-Term labour market statistics>Employed
population YUR1519 percentages. Youth Unemployment 15-19
Labour>LFS>LFS by sex and age- indicator>unemployment rate
YUR2024 percentages. Youth Unemployment 20-24 Labour>LFS>LFS
by sex and age- indicator>unemployment rate YUR1524 percentages.
Youth Unemployment 15-24 Labour>LFS>LFS by sex and age-
indicator>unemployment rate UR1564 percentages. Unemployment
rate 15-64 Labour>LFS>LFS by sex and age-
indicator>unemployment rate ALMP public expenditure as
percentage of GDP Active labour market policies Labour>LAbour
Market programmes>public expenditure as percentage of GDP>
Active PLMP public expenditure as percentage of GDP Passive labor
market policies Labour>LAbour Market programmes>public
expenditure as percentage of GDP> Passive UR2564
percentagesunemployment rate 25-64 Labour>LFS>LFS by sex and
age- indicator>unemployment rate
Slide 33
RD=(YUR1524/UR2564) Relative Deasdvantag Computated APOP
Thousands of persons Active Population aged 15 and over
Labour>LFS>Short-term statistics>short term labour market
statistics>Active population YUPOP Thousand of persons Youth
population (lfs1524/tf) EDU3 percantage Tertiary education
Education & training> Education at Glance> Appendix
A>Atteined tertiary education degree, 25-34 years old(%) EDU2
percantageSecondary education Education & training>
Education at Glance> Appendix A>attained below upper
secondary education, 25-34 years old(%)
Slide 34
Expectation on Betas VariableExpectation on Betas sign
Employment Protection Index Positive GDP & GDP growthNegative
ALMP and PLMPPositive Secondary and Tertiary Education Negative
Youth populationPositive Active Youth PopulationPositive
Slide 35
Descriptive Analysis
Slide 36
GDP
Slide 37
GDP growth
Slide 38
Youth Population
Slide 39
Secondary High School attainment
Slide 40
Tertiary Education attainment
Slide 41
ALMP
Slide 42
PLMP
Slide 43
Ratio of Expenditure for policies
Slide 44
EPI
Slide 45
Active youth population
Slide 46
Expectation on Beta Throughout the descriptive analyses it is
clear that the expectation made theoretically are all fulfilled.
The only variable that could create problems is going to be
tertiary education attainment, found with positive beta.
Slide 47
RE estimation for YUR Variable modRE 1 modRE 2 modRE 3 modRE 4
modRE 5 modRE 6 GDP
growth-0.329***-0.227**-0.208**-0.331**-0.329**-0.310** Youth
population0.6261.8383.736**1.041.2472.156 Southern
Europe0.062-0.402**-0.464*-0.309-0.197 Anglo-
Saxon-0.464*-0.881***-0.928***-0.908***-0.795*** Central
Europe-0.553***-1.011***-1.005***-0.879***-0.765*** Northen
Europe-0.115-0.734***-0.777**-0.37 Secondary
Education-0.232***-0.250***-0.162*-0.137-0.267** Teritary
Education0.203**0.201**0.353***0.354***0.186 PLMP0.341***0.372***
EPI0.0750.501***0.587*** ALMP0.0850.0290.107
constant2.782***2.547***1.879**1.584*1.1821.993*
Slide 48
FE estimation for YUR
VariablemodFE1modFE2modFE3modFE4modFE5modFE6 GDP
growth-0.337***-0.074-0.083-0.312** -0.318** Youth
population-1.19731.182***30.614***14.507** 14.579** Southern
Europe(omitted) Anglo-Saxon(omitted) Central Europe(omitted)
Northen Europe(omitted) Secondary
Education-0.330***-0.342***-0.314** -0.298** Tertiary
Education0.335***0.354***0.353** 0.325** PLMP0.528***0.502***
EPI-0.299*0.324 ALMP0.242*** 0.269***
_cons3.334-10.740***-10.865***-4.171 -3.774 N229223 203
ll-4.59281.96980.30817.234 16.034 AIC15.184-149.937-148.615-20.468
-20.068
Slide 49
Hauseman Test on model 2
Slide 50
Hauseman test on model 3
Slide 51
LSDV estimation for YUR VariableLSDV1LSDV2LSDV3LSDV4LSDV5LSDV6
GDP growth-0.382***-0.390***-0.384***-0.295**-0.288**-0.257***
Youth pop0.909*0.8450.991*1.900**1.609**2.769*** Southern
Europe0.057-0.107*-0.058-0.079-0.151
Anglo-Saxon-0.466***-0.737***-0.712***-0.666***-0.752*** Central
Europe-0.551***-0.835***-0.701***-0.692***-0.777*** Northern
Europe-0.108*-0.02-0.351***0.215* Secondary
Education-0.003-0.071*0.0220.001-0.176*** Tertiary
Education0.298***0.248***0.378***0.399***0.049 PLMP-0.0280.052
EPI0.821***0.919***0.840*** ALMP-0.208**-0.145**-0.223**
_cons2.657***0.909*2.116***-0.2080.1061.771** N229223 203
ll-73.846-31.306-62.637-22.076-23.926-94.75
aic159.69382.613143.27564.15265.851199.5
Slide 52
YUR: to sum up VariablemodFE2modRE2LSDV2 GDP
growth-0.074-0.227**-0.390*** Young pop31.182***1.8380.845 Southern
Europe(omitted)-0.402**-0.107*
Anglo-Saxon(omitted)-0.881***-0.737*** Central
Europe(omitted)-1.011***-0.835*** Northern
Europe(omitted)-0.734***-0.02 Secondary
education-0.330***-0.232***-0.003 lTertiary
education0.335***0.203**0.298*** PLMP0.528***0.341***-0.028
EPI-0.299*0.0750.821*** _cons-10.740***2.547***0.909* N223
ll81.969-31.306 AIC-149.937.82.613
RE estimation for RD Variable modRE a modRE b modRE c modRE d
modRE emodREf GDP growth0.577***0.664***0.667***-0.0090-0.002 Youth
pop-2.479**-2.534**-2.499**-1.073-0.843-0.625 Southern
Europe0.1210.2410.2510.209**0.116
Anglo-Saxon0.3050.510**0.519***0.459***0.359** Central
Europe0.1490.450***0.471***0.295***0.217** Northen
Europe0.160.520**0.486**0.371*** Secondary
education-0.113-0.117-0.022-0.031-0.018 Tertiary
education-0.351***-0.363***-0.029-0.0320.009 PLMP-0.151***-0.142***
EPI0.103-0.031-0.142 ALMP-0.117***-0.099**-0.107***
_cons2.559***3.735***3.880***1.951***2.115***1.821*** N229223
203
Slide 55
FE estimation for RD
VariablemodFEamodFEbmodFEcmodFEdmodFEemodFEf GDP
growth0.605***0.683*** 0.013 0.016 Youth pop5.519-9.009-9.0374.293
4.253 Southern Europe(omitted) Anglo-Saxon(omitted) Central
Europe(omitted) Northern Europe(omitted) Secondary
education-0.327** -0.013 -0.022 Tertiary
education-0.639***-0.638***-0.003 0.012 PLMP-0.179***-0.181***
EPI-0.015-0.179 ALMP-0.121*** -0.136***
_cons-0.8428.598***8.592***-0.132 -0.352 N229223 203
ll-22.284-5.769-5.77144.765 143.473 aic50.56825.53723.541-275.53
-274.945
Slide 56
LSDV estimation for RD Variabl e LSDVaLSDVbLSDVcLSDVdLSDVeLSDVf
GDP growth 0.765***0.756***0.758***-0.146***-0.139**-0.112** Youth
pop -2.799**-2.403**-2.361*-2.152***-2.481***-2.294*** Southern
Europe 0.124***0.209***0.223***0.099***0.017 Anglo-Saxon
0.299***0.338***0.345***0.340***0.243*** Central Europe
0.148***0.301***0.340***0.134***0.038 Northern Europe
0.1430.414***0.317***0.242*** Secondary education
-0.028-0.047-0.033-0.058**-0.053** Tertiary education
-0.039-0.054-0.02400.076** PLMP -0.158***-0.135*** EPI
0.240***0.082-0.007 ALMP -0.010.061***0.050*** _cons
2.703***2.335**2.687**2.474***2.828***2.500*** N 229223 203 ll
-55.162-42.703-45.45650.65145.89537.348 aic
122.324105.407108.913-81.301-73.789-64.696
Dual educational system result to be the best Reforms are
needed in order to improve educational system Youth Guaratee:
-apprenticeship -2015-2020, funds for ALMP CONCLUSIONs
Slide 59
Future research The next goal is to do the same work also
taking into account the Relative Disadvantage