New
Techniques for the Evaluation of European Labour Market Policies
Research
Training Network
Research
The
Research and Training Network is organised around five themes:
(1)
Youth Education, Youth Programmes and the transition from School to
Work
(2)
Training Programmes and Adult Education
(3)
Unemployment Insurance and Social Insurance
(4)
Subsidies and Tax Breaks for the Less Skilled
(5)
Work Time Reductions.
Participants
of the Network have been engaged in activities in all these areas in the
second year of the network’s life.
The
CEPR team, led by Juan Dolado,
has analysed the implications of over-education for the wage
distribution and the unemployment rate of high and low-educated -educated
workers in countries where a large educational upgrading has taken place
in the last two decades, as is the case of Spain.
The
Uppsala team is involved in
research concerning all five themes. Roope Uusitalo (Young Researcher) is working on several aspects of youth education
in Sweden and Finland. Together with Iida Häkkinen (Young Researcher) he has also been working on the effects of a
student aid reform on graduation rates from university education in
Finland.
Hielke
Buddelmeyer (Young Researcher, IZA)
has concluded his project on investigating the sensitivity of labour
supply to program changes of Dutch disabled workers which resulted in two
working IZA papers (one of them to come out soon). Both papers have been
submitted for publication. Currently he is involved with three ongoing
projects. One project tries to investigates the inherit joint nature of
the decision to supply (market) work and care within a household context,
acknowledging that care encompasses not only caring for children but also
caring for a needy spouse, elderly parents or others. The second project
involves the use of data from a randomized experiment in Mexico to analyze
the performance of several nonparametric estimators. Particular emphasis
is placed on the newly rediscovered Regression Discontinuity design.
Preliminary results were already obtained and submitted for the upcoming
Royal Economic Society Meeting in Warwick to take place in the spring of
2003. The goal of the third project is to analyze the female labor force
participation decision along two dimensions: Over time and by full time,
part time and non-participation. Using cutting edge (MCMC) estimation
techniques it is possible to explicitly account for the correlation of
labour supply over time implying a significant improvment over the current
literature. For this analysis the European Community Household Panel is
used. Preliminary results have been obtained for Germany and the analysis
for several other European countries is now under way.
Euwals
and Winkelmann (both IZA) study
the apprenticeship period and the first labour market outcomes of a
particular cohort of males born between 1960 and 1965. The register data
allows the observation of labour market outcomes from 1975 to 1995. They
investigate training intensities of apprenticeship programmes, and relate
these to retention behaviour, first job durations and post-apprenticeship
wages. The quality of the data guarantees new insights in the functioning
of the apprenticeship system
At
GRECSTA-CNRS, four members
belong to the team: Bruno Crépon, Philippe Lagarde, Eric Maurin, and the
leader of the team, Francis Kramarz. In the “Effect of Payroll Tax
Subsidies on Low-Wage Workers on Firm-Level Decisions”, Bruno Crépon
(with Rozenn Desplats) studies the effect of the large increase of payroll
tax subsidies for low-wage workers that occurred in France in 1995 and
1996. The analysis is based on a key treatment variable :
the ex ante changes in average labor costs in 1994 solely due to
the changes in the tax subsidies between 1994 and 1997. This ex ante
reduction in average labor cost is computed using the “Déclarations
Annuelles de Données Sociales” (DADS) in 1994, an exhaustive employee
level file providing us with the wage of each worker in each firm. To
evaluate this program, they extend the Rubin causal framework to the case
where the economic policies involve a continuous treatment and we define
the ensuing parameters of interest. They make the assumption of
independence conditional on observable and generalize the Rosenbaum and Rubin (1983) propensity
score property. They propose an estimation method based on the
implementation of nonparametric series estimators. They find that, between
1994 and 1997, payroll tax subsidies are associated with very strong
employment effects in the economy as well as on other firms outcomes like
the stock of capital, the share of unskilled workers and the average labor
cost.
In
addition, Bruno Crépon and Francis
Kramarz have had an article on the 1982 workweek reduction in France
(“Employed 40 Hours or Not Employed 39”) accepted by the Journal of Political Economy. Their paper will come out in the
December 2002 issue. In this paper, the authors use longitudinal
individual wage, hours, and employment data to investigate the effect of
the February 1, 1982 mandatory reduction of weekly working hours in
France. Just after François Mitterrand’s election in May 1981, the
government decided to increase the minimum wage by 5%. Then, as promised
in its electoral program, the socialist government reduced the workweek
from 40 to 39 hours. At the same time, it mandated stable monthly earnings
for minimum wage workers and recommended the stabilization of monthly
earnings for other workers (recommendations followed by 90% of the firms).
They show that workers directly affected by these changes—those working
40 hours in March 1981 as well as those working overtime at the same date
—were more likely to lose their jobs between 1981 and 1982 than workers
not affected by the changes—those working 36 to 39 hours in March 1981.
Moreover, because the decree enforcing the new standard was issued
faster than earlier promises, some firms had no time to complete
negotiations and their workers were still working 40 hours after February
1, 1982. They show that these workers were also strongly affected by the
reduction in standard hours. The estimates of the impact of this one-hour
reduction of the workweek on employment losses vary between 2% and 4%,
depending on the methodology or the data used. Furthermore, minimum wage
workers were most affected by the changes. This result, consistent with
the model, is due to the impossibility of adjusting their monthly wage,
which results in excess job destruction and creation. These results should
help us understand the possible effects of the upcoming mandatory
reduction of hours in France, where the maximum weekly working hours
declined from 39 to 35 hours beginning in January 2000. Similar programs
are envisaged in other European countries, which hope that hours
reductions will be an efficient policy for reducing unemployment.
Finally,
with Philippe Lagarde, Bruno Crépon and Francis Kramarz have finished the
construction of a matched employer-employee data base that will allow them
to study changes in the minimum wage at the end of the sixties and at the
beginning of the seventies. This group has already started to analyze the
firm behavior when faced with hikes in the minimum wage of approximately
20%.
Eric
Maurin has pursued a different research program focusing on job
instability in France. In a paper jointly written with Pauline Givord,
submitted to the European Economic
Review (“Changes in Job Security and their Causes: An Empirical
Analysis Method applied to France, 1982-2000”), he analyzes the changes
in the risks of involuntary job loss in France between 1982 and 2000. They
find that these risks are higher in the 1990s than they were in the 1980s.
They develop a model of hiring and separations to interpret these
statistical findings and to separate the effects of institutional changes
from the effects of new technologies. Their estimates on micro-data
suggest that the diffusion of new technologies increases the degree of
substitutability between low and high-seniority workers and has become the
main driving force for the decline in job security in France.
At
FRDB, Tito Boeri (team leader)
and Pietro Garibaldi, have worked on the effects institutional rigidities
on shadow employment in south Italy. The paper has been submitted and
published in the CEPR discussion series (DP 3433) Pietro Garibaldi and
Gianluca Violante (part of the UCL
team) have worked on the difference between severance payment and firing
taxes on aggregate employment. The paper has been submitted and published
in the CEPR discussion paper series (DP 3636). Further, in the area of
unemployment assistance, Tito Boeri and Ignacio Conde Ruiz (Young Researcher, FRDB) have worked on the trade off between
unemployment benefits and severance payments
Garibaldi also wrote an IZA-published
paper with Etienne Wasmer (CEPR)
on ‘Labour Market Flows and Equilibrium Search Unemployment’.
The
NHH team has been working on
several topics over the last year. One of these studies, completed by O.
Raaum, Kjell Salvanes (Team Leader), and Erik Ø. Sørensen, was on The
Neighbourhood is not what it used to be: Has there been equalisation of
opportunity across families and communities in Norway? Parents
influence their children's adult outcomes through economic and genetic
endowments, transmission of cultural values and social skills, and through
choice of residential location. Using a variance decomposition framework
which provides bounds on the effect of families and neighbourhoods, we
find important effects of family characteristics as well as residential
location on educational attainment and adult earnings in Norway. Families
are more important than neighbourhoods as the correlations among siblings
are significantly higher than among children growing up in the same local
community. Sibling correlations are estimated to be a little lower than
for the US, while correlations between neighbourhood children in Norway
are found to be significantly weaker than in the US. Unlike previous
studies, we also assess changes over time by studying children growing up
around 1960 and 1970. While family effects are permanent over time, the
impact of neighbourhoods is reduced by half in size from 1960 to 1970 and
we link this result to several policy changes in the 1960s aimed at
increasing equality of opportunity in Norway. Our results differ from
previous US studies, suggesting that the role of families and
neighbourhoods in explaining the degree of equality of opportunity and
social mobility depends on labour market institutions and redistributive
policies.
The
UCL team have produced work on
a number of areas in the last year or so:
Richard
Blundell has worked on a paper entitled ‘Alternative Approaches to the
Evaluation Problem’ in which he has considered four distinct but closely
related approaches to the evaluation problem in empirical microeconomics:
(i) social experiments, (ii) natural experiments, (iii) matching methods,
and (iv) instrumental methods. The first of these approaches is closest to
the `theory' free method of medical experimentation since it relies on the
availability of a randomised control. The last approach is closest to the
structural econometric method since it relies directly on exclusion
restrictions. Natural experiments and matching methods lie somewhere in
between in the sense that they attempt to mimic the randomised control of
the experimental setting but do so with non-experimental data and
consequently place reliance on strong independence and/or exclusion
assumptions.
Christian
Dustmann addressed the question
of the limitation of union power and minimum wage laws within the context
of the German apprenticeship system. A crucial feature of the system is
that - despite training workers in mostly general skills - it is, at least
partly, financed by firms. It has been accredited with smoothing the
transition from school to work, lowering the youth unemployment rate, and
providing incentives for the non-college bound youth to take up
post-secondary education. His hypothesis is that it is unions which induce
firms to sponsor training. To test this hypothesis, he has set up a model
of firm-financed training that incorporates the particular features of the
German wage setting system. Based on this model, he has derived empirical
implications, and tested them using data on both firms and workers. His
results provide strong evidence for the hypothesis.
Stephen
Machin has looked at the impact of the minimum wage on both wages and
employment in the UK. The
results suggest that the minimum wage raised the wages of a large number
of care homes workers, causing a very big wage compression of the lower
end of the wage distribution, thereby strongly reducing wage inequality.
There is some evidence of employment and hours reductions after the
minimum wage introduction, though the estimated effects are not that
sizeable given how heavily the wage structure was affected.
Machin
also considered to what extent union decline in Britain has been
characterised by convergence or divergence in union membership rates for
people with different personal and job characteristics. He compared data
on individual union membership in 1975, from a period when union
membership was high and growing, to data in 2001 data when it is low and
has been falling for over twenty years. Some factors of both convergence
and divergence are identified which should be useful to many parties,
including industrial relations scholars and union organisers.
Return to Introduction
Research