Unemployment Duration and the Interactions Between
Unemployment Insurance and Social Assistance∗
Michele Pellizzari†
March, 2005
Abstract
Existing studies of unemployment benefits and unemployment duration suggest that reforms
that lower either the level or the duration of benefits should reduce unemployment. Despite the
large number of such reforms implemented in Europe in the past decades, this paper presents
evidence that shows no correlation between the reforms and the evolution of unemployment.
This paper also suggests an explanation for this fact by exploring the interactions between
unemployment benefits and social assistance programmes. Unemployed workers who are also
eligible for some social assistance programmes should be less concerned about their benefits
being reduced or terminated. Data from the European Community Household Panel are used
to provide evidence to support this argument by exploiting variation in a particular feature
of family cash benefits, which are universal in some European countries and means-tested
in others. Comparing similar persons subject to different family assistance schemes shows
significant differences in the elasticity of the probability of finding employment to both the
generosity and the duration of the unemployment benefit.
Keywords: Unemployment duration, unemployment insurance, social assistance, duration
models.
JEL Codes: J64, J65
1
Introduction
The effect of unemployment insurance (UI) on unemployment duration is the object of many
studies in a rather large literature. Two empirical findings are now widely accepted. First,
∗
I would like to thank Tito Boeri, Marco Leonardi, Alan Manning, Steve Pischke, Federico Perali and Bar-
bara Petrongolo for useful comments and suggestions. This paper has received useful comments from seminar
participants at the IZA Summer School, CEP-London School of Economics and University of Verona. Financial support from the ESRC, Bocconi University, the Centre for Economic Performance and the University
of Verona is gratefully acknowledged. All errors are my own responsibility.
†
IGIER-Bocconi University, IZA, CEP and fRDB. Contact address: michele.pellizzari@unibocconi.it
1
as initially showed by Nickell [1979] and Lancaster [1979], higher benefits are associated
with longer unemployment spells. Later, Moffit [1985] and Meyer [1990], having access to
information about both the level and the duration of benefit entitlement at the individual
level, were able to show a second important empirical finding, that the probability of exiting
unemployment increases around the time of benefit exhaustion.
The literature also provides a comprehensive theoretical framework for interpreting these
results. Mortensen [1977] develops a simple search model that easily delivers negative correlation between exit rates and unemployment benefits via search effort and reservation wages
being respectively negatively and positively affected by income out of work. When UI entitlement expires, income out of work suddenly drops, inducing an instantaneous increase in
search effort and a decrease in the reservation wage, thus providing an explanation for the
observed higher exits rates around the time of exhaustion. Cahuc et al. [2000] recently extended the model by endogenising wages and allowing the unemployment benefit to gradually
decline over time.
On the basis of these findings, one would expect reforms that reduce either the level
or the duration of unemployment benefits to have a positive impact on the unemployment
rate. However, this prediction does not seem to conform with the recent experience of many
European countries. Figure 1 plots the time series of the unemployment rate for selected
European countries. The vertical lines indicate the implementation years of reforms that
have modified either the level or the duration of the unemployment benefits. The solid
bars refer to changes in the amount of the benefits and the dashed lines to changes in their
duration. The thickness of the lines indicate the direction of the change: thick for reductions
(either in the amount or duration of benefits) and thin for increases.
A first simple fact is evident from Figure 1: the past 20 years have been constellated by
labour market reforms in virtually all European countries. Moreover, despite the coexistence
in many countries of reforms of opposite sign, often implemented close to each other (in
Finland and France, for example), most of the changes (19 out of 29) modified the system
towards less generous benefits paid for a shorter time. However, already a simple visual
inspection of Figure 1 suggests that the correlation between these reforms and the evolution
of unemployment is rather weak.
2
More convincing evidence can be produced for those countries where reforms took place
during the years covered by the European Community Household Panel (ECHP)1 . This is
possible for seven countries: Austria, Belgium, Finland, Germany, Greece, Ireland and the
United Kingdom. Table 1 shows the conditional difference in the probability of finding
employment between individuals who entered unemployment before and after the reform.
The estimates are produced with a standard hazard model controlling for gender, age, health
status, education, marital status, family size, presence of children, household income and
regional unemployment2 . Results confirm the visual impression from Figure 1: reforms that
reduced the level or the duration of unemployment benefits did not have a significant effect
on the probability of finding a job. There are marginally significant effects only in Austria
(where the estimate is actually negative) and Ireland.
The scarce success of labour market reforms in Europe has been acknowledged by other
authors as well, for example Coe et al. [1997] and Orszag et al. [1999]. This papers suggests
an explanation for the failure of so many reforms of the unemployment compensation system. European countries all have complex welfare states and unemployment compensation
is only one element of the system that necessarily interacts in various ways with all the other
programmes. In particular, many unemployed persons receive other social assistance benefits
together with their unemployment insurance. Most of these other benefits are means-tested,
therefore a reduction or an anticipated withdrawal of unemployment insurance is often compensated, at least partly, by higher transfers from other programmes. Moreover, even those
unemployed who only receive unemployment insurance may still expect to become eligible
for some social assistance programmes when their benefits expire. For these workers, too,
reductions or exhaustion of unemployment insurance are less of a concern and do not affect
much their search effort nor their reservation wages.
These arguments will be tested empirically using data from the European Community
Household Panel, which allow to reconstruct monthly labour market histories for samples
of individuals from all EU countries. Contrary to most studies in this literature that use
data from administrative sources, this paper exploits survey data which have two main
1
2
The data are described in details in Section 4.
A fuller description of the econometric model is provided in Section 6.
3
advantages. First, the ECHP contains information on numerous types of social transfers
and thus permits to identify UI recipients who are also entitled to other social assistance
programmes. Second, by relying on self-reported information about one’s labour market
status it is possible to reconstruct unemployment spells that end into employment and into
inactivity separately. Hence, I will be able to estimate the probability of finding a job rather
than that of leaving unemployment or the UI registry. This is an important innovation: in
the largest sample used in this study exits into inactivity represent about 15% of total exits
from unemployment.
However, using survey rather than administrative sources necessarily lowers the quality
of the data on UI payments and durations. In particular, the total amount of UI benefits
received by each individual in the ECHP is only recorded annually (not monthly) and the
duration of payments is not known. This is solved by imputing benefit entitlements on the
basis of individual characteristics available from the ECHP and institutional information
about the functioning of the welfare system in each country and year.
Moreover, the ECHP, by providing data that are highly comparable across countries,
allows to use variation in the administrative rules of the welfare systems to identify causal
effects. In particular, the institutional features of family cash benefits offer an interesting
source of variation. In 10 out of the 15 pre-enlargement EU countries eligibility for family
benefits is determined only by the number and the age of the children in the household,
regardless of income. Only in the remaining 5 countries - Germany, Greece, Italy, Portugal
and Spain - family benefits are subject to means testing. This paper exploits this variation
to identify the differential effect of unemployment benefits on the job finding probabilities
of potential recipients of family benefits. Intuitively, in countries where family benefits are
means tested, UI recipients who are potentially eligible for child assistance will be less worried
about reductions or withdrawal of their benefits, as these could be compensated by higher
social assistance payments. In countries were family benefits are ”universal”, i.e. not means
tested, there should be no such interaction effect. Results indicate significant differences in
the elasticity of the job finding probability to the unemployment benefit and to the time of
benefit exhaustion between male heads of households with children subject to means tested
versus universal family benefits.
4
The importance of interactions between welfare programmes has lately been recognised
by both academicians and policy makers. Belot and van Ours [2000] provide evidence from
macro data showing that countries where unemployment has fallen often owe their success
to comprehensive rather than piecewise reforms of labour market policies. A theoretical justification for the importance of these interactions is discussed in Coe et al. [1997] within a
search and matching framework. Despite the acknowledged importance of potential overlappings between welfare programmes, specific evidence from micro data is still lacking. This
paper aims at filling this gap by providing detailed evidence on one specific interaction.
The paper is organized as follows. Section 2 briefly describes the institutional details of
unemployment benefits and other social assistance programmes in Europe. Section 3 shows
how interactions between welfare programmes can be analysed in a standard search model.
Section 4 describes the data while Section 5 discusses the empirical strategy. The results of
the empirical analysis of re-employment probabilities are presented in Section 6. Section 7
concludes.
2
Unemployment Compensation and Social Assistance
in Europe
2.1
Institutional Framework
All European countries have large and well developed welfare states, nevertheless there still
exist important differences in the institutional details across countries. The comparison of
different types of welfare states in Europe has been the object of a large number of studies
in both the economic and the political literature3 . This section simply aims at providing a
very general overview of the welfare programmes available in the European countries during
the period covered by the empirical analysis that follows.
Excluding old age pensions (which are still the major component of social expenditure in
many countries), welfare benefits are generally grouped into 5 large categories: unemployment related benefits, family cash benefits, invalidity benefits/pensions, housing benefits and
3
Bertola et. al [2000], Esping-Andersen [1990], Ferrera [1998].
5
general social assistance. In kind benefits are not considered here, even if they might play
an important role for some groups of beneficiaries (disabled persons, large families, et.).
Unemployment benefits are generally distinguished into unemployment insurance and
unemployment assistance. Unemployment insurance is a rather standard insurance scheme
by which workers who have paid sufficient contributions out of their salaries are allowed
to receive a compensation if they become unemployed. Unemployment insurance is usually
rather generous but benefits are typically paid only for a limited period of time (with Belgium,
where the duration of payments is unlimited, being a notable exception) and various reeligibility conditions apply for repeated spells of unemployment. Clearly then, unemployment
insurance does not cover the whole population of jobseekers: young workers - and anybody
who has not paid enough contributions - and the long term unemployed - who have exhausted
their benefits - cannot claim the benefit. For this reason many countries have introduced a
parallel unemployment assistance programme that pays a (typically lower) benefit to those
who, for any reason, are not entitled to unemployment insurance. Along with these general
characteristics, the details of the unemployment benefit schemes vary widely across European
countries.
Invalidity benefits are another important component of the welfare system and often take
the form of pensions. The typical scheme pays a benefit to individuals whose capacity to
work and earn is substantially reduced by some sort of invalidity. Almost all countries also
require some kind of contribution conditions. There are few exceptions to this general rule:
in the Netherlands there is no qualifying conditions while in Finland and Sweden the main
conditions relate to residence in the country rather than to contribution records. The level
of the benefit is usually determined on the basis of a measure of ”normalised” earnings,
i.e. earnings of a similar person who does not suffer from the invalidity. In some countries
invalidity benefits also vary with age but payments are always carried over to retirement, at
which point an invalidity pension is typically converted into an old age pension.
Family cash benefits are the most important welfare programme that is not related to
employment. In the majority of countries (10 out of 15) family benefits are actually paid
to any household with children, regardless of their income. These benefits are paid until
the child reaches a certain age and the amount varies according to the child’s age and to
6
the number of children in the household. Some countries also offer supplements for single
parents. Children who undertake higher education or training are often allowed to receive
benefits for some additional years above the age limit.
The provision of housing benefits is more varied. Some countries offer a generalised
housing benefit available to everyone whose income is sufficiently low (Germany, France,
Netherlands, Finland) while others simply provide specific housing supplements for those on
low-income benefits (Ireland, Luxembourg, Austria and Portugal). Denmark and Sweden
have both. Some countries do not offer any housing benefit but often social housing is
available for low income families.
Finally, all countries, with the exception of Greece and Italy4 , also provide a general
social assistance scheme that aims at preventing poverty for those individuals or families
who do not qualify for any of the other ”categorical” benefit or who still remain under a
variously defined income threshold. Qualifying conditions for this type of assistance are
usually related to nationality, residence and age. All countries also require those who are
able to work to prove that they are actually willing to take up job offers and beneficiaries
are often required to participate in training or other active labour market programmes.
Synthetic tables that summarise the details of welfare programmes in the European
countries can be found in Appendix B.
2.2
Welfare Reforms in Europe
Table 2 presents a more complete description of the reforms indicated in Figure 1, including
additional details on changes in social assistance programmes as well. In the past two decades
policy makers have apparently been listening to economists as many reforms have actually
changed the unemployment compensation system in the direction of lower benefits paid for
shorter periods.
Much less effort has been put in reforming other welfare programmes and one of the main
claims of this paper will be that the scarce success of UI reforms is partly due to the lack of
coordination with other parts of the welfare system.
4
A minimum income scheme was experimented in Italy between 1998 and 2002 (Reddito Minimo
d’Inserimento) but was never introduced on a universal basis.
7
Looking at Table 2 some notable facts emerge. First, in the effort to reduce unemployment
in the recession of the early 90s, there has been a clustering of reforms between 1992 and
1996. Secondly, the large majority of reforms clearly focused on unemployment insurance and
have typically taken the form of reduction of either the duration or the level of the benefits.
Many countries have also tightened eligibility conditions or strengthened work requirements
for the unemployed. Only few reforms addressed other welfare programmes and even fewer
tried to comprehensively change several programmes at the same time (Germany in 1998,
Ireland in 1993, the United Kingdom in 1996).
3
A simple theoretical framework
The interaction between unemployment benefits and social assistance can be analysed in the
framework of a very standard search model. The theory in this section is a mere revised and
simplified version of that in Mortensen [1977].
In each period t, an unemployed worker needs to invest leisure time (st ) to search for job
offers. These arrive with per-period probability αst , where α is a constant, from an exogenous
wage distribution F (·) with support (0, w]. In each period utility is a non-decreasing function
of income (yt ) and leisure (lt ), u(yt , lt )5 . For simplicity, assume that working time is constant
and equal to h in all jobs and that the total endowment of leisure is normalised to 1. Then,
per-period utility while unemployed with benefit bt and searching for a job is u(bt , 1 − st ),
while a job that pays wt generates a utility flow equal to u(wt , 1 − h). Unemployed workers
also need to choose a reservation wage wt∗ : only wage offers above wt∗ are accepted, all the
others are turned down.
Under these assumptions the value of unemployment U at time t can be written as:
(1 + r)Ut = u(bt , 1 − st ) + αst [Pr {w < wt∗ } Ut+1 + Pr {w ≥ wt∗ } E {V (w)|w ≥ wt∗ }]
(1)
where V (w) is the value of employment at wage w. For simplicity assume that there is
5
The per-period utility satisfies the standard assumptions, i.e. twice differentiable with:
∂u(y,l)
∂l
> 0 and
∂u(y,l)
∂y∂y
≤ 0,
∂u(y,l)
∂l∂l
≤ 0.
8
∂u(y,l)
∂y
> 0,
no job destruction: once workers enter employment they stay in the job forever at constant
wage6 . Equation (1) can be rewritten as:
(1 + r)Ut = u(bt , 1 − st ) + [1 − αst (1 − F (wt∗ ))] Ut+1 + αst
Zw
V (w)dF (w)
(2)
wt∗
The optimal levels of wt∗ and st are then chosen in order to maximise equation (2), according
to the following first order conditions:
V (wt∗ ) = Ut+1
⎤
⎡
Zw
∂u(bt , 1 − st )
⎥
⎢
= α ⎣ V (w)dF (w) − (1 − F (wt∗ ))Ut+1 ⎦
∂st
(3)
(4)
wt∗
The interpretation of these two conditions is very intuitive. Equation (3) shows that the
optimal reservation wage is set at a level that equalises the value of employment and unemployment. An unemployed worker can allocate time to two different activities, search and
leisure, hence optimal search time equalises the marginal utilities of search and leisure, as
shown in equation (4). Note that both st and wt∗ are time-varying: equations (3) and (4)
hold for all t and, for any known sequence of benefits, {bt }∞
0 , identify a series of reserva-
∞
tion wages, {wt∗ }∞
0 , and optimal search times, {st }0 . The per-period probability of exiting
unemployment - the hazard rate - is then calculated as:
qt = αst · [1 − F (wt∗ )]
(5)
These results are useful to analyse the implications of different assumptions about the
sequence of benefits for the exit rate. Equation (5) shows that the exit rate is higher when
job search is more intense and when the reservation wage is lower, i.e. when unemployed
workers are less choosy about wage offers.
6
In this case:
V (w, h) =
∞
X
u(w, h)
(1
+ r)j
j=t
Introducing exogenous job destruction does not modify the empirical implications of the model.
9
Let us now analyse how search time, reservation wages and exit rates look like for different
time profiles of the benefit.
Unemployment benefit without social assistance Consider the standard case of an
unemployed worker who receives a constant unemployment benefit (b) for a given number of
periods, T , and nothing after that (this is the specific case discussed in Mortensen [1977]).
For such worker the value of unemployment decreases over time as periods of positive
benefit payments run out and expected future income out of work decreases, i.e. Ut+1 < Ut .
Equation (3) implies that the reservation wage also decreases over time. Similarly, the right
hand side of equation (4) increases with time (the value of unemployment enters with a
negative sign and the derivative with respect to wt∗ , given equation (3), is zero7 ), therefore
in order for the equality to hold the left hand side must increase as well and this can only
be achieved with higher st , i.e. search time also increases with time.
At time T , when unemployment benefit entitlement expires, bt discontinuously drops to
zero. This requires the reservation wage to jump down and optimal search time to jump
up. Note incidentally that for these effects to be non-ambiguous leisure and income must be
complements (i.e. u21 (bt , 1 − st ) ≤ 0). These results are represented in Figure 2.
Unemployment Benefit and Social Assistance The previous analysis can be easily
extended to a worker who receives social assistance together with his/her unemployment
benefit, or, similarly, to somebody who expects to become eligible for some social assistance
programmes once his/her unemployment insurance expires. Eventually all changes from
one scheme to another simply generate jumps in the time profile of the benefit and can be
analysed within the same framework used for understanding exhaustion of unemployment
benefits in the previous paragraph.
It may for example be the case that, given the particular rules and household composition
of applicants, family cash benefits top up household income once unemployment insurance
7
The derivative with respect to wt∗ of the right hand side of equation (4) is:
∂RHS
= α [V (wt∗ ) − Ut+1 ]
∂wt
which is zero at the optimum.
10
expires, leaving the time profile of benefit payments flat. In this case the model predicts no
discontinuous jumps in reservation wage, search effort and exit rate, which will all remain
constant throughout the entire unemployment spell.
In other instances it might happen that payments under social assistance are actually
higher than under unemployment insurance. This possibility, although rare, can occur in
some countries where social assistance is particularly generous (see OECD [2002]). In such
an extreme case the value of unemployment increases with time and all the effects derived
previously are reversed, as shown in Figure 3.
4
The data: the European Community Household Panel
Most of the existing studies of unemployment insurance and unemployment duration make
use of administrative data8 . The advantage of these data usually consists in having detailed
information about the amount and the sequence of payments as well as about individual
eligibility and entitlement conditions.
However, for the purpose of this paper the use of administrative data would be problematic for at least two reasons. First, in many countries unemployment insurance and social
assistance programmes are administered by different governmental bodies and data available
from one body rarely include information about benefits paid by others. Therefore, using
administrative data would make it very difficult to look at interactions between different programmes. Secondly, and probably more importantly, even if comprehensive administrative
data were available, in order to explore the interactions between different programmes one
would need to compare similar individuals facing different unemployment benefits and social
transfers: in other words one would need enough variation in the rules and regulations of
both unemployment insurance and social assistance. However, there is typically little variation in such rules within one country9 and for identification purposes it would be helpful to
8
Boeri and Steiner [1998], Katz and Meyer [1990], Lancaster [1979], Meyer [1990], Moffit [1985], Naren-
dranathan and Stewart [1993b].
9
This is especially true for unemployment benefit while social assistance is more varied, being often
administered at the local level (but this also makes it more difficult to obtain information about the system
as well as about the beneficiaries).
11
use some cross-country variation as well. Unfortunately, cross-country comparable administrative microdata are simply not available. Alternatively one could use some exogenous time
variation induced, for example, by a reform but, as already mentioned in Section 2, there
hasn’t been much reforming in social assistance programmes over the past years.
In order to overcome these problems, comparable cross-country survey data are utilized in
this paper. Data come from the European Community Household Panel (ECHP), a panel of
households and individuals from EU countries produced by Eurostat in cooperation with the
member states statistical offices. The main advantage of this data source is the high level
of cross-country comparability. This is guaranteed by standardised sampling procedures,
defined by Eurostat and implemented by each country’s national statistical office. Moreover,
identical questions are asked to households sampled in each country, merely translated into
the local language. Sample sizes differ from country to country, with the highest sample-topopulation ratios for the largest and the poorest countries10 .
The ECHP started in 1994 and 8 waves of data have been released so far, covering the
period from 1994 to 2001. Not all countries entered the survey at the same time and for three
of them - Germany, Luxembourg and the United Kingdom - the original sample has been
replaced after the first three waves with harmonised versions of household panels already
been produced nationally: the German Socio-Economic Panel (GSOEP), the Luxembourg’s
Socio-Economic Panel (PSELL) and the British Household Panel Survey (BHPS). When
possible data from the existing panels have been provided for the first three years too.
For the purpose of this paper it is important to note that the ECHP includes information
about unemployment benefit payments and social assistance transfers received both at the
individual and at the household level. Moreover, it also contains retrospective information
which allow to reconstruct employment/unemployment/inactivity monthly spells. In fact,
individuals are interviewed once per year and at that time they are asked to report their
monthly labour market status over the previous calendar year. One drawback of these data
is the fact that all the variables are recorded annually and, as we will see later on, this will
make it difficult to attach the correct numbers to each unemployment spell.
10
see Peracchi [2002] for a detailed description of the ECHP.
12
5
The empirical strategy
The theory of Section 3 suggests that unemployment insurance recipients who also receive
some social assistance should be less sensitive to both the level and the duration of their
benefits. However, simply comparing UI recipients who receive some social assistance to
those who do not receive any additional benefits might not identify any causal effect. The
selection into social assistance is, in fact, likely to be highly non-random.
Ideally, one would like to rely on a natural experiment where UI recipients are randomly
assigned to a treatment group of persons eligible for means-tested social assistance and
a control group of persons who are either not-eligible for social assistance or eligible for
universal, non-means tested, benefits. UI recipients in the treatment group will be less
concerned about their benefits being reduced or withdrawn, as they could be compensated
by higher social assistance payments. Such interaction effect should not affect the control
group.
The ECHP allows to exploit variation in the institutional details of family cash benefits
across European countries that mimic this ideal experiment. Specifically, in 5 countries Germany, Greece, Italy, Portugal and Spain - family cash benefits are means tested while
in all the other EU-15 countries family assistance is universally assigned according to the
number and the age of the children but regardless of household income. This paper focuses
on the subsample of UI recipients who are male heads of households with children and assigns
them to a treatment group of those who are eligible for means-tested family benefits and a
control group of those who are either eligible for universal family assistance or not eligible
altogether because the children are above the age limit. As long as the generosity of family
benefits does not affect the decision about one’s country of residence11 and about when to
have children, the selection into these treatment and control groups can be considered fairly
exogenous.
Table 3 reports the eligibility rules of family cash benefits in all EU-15 countries and
the distribution of the sample by treatment and control groups. Eventually, the sample
11
An attraction effect of welfare generosity could be present for migrants, which are a minority in this
sample, but is likely to be negligible for Europeans, who are highly immobile in general and especially when
they have children.
13
includes 2,703 unemployment spells experienced between January 1994 and December 2001
by male heads of households who received unemployment benefits at some point during their
unemployment spells. The sample is divided into a treatment group of 1,042 persons who
are entitled to means-tested family assistance and 1,661 persons who are either entitled to
universal assistance (because they live in countries where family benefits are not conditional
on income) or are not eligible at all (because the children are above the age limit).
In order to avoid further issues related to recipiency of other benefits, the sample is also
restricted according to age, health status and home ownership. In particular, the persons in
the sample are aged between 25 and 50 and are thus very unlikely to receive any type of old
age benefits. They are also in good health conditions and they report owing their houses for
the entire duration of their unemployment spell, thus making them ineligible for both sickness
and housing benefits. Some of the households in the sample might still be entitled to some
minimum income schemes, so the estimates reported below should be interpreted as lower
bounds because potential actual or expected entitlement to minimum income programmes
could affect the sensitivity to the level and the duration of unemployment benefits in the
control group as well.
Furthermore, left-censored spells have been dropped from the sample to avoid stocksampling bias. Unemployment spells, thus, end into either employment or inactivity or are
right-censored. Spells experienced by new entrants in the labour market are also excluded
because these workers are rarely entitled to unemployment insurance. Greece, Luxembourg,
the Netherlands and Sweden have been excluded because information on retrospective employment status is lacking for these countries.
The ECHP is a collection of country samples drawn from the total population following
common procedures but independently. This means that the sample-to-population ratios
differ from country to country and the observations need to be weighted accordingly when
pooled together across countries. The 2,703 observations represent 685,888 unemployment
spells which are distributed across countries as shown in Figure 4. This distribution is obviously influenced by both the relative size of each country and the level of the unemployment
rate experienced in each area (plotted in Figure 4 along the dashed line and scaled on the
right hand side vertical axis).
14
As already mentioned, this paper uses cross-country variation in the administrative rules
of the welfare system (unemployment insurance and family cash benefits, in particular) to
identify causal effects. This approach, however, requires controlling for all other possible institutional differences across countries. A lot of effort has been put in recent years, especially
by the OECD, in producing quantitative measures of labour market institutions and a rather
complete set of indicators is now available for statistical analysis. This paper fully exploits
these indicators and the set of explanatory variables used for the estimation always include
the following yearly measures: the regional rate of unemployment, the regional rate of longterm unemployment, expenditure in active labour market programmes (as a % of GDP), the
share of part-time dependent employment, the share of temporary dependent employment,
the legal minimum wage (normalised by the median wage)12 , the average effective tax rate13 ,
union density and the index of the strictness of employment protection legislation14 .
Before proceeding with the empirical analysis two issues need to be discussed. First,
although the ECHP allows the identification of monthly unemployment spells, it reports
all other variables, included the amount received in ”unemployment related benefits”, only
annually. Hence, it has been necessary to compute a monthly measure of unemployment
benefit entitlement. Second, the pre-unemployment wage is a crucial variable for the analysis
of the duration of unemployment but in the ECHP only workers who are employed at the
time of the interview report their current wages. Thus, this variable is missing for many
observations. These two issues are discussed in the following sub sections.
5.1
The imputation of monthly unemployment benefit payments
The most obvious solution to compute a monthly unemployment benefit consists in simply dividing the annual amount by the number of months spent in unemployment during a
particular year. However, this approach would generate zero variation in individual unemployment benefits over time, unless a spell spans over more than one year. Moreover, unless
12
13
This variable is set ot zero for those countries where there is no legal minimum wage.
This is computed as the total tax wedge (income taxes plus all social security contributions) over total
gross labour costs for a single production worker who earns the average wage in the manufacturing sector
(the Average Production Worker, in the OECD terminology).
14
See OECD[2004a] and OECD [2004b] for more details about these indicators.
15
both the amount of the benefit and the number of months of unemployment are exactly measured, this approach is likely to generate some spurious covariance between monthly benefits
and unemployment durations. In fact, the duration of a spell is very highly correlated with
the number of months spent in unemployment in one year (it is actually exactly equal to
that number if the unemployment spell begins and ends in the same year).
In order to solve this problem, monthly unemployment benefits have been imputed on the
basis of country specific rules and regulations. In fact, in all countries the amounts and the
duration of unemployment benefits are calculated on the basis of individual characteristics,
most of which are easily available from the ECHP: previous employment records, previous
wage, age, family composition, et.. Combining these data with the rules of each country’s
unemployment benefit system, a rather precise imputation of both the levels and the duration
of payments can be obtained.
In econometric terms, this procedure is equivalent to instrumenting the benefits with
the country’s unemployment insurance regulations. This means that, beyond providing a
monthly measure of benefits, using these imputed values makes the estimates robust to
potential endogeneity of unemployment benefit receipt.
The imputation routine requires two basic ingredients: a detailed description of the unemployment compensation system in all countries and years, and all relevant personal characteristics. The institutional features of all welfare programmes in the member countries
of the European Union are systematically collected in the MISSOC15 , a yearly publication
of the European Commission that contains comparative descriptions of rules and regulations of welfare programmes in the member states. Additional complementary information
can be extracted from institutional databases created by other research institutions, like
the Fondazione Rodolfo Debenedetti and the CESIfo centre. As for the relevant personal
characteristics, most of them are directly available from the ECHP.
Combining these two sources - institutional details from the Missoc and other sources
and personal characteristics from the ECHP - it is possible to write imputation procedures
for each country and year. The outcome of these procedures is a vector of imputed variables including the duration of benefit entitlement in months and the monthly sequence of
15
Mutual Information System on Social Protection in the Member States of the European Union.
16
payments16 for every individual in the sample.
Evaluating the goodness of this imputation is not an easy task. One possibility is to
compare the cumulated annual amount of imputed benefits with annual income from ”unemployment related benefits”, as recorded in the ECHP. This is done in Figure 5 where
imputed annual benefits are shown on the vertical axis against reported annual benefits on
the horizontal axis. Although the two measures appear to be strongly correlated, there still
is a large dispersion around the 45-degree line.
For two countries - Greece and Italy - the figures are particularly odd and require some
explanations. Unemployment benefits in Greece, like in many other countries, are subject
to a minimum and a maximum level. Here, however, the distribution of annual benefits is
highly concentrated around the minimum. It seems like the large majority of the unemployed
in Greece report receiving only the minimum benefit, even when the imputation suggests
they should be entitled to higher transfers.
Italy is another anomalous case: in this country there are many individuals who appear
to be entitled but receive no benefits. This result is likely to be due to the high degree
of discretionality of the Italian system. In fact, in this country unemployment benefits are
highly differentiated by sector of industrial activity and firm size. Moreover, access to the
most generous programmes (Cassa Integrazione Guadagni Ordinaria and Straordinaria) is
often subject to government approval. This particular institutional setting, characterised by
a high degree of discretionality, necessarily leads to a poor imputation.
It is unclear, however, whether one would actually expect the points in Figure 5 to lie on
the 45-degree line. The imputed measure of benefits is more an indicator of entitlement than
receipt. It is a known fact that the degree of benefit take-up (i.e. the fraction of persons
entitled to a benefit who actually claim it) varies largely across countries and can be very low.
The recent survey of Hernanz et al. [2004] reports estimates of take-up of unemployment
benefits that range between 65% and 76% in the United States and between 62% and 71%
in the United Kingdom. Moreover, insurance-based programmes, like most unemployment
benefits, are typically the least affected by low take-ups. Social assistance schemes, like
some unemployment assistance programmes that are considered in this analysis, can reach
16
The programmes are written in Stata8.2 and are available from the author.
17
take-up rates as low as 37%. These estimates are only available for few countries (United
Kingdom, Germany, Netherlands and a few others) which are likely to be the most efficient
in the administration of social security programmes in Europe.
Moreover, even if imputed and reported benefits were actually measures of the same
underlying quantities, measurement error issues affect both of them and it is unclear which
one of the two is preferable. Survey respondents will tend to underreport the amounts they
receive from public transfers, typically because of social stigma, one of the most popular
explanations put forward for the observed low take-up rates. Moreover, the ECHP only
asks the annual amount received in ”unemployment related benefits”, thus, it is potentially
subject to recall bias and, possibly, misunderstanding about what should fall into this group
of transfers.
The imputation, on the other hand, does not take into account various elements of
the system for which no information is easily available. For example, in many countries
benefits might be reduced or withdrawn in case the recipient does not fulfill various job
search requirements. These sanctions are normally applied with some discretion and it is
extremely hard to identify those persons in the ECHP who might have been subject to them.
Another potential problem concerns the timing of the reforms. Changes in the unemployment
compensation system could in principle be applied to the newly unemployed only, i.e. those
who lost their jobs after the reform, or to all recipients. Information about these details
of the reforms is very difficult to obtain. For simplicity, the imputation procedures used
here assume that all changes always affect all recipients, regardless of when they entered
unemployment. Finally, in the ECHP employment histories are perfectly known since the
time the individual joined the survey but little is known about previous periods and some
assumptions need to be made. Specifically, it has been assumed that individuals have always
worked and paid contributions since the start of their first job, a piece of information available
from the data17 .
For the purpose of this paper, however, what matters most is that imputed benefits work
well as instruments for actual benefits. The results in Figure 5 indicate that imputed benefits
17
This assumption is particularly problematic for the United Kingdom, where contribution conditions are
particularly complex.
18
are generally strongly correlated with reported benefits, which, although imperfect, are the
best available measure of actual benefits. Moreover, conditional on individual characteristics,
the variation in imputed benefits across individuals is only due to differences, across countries
and over time, in the administrative rules of the benefit systems, a type of variation which
is hardly correlated with unobserved individual effects.
5.2
Measuring the earning potential of the unemployed
One of the crucial controls that needs to be included in the estimation is a measure of the
previous wage, as an indicator of the earning potential. In fact, the same benefit amount
affects differently people who can earn different wages. What really counts in determining
the incentives/disincentives to work is the actual difference between income in work and
income out of work. This is why in the empirical analysis that follows, the replacement rate
(the ratio between the unemployment benefit and the previous wage) will be used instead of
the level of the benefit itself.
However, in the ECHP unemployed workers do not report their previous wages. Only individuals who are working at the time of the interview are asked about their current monthly
wage. In the estimation, the most recent observed current wage from previous interviews has
been used as previous wage. Obviously there are many individuals who happen to be unemployed at all interviews, even if they report some employment spells between subsequent
interviews. For these individuals no previous wage is observed. Discarding them would reduce the sample size dramatically and, even more worrisome, it would introduce a potentially
large sample bias: the probability of having been unemployed at all interviews is obviously
higher for individuals at high risk of long and/or repeated unemployment. Alternatively,
one can use the average wage earned by individuals with similar characteristics. This is the
approach taken here: missing previous wages are replaced by the average wage of full-time
workers with the same level of education, age, experience, gender and region of residence18 .
18
This is computed by running a series of year-by-year, country-by-country OLS wage regressions including
education, age, experience and regional dummies and run separately for males and females.
19
6
The empirical analysis
Table 4 shows the descriptive statistics for the entire sample and for the treatment and the
control group separately. Encouragingly, the differences between the two groups are not
particularly marked. Education and income are significantly higher in the control group,
while age and experience are slightly lower. Interestingly, household composition (household
size and number of children) is not particularly different between the two groups, suggesting
that the selection into treatment and control is mostly due to cross-country differences in the
means-testing procedures. As expected, the fraction of families that report receiving family
assistance is higher in the treatment group while UI replacement rates and durations are
similar. Important differences emerge for unemployment assistance. In some countries, once
unemployment insurance expires, one can receive a non-insurance based benefit, conventionally named unemployment assistance, which is normally unlimited (although subject to job
search requirements). Most of the countries that offer unlimited unemployment assistance
happen to also have universal family benefits, hence, 72% of the persons in the control group
are potentially entitled to unlimited unemployment assistance versus only 16% among the
treated. Finally, the average duration of the unemployment spells is only marginally different, with the treated experiencing on average spells longer by 1.5 months. The differences
in the regional and country variables are more marked, but this is not surprising, being the
treatment and the control group selected mostly on the basis of the features of the countries’
welfare systems.
This brief inspection of the descriptive statistics already suggests that there should be
no systematic differences in the individual characteristics of the treated and the controls. In
order to clear any further doubt, Table 5 reports the results of a simple probit regression
where the dependent variable is equal to 1 if the individual is in the treatment group and zero
otherwise. Most of the individual characteristics turn out to be not significant, apart from
the number of children and the size of the household. A test of the joint significance of all
the personal and family variables does not reject the null hypothesis of the coefficients being
all zero. On the other hand, all the regional and country characteristics (apart from the
regional long term unemployment rate) are strongly significant. Once again this result is not
20
surprising. The crucial identification assumption here is that, after controlling for a large set
of controls of national institutional differences, cross-country variation in the administrative
rules of the welfare system can safely be used to identify causal effects.
Obviously, there still remain a possibility that the treated and the control might significantly differ according to some unobserved characteristics (culture, attitudes towards family
and work, etc.). However, since the variables of interest - UI replacement rate and duration
of UI entitlement - have been imputed only on the basis of the country regulations, endogeneity could only arise if these unobserved characteristics influence the design of the welfare
programmes considered here.
Let us now move on to the analysis of the exit rates from unemployment for the two
groups. Figures 6 and 7 show the empirical hazards19 for the entire sample and for the treatment and control group separately. The distribution of imputed duration of unemployment
benefit is also reported in the figures, scaled on the right hand side vertical axis. The figure
for the entire sample (Figure 6) shows the expected peak in the hazard around the time of UI
exhaustion. In fact, the almost monotonic declining trend in the hazard rate is interrupted
between the 12th and the 18th month, in correspondence with a series of important mass
point in the distribution of UI duration at 12, 15, 16 and 18 months. The later mass point
at 24 months, however, is not associated with any peak, but at this duration the size of the
sample is already dramatically reduced and the hazard rate is estimated rather imprecisely.
By looking at the same picture for the two groups (Figure 7) it is already evident that these
peaks are mostly due to the control group. The hazard rate for the treated, in fact, is a lot
flatter, especially between the 12th and the 18th month. Moreover, the control group shows
an earlier peak at the 6th month of unemployment, in correspondence with mass points in
the distribution of UI duration at 6 and 7 months. The hazard rate of the treated is a much
more fluctuating at the same duration.
Evidence from the empirical hazards, although already suggestive, is obviously not fully
convincing because the extent to which these graphs are influenced by personal characteristics
19
The empirical hazard at time t is computed as the ratio of individuals who actually exit unemployment
at time t over the number of all individuals who have been unemployed at least until t, i.e. all individuals
who could have exited at time t.
21
and duration dependence is not taken into account. Table 6 reports the results from a
proportional hazard model estimated for various samples. The econometric model is rather
standard, with one minor difficulty arising because in the ECHP unemployment durations
are recorded in months - i.e. in discrete intervals of time - whereas the underlying process
of job search occurs essentially in continuous time (workers can find a job at any moment
within a month). Following the custom in the literature, let us assume that the hazard rate ϑ(t|Xi , β) - of the underlying continuous process for individual i, i.e. the instant probability
that the spell ends at time t, can be written as the product of two parts: a baseline hazard
that depends on duration only, h0 (t), and a ”proportional shifter”, eXi β , that, according to
each individual’s characteristics Xi , shifts the baseline up or down:
ϑ(t|Xi , β) = h0 (t) · eXi β
(6)
The discrete time analog of ϑ(t|Xi , β) for spell i that ends between month Ti and Ti + 1, is
usually written as:
h(Ti | Xi , β) = Pr {Ti < t < Ti + 1 | t > Ti , Xi , β} =
¤
£
S(Ti |Xi , β) − S(Ti + 1| Xi , β)
= 1 − exp eXi β (Hi − Hi+1 )
=
S(Ti | Xi , β)
where Hi =
ZTi
(7)
h0 (u)du. It is useful to apply to equation (7) the following transformation:
0
log(− log [1 − h(Ti | Xi , β)]) = Xi β + τ i
(8)
Equation (8) allows to recover for (a transformation of) the discrete time hazard h(Ti | Xi , β)
the separability property of its continuos-time analog. In fact, this transformation of h(Ti | Xi , β),
just like ϑ(t|Xi , β), can now be separated into two parts: one, τ i = log [H(Ti ) − H(Ti + 1)] ,
that depends on the shape of the baseline hazard only, and another one, Xi β, which depends
only on individual’s characteristics (possibly time-varying).
According to equation (8), the discrete time hazard can then be rewritten as:
¢
¡
h(Ti | Xi , β) = 1 − exp −eXi β+τ i
22
(9)
Then, it is possible to express the likelihood contributions of completed and uncompleted
spells in terms of the discrete-time hazard and apply the transformation of equation (8) for
the estimation:
completed spells : Pr {Ti < t < Ti + 1 | Xi , β} = h(Ti | Xj , β)
uncompleted spells:
Pr {t > Ti | Xi , β} =
Ti
Y
k=0
TY
i −1
k=0
[1 − h(k | Xi , β)]
[1 − h(k | Xi , β)]
In our data, a spell can end either into employment or into inactivity. Assuming that
the probabilities of ending in any of these two states are independent, Narendrenathan and
Stewart [1993] showed that, by making the additional, harmless but greatly simplifying
assumption, that exits can only occur at the boundaries of the interval (i.e. either at the
beginning or at the end of each month), the correct hazard for exits into employment can
be estimated by considering as censored all those spells that end into inactivity. This is also
the approach taken here and the investigation of the determinants of exits into inactivity is
left for future research.
The set of controls included in the model contains all the relevant observable personal
characteristics, country and region specific controls and year dummies. Table 6, however,
only reports the estimated coefficients on the variables of main interest - the UI replacement
rate and the ”time to exhaustion” dummies - while the full set of results can be found in
Table A.1 in Appendix A. The estimates in Table 6 are obtained by imposing a third order
polynomial form for the baseline hazard, τ i = Ti + Ti2 + Ti3 .
The first column in Table 6 shows the estimates obtained for the large sample all UI
recipients, i.e. including males and females, all persons aged between 16 and 64 and all
households with and without children. The purpose of this first set of results is to show that,
using cross-country variation in the administrative rules of the unemployment insurance
system, it is possible to replicate the usual results in the literature. In fact, higher UI
replacement rates (UIρ) reduce the probability of finding a job, but only after the 6th
month of unemployment, while there is no statistically significant effect during the earlier
months. This non-monotonic effect of the generosity of unemployment benefit is an already
known result. Once UI expires, some of the individuals in the sample are still eligible for
23
unemployment assistance but its generosity (UAρ) does not have a significant impact on the
hazard rate.
To capture the presence of peaks in the hazard rate around the time of benefit exhaustion, a set of ”months to exhaustion” dummies has been included in the model for both
unemployment insurance and unemployment assistance. The coefficients on these dummies
indicate a significant increase in the hazard rate between 4 to 6 months from exhaustion of
UI and a similar but opposite effect for UA.
The following column in Table 6, repeats the same estimation for the sample of 2,703 UI
recipients who are male heads of households with children. Results are generally confirmed,
apart form the peak in the hazard at 4 to 6 months from UI exhaustion, which has now
disappeared. The most interesting part of Table 6, however, is the comparison between
column 3 and 4, where the estimation is repeated for the control and the treatment group
separately. The results are now dramatically different: while the negative impact of high UI
replacement rates is confirmed for the control group, this disappears for the treated, who
instead show significantly higher exit rates associated with more generous UI in the first 6
months of unemployment and no effect afterwards. Moreover, the estimates for the control
group show peaks in the hazard between 2 and 4 months to UI exhaustion while the same
coefficient is actually negative for the treated who, instead, show a jump in the probability
of finding a job around the time of UA exhaustion.
As indicated by the summary statistics in Table 4, there are actually few individuals who
are eligible for UA among the treated, so this last result could be due these few observations
all exiting around the time of UA exhaustion. The dip in the hazard between 4 to 6 months
from UI exhaustion for the treated could be explained by a sort of ”wait unemployment”
effect: if family benefits guarantee a similar income to UI for a longer period and without
the hassle of the job search requirements, there could actually be an incentive to wait for
the unemployment benefit to expire and receive more family benefits. The last column of
Table 6 reports the significance level of the difference between the estimated coefficients for
the treated and the control.
The results of Table 6 are obtained without controlling for unobserved heterogeneity. Although the identification strategy is robust to correlation between unobservable heterogeneity
24
and the variables of interest (UI replacement rate and exhaustion dummies), unobservable
effects could increase the variance of the model. This is likely to be particularly important
in the sample used here, where about 60% of the individuals experience more than one unemployment spell during the period of observation. As it is customary in this literature,
unobserved heterogeneity could be controlled for by making an arbitrary assumption about
the distribution of the unobservables. The likelihood of the model is then estimated by
integrating it out. In Table A.2 the model of Table 6 is estimated using this approach and
assuming normally distributed heterogeneity. The results are only marginally different from
those in the main text and, if anything, they are more precisely estimated.
7
Conclusions
This paper investigates how interactions between unemployment insurance and social assistance affect the job search behaviour of the unemployed. The theoretical framework
presented in Section 3 formalizes the idea that people are eventually interested in total payments (i.e. unemployment benefit and social assistance) and their time profile. Unemployed
workers will react differently to changes in the rules of the UI system depending on what
alternative or complementary welfare programmes are available.
These theoretical predictions are tested using monthly labour market histories of European unemployed persons and exploiting variation in a particular feature of family cash
benefits across Europe. While family benefits are universal in most countries, i.e. allocated
according only to the number and the age of children but regardless of income, in 5 countries
they are actually means tested. Comparing similar unemployed male heads of households
with children who are subject to different family assistance schemes, shows significant differences in the elasticity of the probability of finding employment to both the generosity and
the duration of the unemployment benefit. These results suggests that interactions between
welfare programmes are, indeed, important.
From a more general policy perspective, this paper highlights the need to design welfare
reforms with very careful consideration for the interactions between different programmes
in the system: reducing the level or the duration of unemployment benefit may not be very
25
effective in inducing unemployed workers to search harder if they can easily shift into other
social assistance programmes. This result is consistent with some recent studies that have
shown how wide and comprehensive reforms of labour market policies, even if politically
harder to implement, are often more effective than piecewise reforms.
References
[1] A.B., Gomulka J., Micklewright J., Rau N., (1984) ”Unemployment benefits, duration
and incentives in Britain: how robust is the evidence?”, Journal of Public Economics,
1984, vol. 23, 3-26.
[2] Atkinson A., Micklewright J., (1991) ”Unemployment compensation and labour market
transitions: a critical review”, Journal of Economic Literature, 1991, vol. 29, 1679-1727.
[3] Belot M., Van Ours J. C., (2000) ”Does the recent success of some OECD countries
in lowering their unemployment rates lie in the clever design of their labour market
reforms?”, Centre for Economic Policy Research Discussion Paper: 2429, June 2000.
[4] Bertola G., Jimeno J.F., Marimon R., Pissarides C., (2000) ”EU Welfare Systems and
Labour Markets: Diverse in the Past, Integrated in the Future?”, in Bertola G., Boeri
T., Nicoletti G., Welfare and Employment in a United Europe”, MIT Press.
[5] Blundell R., Fry V., Walker I., (1988) ”Modelling the take-up of means-tested benefits:
the case of housing benefits in the United Kingdom”, Economic Journal, vol. 98 (390
Supplement), 1988, 58-74.
[6] Boeri T., Steiner V., (1998) ””Wait unemployment” in economies in transition: the case
of Poland”, Konjunkturpolitik; 44(3), 1998, 287-311.
[7] Cahuc P., Lehemann E., (2000) ”Should unemployment benefit decrease with the unemployment spell?”, Journal of Public Economics, 2000, vol. 77, 135-153.
[8] CESIfo (2004), Ifo’s Database for Institutional Comparisons in Europe (DICE),
www.cesifo.de
26
[9] Coe D. T., Snower D. J., (1997) ”Policy Complementarities: The Case for Fundamental
Labour Market Reform”, Centre for Economic Policy Research Discussion Paper: 1585,
February 1997.
[10] Craig P., (1991) ”Costs and Benefits: a review of research on take-up of income related
benefits”, Journal of Social Policy, 1991, vol. 20(4), 537-565.
[11] Daveri F., Tabellini G., (1998) ”Unemployment, growth and taxation in industrial countries”, Labour Economics, 1998, vol. 5, 47-88.
[12] Dormont B., Fougere D., Prieto A., ”The effects of time profiles of unemployment
insurance benefits on reemployment rate”, mimeo.
[13] Elmeskov J., Martin J. P., Scarpetta S., (1998) ”Key lessons for labour market reforms:
evidence from OECD countries’ experiences”, Swedish Economic Policy Review, 1998,
vol. 5, 205-252.
[14] Esping-Andersen G., (1990) ”The Three Worlds of Welfare Capitalism”, Cambridge,
UK, Policy Press.
[15] Ferrera M., (1998) ”The Four Social Europes: Between Universalism and Selectivity”,
in Rhodes M., Meny Y., ”The Future of European Welfare: A New Social Contract?”,
Basingstoke, Macmillan.
[16] Fondazione Rodolfo Debenedetti (2004), fRDB Social Reforms Database, www.frdb.org
[17] Heckman J., Singer B., (1984) ”A method for minimizing the impact of distributional
assumptions in econometric models of duration data”, Econometrica, vol. 52, 1984,
271-320.
[18] Hernanz V., Malherbet F., Pellizzari M. (2004), ”Take-up of welfare benefits in OECD
countries: a review of the evidence”, OECD Social, Employment and Migration Working
Papers n. 17.
[19] Holmlund B., (1998) ”Unemployment insurance in theory and practice”, Scandinavian
Journal of Economics, 1998, vol. 100(1), 113-141.
27
[20] Katz L. F., Meyer B. D., (1990) ”Unemployment insurance, recall expectations and
unemployment outcomes”, Quarterly Journal of Economics, November 1990, vol. 105(4),
973-1002.
[21] Lancaster T., (1979) ”Econometric models for the duration of unemployment”, Econometrica, 1979, vol. 47, 939-956.
[22] Meyer B.D., (1990) ”Unemployment insurance and unemployment spells”, Econometrica, July 1990, vol. 58(4), 757-782.
[23] MISSOC, ”Social Protection in the Member States of the European Union”, 1994 to
1997 editions, European Commission, Directorate-General for Employment Industrial
Relations and Social Affairs, 1994.
[24] Moffitt R., (1983) ”An economic model of welfare stigma”, American Economic Review,
1983, vol. 75(5), 1023-1035.
[25] Moffitt R., (1985) ”Unemployment insurance and the distribution of unemployment
spells”, Journal of Econometrics, April 1985, vol. 28, 85-101.
[26] Mortensen D., (1977) ”Unemployment Insurance and Job Search Decisions”, Industrial
and Labor Relations Review, July 1977, vol. 30, 505-517
[27] Narendranathan W., Stewart M. B., (1993a) ”Modelling the probability of leaving unemployment: competing risks models with flexible base-line hazards”, Applied Statistics,
1993, vol. 42(1), 63-83.
[28] Narendranathan W., Stewart M. B., (1993b) ”How does the benefit effect vary as unemployment spells lengthen?”, Journal of Applied Econometrics, 1993, vol. 8(4), 361-381.
[29] Nickell S., (1979) ”Estimating the probability of leaving unemployment”, Econometrica,
September 1979, vol. 47(5), 1249-1266.
[30] Nickell S., Van Ours J. C., (2000) ”The Netherlands and the United Kingdom: A
European Unemployment Miracle?”, Economic Policy, April 2000, vol. 0(30), 135-175.
28
[31] OECD (1998), ”Benefit Systems and Work Incentives - 1998 Edition”, Paris.
[32] OECD (1999), ”Benefit System and Work Incentive - 1999 Edition”, Paris.
[33] OECD (2002), ”Benefit System and Work Incentive - 2002 Edition”, Paris.
[34] OECD (2004a), OECD Employment Outlook, Paris.
[35] OECD (2004b), Taxing Wages 2003/2004, Paris
[36] Orszag J. M., Snower D. J. (1999), ”Anatomy of Policy Complementarities”, IZA Discussion Paper No.41.
[37] Peracchi F. (2002), ”The European Community Household Panel: A review”, Empirical
Economics, vol. 27, 63-90.
29
Figures and Tables
1984
1984
1996
1996
2000
2000
2004
2004
1980
1984
1992
year
1996
Denmark
1988
1992
year
1996
Germany
1988
2000
Type of reform:
— UI benefit
– – – UI duration
2004
2004
20
2000
unemployment rate
15
10
1984
5
1980
0
11
unemployment rate
10
9
1992
year
France
1992
year
12
8
Belgium
1988
1988
unemployment rate
10
8
1980
1980
1984
1984
1988
1988
Finland
1992
year
Greece
1992
year
1996
1996
2000
2000
2004
2004
Direction of the reform:
thick: less generous (i.e. lower and/or shorter payments)
thin: more generous
7
12
Figure 1: Unemployment and social reforms in Europe
1980
1980
6
12
4
10
unemployment rate
8
6
4
2
unemployment rate
10
8
6
12
unemployment rate
10
8
6
1984
1988
Ireland
1992
year
1996
2000
2004
1984
1992
year
1996
Portugal
1988
2000
2004
1980
1980
1984
1984
1988
Italy
1992
year
1996
1996
2000
2000
Type of reform:
— UI benefit
– – – UI duration
1992
year
Sweden
1988
Figure 1: Unemployment and social reforms in Europe (continued...)
1980
1980
2004
2004
15
unemployment rate
10
5
0
12
1980
1980
1984
1984
1992
year
1996
Netherlands
1988
1992
year
1996
United Kingdom
1988
2000
2000
2004
2004
Direction of the reform:
thick: less generous (i.e. lower and/or shorter payments)
thin: more generous
4
6
unemployment rate
10
8
unemployment rate
11
10
9
12
8
10
2
unemployment rate
8
6
4
unemployment rate
15
10
20
5
9
8
unemployment rate
7
6
5
4
Figure 2: Time profiles of benefit, reservation wage, search time and hazard rate — Case 1
reservation
wage
benefit
wt*
bt
T
unemployment
duration
T
unemployment
duration
hazard
rate
search
time
qt
st
T
unemployment
duration
T
unemployment
duration
Figure 3: Time profiles of benefit, reservation wage, search time and hazard rate — Case 2
reservation
wage
benefit
bt
wt
T
T
unemployment
duration
search
time
unemployment
duration
hazard
rate
st*
w t*
T
unemployment
duration
T
unemployment
duration
46058
Austria
19606
Belgium
24877
Finland
95612
France
84908
Germany
Ireland
14847
50496
Italy
Portugal
9023
Figure 4: Distribution of spells by country
15460
Denmark
282122
SpainUnited Kingdom
42879
20
16
8
12
Average un. rate
4
0
300000
200000
100000
0
Frequency
Figure 5: Comparison between imputed and reported annual unemployment benefit (national currencies)
Figure 6: Empirical Hazard for unemployment-to-job transitions
30
0.08
0.07
25
20
hazard
0.05
15
0.04
0.03
10
frequency of UI duration (%)
0.06
0.02
5
0.01
0
0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
32
34
36
month of unemployment
Males heads of households with children
[95% Conf. Int.]
Figure 7: Empirical Hazard for unemployment-to-job transtions
by groups
0.1
30
0.09
25
0.08
20
hazard
0.06
0.05
15
0.04
10
0.03
0.02
5
0.01
0
0
2
4
6
8
10
12
14
16
18
20
22
24
26
month of unemployment
CONTROL GROUP
TREATMENT GROUP
28
30
32
34
36
frequency of UI duration (%)
0.07
Table 1: The effect of UI reforms on the probability of leaving unemployment
Country
Reform
Austria
reduced duration - 2000
Belgium
reduced benefit — 1999
reduced benefit — 1997
Finland
reduced benefit — 2000
Germany
reduced duration — 1995
Greece
reduced benefit - 1996
Ireland
reduced benefit — 1995
United Kingdom
reduced duration and
benefit — 1995
effect1
-0.459*
(0.252)
-0.292
(0.361)
0.156
(0.142)
0.258
(0.253)
-0.031
(0.169)
0.035
(0.184)
0.486*
(0.279)
0.291
(0.334)
1. Coefficient of a dummy for unemployment spells that started after the implementation of the reform.
The estimates are obtained from a proportional hazard model with a fully non-parametric baseline and the
following explanatory variables: sex, age (linear and squared), health status, education, marital status, family
size, dummy for children in the household, household income, regional unemployment rate and rate of long
term unemployment, year dummies.
Source: ECHP 1994-2001
Table 2: Welfare reforms in Europe
COUNTRY
Austria
REFORMS
•
•
•
•
•
•
•
Belgium
•
•
•
•
•
•
Denmark
•
•
•
Finland
•
•
•
•
•
•
•
France
Germany
Greece
Ireland
•
•
•
•
•
•
•
•
•
•
•
•
•
•
1995: Unemployment benefits are cut (particularly for those with high incomes)
1999-2000: Family Benefits are made more generous
2000: Unemployment benefits’ duration is increased for persons with at least 15 years of
contributions.
2002: Unemployment benefits made more generous.
2002: Child-care support made more generous.
1988: Replacement rates for unemployment benefits are reduced
1992: Reform of unemployment benefit: access for those on temporary or part-time jobs;
redefinition of “suitable offer”; new rules for early retirement.
1998: Improved incentive for those on Income Support (Minimex) to take up jobs.
1999: Unemployment benefits made more generous for single persons
2001: Introduction of tax deductibility for child-care costs.
2001: Social assistance made more generous for old people.
1987: Unemployment benefits are increased.
1994: Labour Market Reform: unemployment benefits’ duration is reduced and eligibility
conditions are tightened.
1996: Follow-up of Labour Market Reform: unemployment benefits’ duration is shortened
and eligibility conditions are tightened.
1993: Reform of (flat-rate) Unemployment Assistance: duration is limited and eligibility
requirements brought in line with those for the earnings-related supplementary benefit.
1995: Reform of the unemployment benefits system: stricter conditions for the unemployed
to re-qualify for unemployment benefits, shortened duration for older workers
1997: Reform of the unemployment benefits system: stricter access conditions and lower
payments.
1998: Unemployment benefits are increased.
1998: Stricter access conditions for minimum income benefits.
2000: Unemployment benefits are decreased.
1992: Reform of unemployment benefits that introduces a downward sliding scale for
payments.
1993: Unemployment benefits’ duration is reduced.
1993: Increased generosity of general social assistance: housing benefits, family benefits,
employment accidents and occupational illness benefits are increased and made easier to
access.
1997: Minimum unemployment benefit is increased and duration of benefits is also
extended.
1998: Minimum income benefit increased.
1999: Unemployment benefits increased.
2000: Generalised increases in both unemployment and minimum income benefits.
2001: Generalised increases in both unemployment and minimum income benefits.
1994: Unemployment benefits are reduced.
1995: Unemployment benefit duration is reduced.
1998: Several changes: stricter rules for access to the minimum income scheme (RMI); lower
payments for sickness benefits; better incentives for unemployment benefit recipients to
take up jobs (redefinition of suitable offer; incentive to take part-time jobs, et.).
1999: Increased sickness benefits.
1990: Unemployment benefits’ duration is increased.
1996: Unemployment benefits are reduced.
1987: Unemployment benefits are reduced.
1993: Labour Market Reform: eligibility for unemployment benefits is made stricter; child
benefits are increased, family benefits are increased; income support is increased.
1995: Unemployment benefits are reduced.
•
•
1988: Ordinary unemployment benefits are increased.
1991: New unemployment benefit scheme for long-term unemployment is introduced (only
for certain categories)
Italy
• 1994: Unemployment benefits are increased and coverage is extended.
• 1998: Minimum income scheme introduced on an experimental basis.
• 1999: introduced new means-tested family cash benefits.
• 1987: Revision of the Social Security System Act: duration of unemployment benefits is
reduced; stricter rules for invalidity benefits/pensions; conditions for accessing all benefits
are tightened.
• 1991: Social Insurance Organisational Act: administration of all benefits delegated to a
single governmental body.
Netherlands
• 1996: General Social Assistance Act: improved incentives for those on income support to
take up jobs; privatisation of sickness benefits.
• 1997: A new housing benefit is introduced.
• 2001: Employment bonus paid to people on unemployment benefit who find job.
• 1997: Reform of family benefits (completely different structure, cannot say if more/less
generous overall); easier access to employment injuries/occupational disease benefits; a new
minimum income scheme is introduced.
Portugal
• 1998: Introduction of a “partial” unemployment benefit for part-timers.
• 1998: Longer contribution record required for eligibility of unemployment benefits.
• 1999: Extended duration of unemployment benefits.
• 1992: Reform of the unemployment benefits (reduced duration and payments).
• 1993: Unemployment benefits are reduced.
Spain
• 1994: Unemployment benefits are subject to taxation.
• 2000: Extended unemployment benefits for older workers with children.
• 1986: Participation in training programmes is considered equivalent to work for the purpose
of eligibility for unemployment benefits
• 1987: Subsidised jobs are offered to those whose unemployment benefit expires (and the job
must, by law, last at least enough to make the worker eligible for unemployment benefits
again).
• 1993: Unemployment benefits are reduced.
Sweden
• 1997: Reform of unemployment benefits: benefits are reduced and re-qualification through
subsidised jobs no longer available.
• 1998: Increased sickness benefits.
• 1999: Housing benefits made stricter and lower.
• 2002: Child-care made more generous.
• 1988: Unemployment benefits are reduced.
• 1989: Unemployment benefits are reduced.
• 1996: Job Seekers Allowance (JSA) replaces the old unemployment benefit: both payments
and duration of benefits are reduced, income support is replaced by a means-tested
component of the JSA.
United Kingdom
• 1997: Introduction of an employment bonus for older workers.
• 1998: Welfare-to-Work Programme: training for long-term unemployed, hiring subsidies for
employers, sanctions for refusing job offers, etc.
• 1999: Introduction of new means-tested child-care tax credit.
Sources: European Commission Missoc (1992-2001), Missoc-info (1985-2001); Fondazione RDB “Social Policy Reforms
Database”; CESifo “DICE Database”.
Table 3. Family Cash Benefits in Europe and Composition of the sample
Country
Income Test
Age Limit
Austria
No
19 to 26
All male heads of
households with kids
291
Belgium
No
18 to 25
Denmark
No
Finland
1
Control
Treatment
291
(100.0)
0
(0.0)
111
111
(100.0)
0
(0.0)
18
149
149
(100.0)
0
(0.0)
No
16
275
275
(100.0)
0
(0.0)
France
No
18 to 20
187
187
(100.0)
0
(0.0)
Germany
Yes
18
199
36
(18.1)
163
(81.9)
Greece
Yes
18 to 22
-
-
-
Ireland
No
16 to 19
277
277
(100.0)
0
(0.0)
Italy
Yes
18
149
35
(23.5)
114
(76.5)
Luxembourg
No
18 to 27
-
-
-
Netherlands
No
17 to 24
-
-
-
Portugal
Yes
15 to 24
80
5
(6.3)
75
(93.8)
Spain
Yes
18
890
200
(22.5)
690
(77.5)
Sweden
No
16
-
-
-
United Kingdom
No
16 to 19
95
95
(100.0)
0
(0.0)
2703
1661
(61.5)
1042
(38.5)
Total
1. Benefits are paid until the child reaches this age limit, which is extended for children in training or higher
education. Source: European Commission Missoc
Table 4: Descriptive Statistics
All male heads of
households with kids
Mean Std. Dev.
Variable
Individual and spells:
Age
Experience1
1=primary education
1=secondary education
1=tertiary education
Household size
Number of chilren (<15yo)
Log Household income (PPP) 2
Control group
Treatment group
Mean
Std. Dev.
Mean Std. Dev.
37.22
(7.41)
36.33
(7.68)
38.66 (6.70)
20.08
0.49
0.37
0.14
4.35
1.31
(8.93)
(1.32)
(1.08)
18.67
0.38
0.45
0.17
4.34
1.19
(8.94)
(1.41)
(1.18)
22.30
0.66
0.25
0.09
4.37
1.51
8.15
(3.02)
9.02
(1.86)
6.76 (3.88)
0.21
-
0.20
-
0.24 -
0.64
(0.23)
0.64
(0.25)
0.64 (0.21)
20.89
(18.18)
20.40
(18.63)
21.66 (17.42)
0.50
-
0.72
-
0.16 -
Duration of all unemployment spells
Duration of completed unemployment spells
11.13
10.61
(11.53)
(9.65)
10.67
10.05
(11.73)
(8.69)
11.85 (11.17)
11.43 (10.88)
Fraction of completed spells
ending into employment
ending into inactivity
81.98
0.95
0.05
-
79.77
0.94
0.06
1=Family cash benefit recipients
3
UI replacement rate
Maximum entitlement of UI (in months)
1=unlimited UA
4
4
Number of spells
2703
Countries and regions:
5
Regional unemployment rate
Regional long-term unemployment rate
Expenditure in ALMP (%GDP)
Share of part-time employment
6
6
Share of temporary employment
Share of public employment
6
Minimum wage / median wage
Average effective tax rate
Uniuon density
7
6
Employment protection index
6
6
6
5
(8.46)
(1.17)
(0.87)
85.51 0.98 0.02 1661
1042
14.09
(7.49)
11.15
(5.76)
18.77 (7.54)
0.39
(0.14)
0.42
(0.15)
0.36 (0.10)
0.95
(0.47)
1.10
(0.50)
0.72 (0.30)
11.86
(4.58)
13.49
(4.44)
9.27 (3.49)
18.19
(11.13)
13.41
(8.35)
25.82 (10.77)
0.21
(0.06)
0.22
(0.06)
0.18 (0.03)
0.19
(0.21)
0.15
(0.23)
0.24 (0.15)
38.00
(5.73)
38.34
(6.42)
37.46 (4.37)
34.88
(22.22)
44.02
(23.53)
20.30 (7.27)
2.45
(0.81)
2.09
(0.82)
3.03 (0.32)
1. Years since first entrance in the labour market
2. Income of all other household members
3. Monthly UI amount / previous wage. UI amounts imputed on the basis of country regulations and personal characteristics
(MISSOC, 1993-2001).
4. Imputed on the basis of country regulations and personal characteristics (MISSOC, 1994-2001).
5. Source: ECHP
6. Source OECD
7. Income taxes plus social security contributions (employer and employee) over total gross labour costs. Source: OECD
Table 5: Selection into treatment and control group
Dependent variable=1 if treated
Personal and family characteristics:
Age
Experience1
1=secondary education
1=tertiary education
Household size
Number of kids (<15yo)
(log) hh income2
1= zero hh income
Country/Region characteristics:
Regional unemployment rate3
Regional long-term unemployment rate3
Expenditure in ALMP (%GDP)4
Share of part-time employment4
Share of temporary employment4
Share of public employment4
Minimum wage (/median wage)4
Average tax rate5
Union density4
Employment protection index4
Year dummies:
1995
1996
1997
1998
1999
2000
2001
Constant
Observations
Log-likelihood
UI recipients, male heads of households with kids
0.013
(0.024)
0.025
(0.020)
-0.206
(0.149)
-0.063
(0.191)
-0.142*
(0.064)
1.774**
(0.129)
-0.054
(0.059)
0.616
(0.600)
0.028*
(0.012)
0.925
(0.779)
1.266*
(0.544)
0.268*
(0.117)
0.143**
(0.038)
36.324**
(8.342)
-30.114**
(3.895)
-0.329**
(0.052)
-0.319**
(0.069)
6.434**
(0.577)
-0.666**
(0.237)
-0.751**
(0.282)
0.045
(0.259)
0.400
(0.316)
-0.636
(0.375)
-0.494
(0.387)
-0.814
(0.445)
-7.706
(4.006)
2703
-319.58
1. Years since first entrance in the labour market
2. Income of all other household members.
3. Source: ECHP
4. Source: OECD
5. Income taxes plus social security contributions (employer and employee) over total gross labour costs. Source: OECD
Robust standard errors (clustered by individual and year) in parentheses
* significant at 5%; ** significant at 1%
(2>time<=6)
106268
16813
-36832.70
yes
yes
yes
(time, time2, time3)
-0.112
(0.250)
-0.159
(0.225)
-1.235**
(0.275)
-0.235
(0.136)
0.072
(0.146)
0.302**
(0.121)
-0.035
(0.111)
0.066
(0.065)
-0.468**
(0.132)
-0.014
(0.156)
[1]
24080
2703
-9146.90
yes
yes
yes
(time, time2, time3)
0.892
(0.601)
-0.745
(0.500)
-1.049*
(0.479)
-0.533
(0.434)
0.256
(0.258)
0.308
(0.256)
0.318
(0.322)
0.150
(0.294)
-0.987*
(0.458)
-0.409
(0.469)
UI recipients, male heads
of households with kids
[2]
13757
1661
-4465.21
yes
yes
yes
(time, time2, time3)
-0.476
(0.511)
-0.851
(0.754)
-0.978
(0.643)
-0.090
(0.337)
0.597*
(0.293)
0.475
(0.305)
0.204
(0.380)
-0.248
(0.336)
-1.368*
(0.559)
-0.924
(1.004)
[3]
Control
10323
1042
-4523.81
yes
yes
yes
(time, time2, time3)
2.554**
(0.883)
1.236
(0.773)
-0.624
(0.726)
-1.417
(0.950)
-1.406*
(0.671)
0.050
(0.420)
2.222*
(0.887)
1.950*
(0.778)
0.613
(0.833)
0.843
(0.707)
[4]
Treatment
All UI recipients
Table 6: Proportional Hazard Model for unemployment duration — Exits into Jobs
UIρ ⋅
(time>6)
Replacement rates:
UIρ1 ⋅ (time<=2)
UIρ ⋅
UAρ1
“Months to UI exhaustion” dummies:
1=less than 2 months
1=2 to 4 months
1=4 to 6 months
“Months to UA exhaustion” dummies:
1=less than 2 months
1=2 to 4 months
1=4 to 6 months
Personal and family characteristics2
Country/region characteristics3
Year dummies
Baseline hazard
Observations
Subjects
Log-likelihood
Difference
[3]-[4]
**
**
*
**
**
1. Monthly UI or UA amount / previous wage. UI and UA amounts imputed on the basis of country regulations and personal characteristics (MISSOC, 1993-2001).
2. Age, experience, education, health status, house ownership, household size, number of kids (<15yo), (log) household income. See table A1
3. Regional unemployment rate, regional long-term unemployment rate, expenditure in ALMP (%GDP), share of part-time employment, share of temporary employment, share of
public employment, minimum wage (/median wage), average tax rate. See table A1.
Robust standard errors (clustered by individual and year) in parentheses
* significant at 5%; ** significant at 1%
Appendix A:
Additional estimation results
Table A1: Proportional Hazard Model for unemployment duration — Exits into Jobs
All UI recipients
[1]
Replacement rates:
UIρ1 ⋅ (time<=2)
UI recipients, male heads
of households with kids
[2]
Control
Treatment
[3]
[4]
-0.035
(0.111)
0.066
UIρ ⋅ (2>time<=6)
(0.065)
-0.468**
UIρ ⋅ (time>6)
(0.132)
-0.014
UAρ1
(0.156)
“Months to UI exhaustion” dummies:
1=less than 2 months
-0.235
(0.136)
1=2 to 4 months
0.072
(0.146)
1=4 to 6 months
0.302*
(0.121)
“Months to UA exhaustion” dummies:
0.318
(0.322)
0.150
(0.294)
-0.987*
(0.458)
-0.409
(0.469)
0.204
(0.380)
-0.248
(0.336)
-1.368*
(0.559)
-0.924
(1.004)
2.222*
(0.887)
1.950*
(0.778)
0.613
(0.833)
0.843
(0.707)
-0.533
(0.434)
0.256
(0.258)
0.308
(0.256)
-0.090
(0.337)
0.597*
(0.293)
0.475
(0.305)
-1.417
(0.950)
-1.406*
(0.671)
0.050
(0.420)
1=less than 2 months
0.892
(0.601)
-0.745
(0.500)
-1.049*
(0.479)
-0.476
(0.511)
-0.851
(0.754)
-0.978
(0.643)
2.554**
(0.883)
1.236
(0.773)
-0.624
(0.726)
-
-
-
0.014
(0.015)
-0.019
(0.012)
-0.066
(0.113)
-0.197
(0.177)
-
0.015
(0.018)
-0.020
(0.015)
-0.057
(0.140)
-0.275
(0.203)
-
-0.008
(0.028)
-0.005
(0.022)
-0.091
(0.183)
-0.097
(0.293)
-
-
-
-
-0.016
(0.042)
0.100
(0.063)
-0.036
(0.052)
-0.169
(0.464)
-0.093
(0.062)
0.206*
(0.091)
0.047
(0.084)
0.538
(0.795)
0.100
(0.065)
-0.071
(0.088)
-0.139*
(0.063)
-0.861
(0.548)
-0.022*
(0.009)
0.072
(0.397)
-0.219
(0.183)
-0.480
(0.037)
-0.005
(0.593)
0.364
(1.846)
0.006
(0.014)
0.261
(0.416)
-0.116
(0.241)
-0.014
(0.038)
-0.606
(0.018)
2.441
(2.172)
-0.034**
(0.011)
-0.663
(0.933)
0.632
(1.837)
-0.037
(0.318)
-0.046
(16.338)
20.758
(39.683)
-0.112
(0.250)
1=2 to 4 months
-0.159
(0.225)
1=4 to 6 months
-1.235**
(0.275)
Personal and Family characteristics:
1=female
-0.290**
(0.054)
Age
-0.002
(0.008)
Experience2
-0.002
(0.007)
1=secondary education
0.074
(0.061)
1=tertiary education
0.286**
(0.081)
1= bad health
-0.391**
(0.126)
1=house owner
0.245**
(0.057)
Household size
0.083**
(0.032)
Number of kids (<15yo)
-0.012
(0.039)
(log) hh income3
-0.067*
(0.033)
1= zero hh income
-0.667*
(0.298)
Country/Region characteristics:
Regional unemployment
-0.011*
rate4
(0.005)
Regional long-term
0.201
(0.236)
unemployment rate4
Expenditure in ALMP
-0.175
(%GDP)5
(0.101)
Share of part-time
-0.054**
employment5
(0.019)
Share of temporary
-0.014*
employment5
(0.007)
Share of public
3.468**
employment5
(0.833)
Minimum wage (/median
wage)5
Average tax rate6
Union density5
Employment protection
index5
Year dummies
Baseline hazard
Observations
Subjects
Log-likelihood
-0.996**
(0.214)
0.022**
(0.008)
-0.014**
(0.003)
-0.158
(0.102)
-0.072
(0.014)
0.039*
(0.022)
-0.011
(0.007)
-0.452
(0.239)
-0.088*
(0.762)
0.005
(0.032)
-0.012*
(0.007)
-0.256
(0.336)
4.824
(0.140)
0.232
(0.171)
-0.054
(0.339)
-2.236*
(0.879)
yes
(time, time2, time3)
yes
(time, time2, time3)
yes
(time, time2, time3)
yes
(time, time2, time3)
106268
16813
-36832.70
24080
2703
-9146.90
13757
1661
-4465.21
10323
1042
-4523.81
1. Monthly UI or UA amount / previous wage. UI and UA amounts imputed on the basis of country regulations and personal
characteristics (MISSOC, 1993-2001).
2. Years since first entrance in the labour market
3. Income of all other household members.
4. Source: ECHP
5. Source: OECD
6. Income taxes plus social security contributions (employer and employee) over total gross labour costs. Source: OECD
Robust standard errors (clustered by individual and year) in parentheses
* significant at 5%; ** significant at 1%
Table A.2: Proportional Hazard Model for unemployment duration with unobserved heterogeneity— Exits
into Jobs
All UI recipients
[1]
Replacement rates:
UIρ1 ⋅ (time<=2)
-0.288*
(0.121)
0.378**
UIρ ⋅ (2>time<=6)
(0.108)
-0.384**
UIρ ⋅ (time>6)
(0.109)
0.044
UAρ1
(0.000)
“Months to UI exhaustion” dummies:
1=less than 2 months
-0.003
(0.091)
1=2 to 4 months
0.272**
(0.070)
1=4 to 6 months
0.110
(0.061)
“Months to UA exhaustion” dummies:
1=less than 2 months
0.205
(0.156)
1=2 to 4 months
-0.590**
(0.187)
1=4 to 6 months
-0.751**
(0.197)
Personal and Family characteristics:
1=female
-0.290**
(0.054)
Age
-0.032**
(0.006)
Experience2
0.005
(0.005)
1=secondary education
0.244**
(0.050)
1=tertiary education
0.245**
(0.076)
1= bad health
-0.367**
(0.090)
1=house owner
0.341**
(0.052)
Household size
-0.071**
(0.017)
Number of kids (<15yo)
0.163**
(0.023)
(log) hh income3
-0.104**
(0.021)
1= zero hh income
-0.811**
(0.173)
Country/Region characteristics:
Regional unemployment
-0.035**
rate4
(0.005)
Regional long-term
-0.904**
unemployment rate4
(0.171)
Expenditure in ALMP
-0.245**
(%GDP)5
(0.080)
Share of part-time
0.020
employment5
(0.014)
Share of temporary
0.003
employment5
(0.006)
Share of public
-1.277
UI recipients, male heads
of households with kids
[2]
Control
Treatment
[3]
[4]
0.193
(0.189)
0.351*
(0.169)
-0.678**
(0.193)
-0.074
(0.236)
0.564**
(0.208)
0.286
(0.186)
-0.954**
(0.222)
-0.229
(0.359)
0.713
(0.583)
1.376*
(0.561)
0.734
(0.576)
0.831
(0.472)
-0.010
(0.154)
0.332**
(0.120)
0.437**
(0.100)
0.618**
(0.164)
1.082**
(0.125)
0.479**
(0.104)
-1.205**
(0.434)
-1.027*
(0.420)
0.240
(0.214)
0.079
(0.246)
-0.915**
(0.256)
-1.026**
(0.250)
-1.935**
(0.336)
-2.276**
(0.316)
-1.199**
(0.329)
1.503**
(0.452)
0.487
(0.486)
-0.940*
(0.395)
-
-
-
-0.053**
(0.009)
0.064**
(0.009)
0.261*
(0.107)
0.831**
(0.106)
-
-0.012
(0.011)
-0.020*
(0.008)
-0.596**
(0.088)
-0.910**
(0.119)
-
-0.090**
(0.016)
0.051**
(0.013)
0.078
(0.000)
1.859**
(0.169)
-
-
-
-
-0.201**
(0.025)
0.124**
(0.029)
-0.125**
(0.032)
-1.067**
(0.268)
0.138**
(0.040)
0.013
(0.039)
-0.270**
(0.041)
-1.418**
(0.394)
0.148**
(0.035)
-0.200**
(0.053)
-0.154**
(0.043)
-0.898*
(0.354)
-0.030**
(0.007)
-1.587**
(0.317)
-1.060**
(0.124)
-0.061**
(0.000)
0.218
(0.007)
-2.591**
-0.034**
(0.010)
-0.368
(0.339)
-1.116**
(0.143)
0.032
(0.010)
0.396
(0.325)
5.604**
-0.070**
(0.009)
-4.154**
(0.763)
-1.993
(1.031)
-0.199*
(0.095)
-0.346
(0.196)
1.269
employment5
Minimum wage (/median
wage)5
Average tax rate6
Union density5
Employment protection
index5
Year dummies
Baseline hazard
Distrbution of
heterogeneity (ν)
variance of ν
Observations
Subjects
Log-likelihood
(0.914)
-0.092
(0.270)
0.026**
(0.007)
0.004
(0.004)
0.028
(0.076)
(0.929)
-0.031**
(0.020)
0.055**
(0.013)
0.004
(0.004)
-0.595**
(0.133)
(1.422)
0.016
(0.025)
-0.022
(0.015)
0.020**
(0.005)
0.532**
(0.178)
(23.688)
-10.536
(9.861)
0.012
(0.095)
-0.267
(0.223)
-2.627**
(0.523)
Yes
(time, time2, time3)
Yes
(time, time2, time3)
Yes
(time, time2, time3)
Yes
(time, time2, time3)
normal
normal
normal
Normal
1.689**
(0.029)
1.935**
(0.039)
2.067**
(0.051)
2.115**
(0.077)
106268
16813
-18275.38
24080
2703
-8419.14
13757
1661
-4826.16
10323
1042
-3401.37
1. Monthly UI or UA amount / previous wage. UI and UA amounts imputed on the basis of country regulations and personal
characteristics (MISSOC, 1993-2001).
2. Years since first entrance in the labour market
3. Income of all other household members.
4. Source: ECHP
5. Source: OECD
6. Income taxes plus social security contributions (employer and employee) over total gross labour costs. Source: OECD
Robust standard errors (clustered by individual and year) in parentheses
* significant at 10%; ** significant at 5%; *** significant at 1%
Appendix B:
Descriptive tables of Welfare programmes
in European countries
Existing
schemes
Qualifying period1
Duration of payment2
TABLE B1. UNEMPLOYMENT BENEFIT IN EUROPEAN COUNTRIES (situation on January 2001)
Country
Rate3
Insurance
Variable according to age
No limit
60% to 55% initially declining to 44%Belgium
35% as unemployment continues
Insurance
52 weeks in the preceding 3 years
1+3 years
90% of reference earnings
Denmark
(optional)
Insurance
12 months in the preceding 3 years
Depending on age and contribution
Insurance: 60%-67% of net earnings
Germany
Assistance
history (between 12 and 64 months)
Assistance: 57%-53% of net earnings
Depending on contribution history
40%-50% of earnings
Insurance
125 days of work during the 14
Greece
(between 5 and 12 months)
months preceding job loss or, at least,
200 days of work during the 2 years
preceding job loss.
Insurance
12 months in the previous 6 years
Depending on contribution history
Insurance: 70%-60% of reference
Spain
Assistance
earnings
Assistance: 75% of minimum wage
Insurance
least 4 months insurance in last 18
Depending on age and contribution
Insurance: 40.4%-57.4% of earnings,
France
Assistance
months.
history (between 4 and 60 months)
declining.
Assistance: lump sum
Insurance
39 weeks’ contributions paid
390 days
Insurance: 98€ per week
Ireland
Assistance
Assistance: 97-98€ per week
Insurance
Varies according to the industry
Depending on the industry (180 days
30%-80% of earnings
Italy
or 90 days or 36 months)
Insurance
26 weeks of employment during the
365+182 days
80% of earnings
Luxembourg
last year
Insurance: 70% of earnings
Insurance
26 weeks of employment during the
6 months+ 9 months to 5 years
Netherlands
Assistance: 70% of minimum wage
Assistance
last 39 months
depending on age and employment
history.
Insurance
52 weeks during the last 24 months
Depending on age and contribution
Insurance: 55% of earnings + lump
Austria
Assistance
history (between 20 to 78 weeks)
sum
Assistance: 92-95% of unemployment
insurance
Insurance
540 days during the last 24 months
Depending on age (between 12 and 30
Insurance: 65% of earnings
Portugal
Assistance
months)
Assistance: 70%-100% of minimum
wage
Insurance
43 weeks of employment during the
500 days
Insurance: lump sum (21€ per day) +
Finland
Assistance
last 24 months
earning related supplement
Assistance: lump sum (21€ per day)
Insurance
6 months of employment
Depending on age (between 300 and
Insurance: 80% of earnings
Sweden
450 days)
182 days
Insurance
Contributions paid in one of the 2 tax
Insurance: lump sum (65-83€ per
United Kingdom
Assistance
years on which the claim is based
week)
amounting to at least 25 times the
Assistance: lump sum (99-130€ per
minimum contribution for that year
week for a couple)
(1)
Qualifying periods refer to working periods during which contributions are paid unless otherwise specified.
(2)
When duration is expressed in days these usually refer to a 5 or 6 day weekly payments. Duration usually refers only to insurance schemes,
unemployment assistance is typically unlimited.
(3)
When different rates are specified for the same scheme, the actual rate depends on family characteristics, age, contribution history or duration of
unemployment. Only basic rates are reported, excluding any supplement (family, old age, etc.)
Source: European Commission Missoc 2001.
Spain
Greece
Germany
Denmark
Belgium
Invalidity pension
Invalidity pension
Invalidity pension
Invalidity benefit
Invalidity pension
Invalidity benefit
Type of benefit
Eligible only after 12 months
of sickness benefit
66.66%
33%
50%
50%
50%
66.66%
minimum level of incapacity1
5 years of contributions
260 contributions’ weeks
12 months of work with sufficient
contributions paid
Worked 1/2 of the time between age 20 and
the claim
Between 5 and 15 years of work with 300 to
1500 contributions’ days (depending on age)
60 months of work with 36 months of
contributions in the previous 5 years
Until retirement
Until retirement
Until retirement
Until retirement
Until retirement
Until retirement
Until retirement
Depends on incapacity
Depends on contribution records
Depends on income and
contribution records
Depends on age (103-128€ per
week)
30%-50% of annual earnings
50%-100% of reference earnings
Depends on incapacity
Amount of benefit2
France
Invalidity pension
•66%
•100%
12 months in the previous 3 years
Until retirement
Depends on income
Duration
Ireland
•Invalidity pension
•Incapacity benefit
Eligible if invalidity prevents
the beneficiary form doing his
last job (or a similar one)
None
Until retirement
Depends on contribution records
Qualifying conditions
Italy
Invalidity pension
25%
60 contributions’ months in the previous 120
months
Until retirement
Depends on age
Country
Luxembourg
Invalidity pension
50%
5 years
Until retirement
25%-100% of basic pension
TABLE B2. INVALIDITY BENEFITS IN EUROPEAN COUNTRIES (situation on January 2001)
Netherlands
Invalidity pension
66.66%
3 years of residence
Until retirement
40-65% of last earnings
Austria
Invalidity pension
40%
Residence in Sweden
Until retirement
Portugal
Invalidity pension
25%
Until retirement
Finland
Invalidity pension
Lump sum
Depends on age.
6 months of contributions with 120 days
worked
3 years of residence
Sweden
•364 days
•Until retirement
Depends on incapacity (from 6.855€
to 18.012€ per year)
Depends on incapacity
United Kingdom
100%
•Short-term invalidity
•Enough contributions paid
•Long-term invalidity
•Having exhausted short-term invalidity
benefit
(1) Reduced capacity of earning or work unless otherwise specified.
(2) When a range is specified this usually varies with age and contribution records.
Source: European Commission Missoc 2001.
Portugal
Austria
Netherlands
Luxembourg
Italy
Ireland
France
Spain
Greece
Germany
Denmark
Belgium
No
Yes
No
No
No
Yes
No
No
Yes
Yes
Yes
No
No
Income Test
16
16
15 to 24
19 to 26
17 to 24
18 to 27
18
16 to 19
18 to 20
18
18 to 22
18
18
18 to 25
Age Limit1
Depends on the number of children
Depends on the number of children
Depends on the number of children
Depends on the number of children
Depends on the number of children
Depends on the number and age of the children
Depends on the number and age of the children
Depends on family and number of children
Depends on the number and age of the children
Depends on the number of children
Depends on the age of the child
Depends on the number of children
Depends on the number of children
Depends on the age of the child
Depends on the age of the child
Calculation of benefit
Yes
Yes
Yes
No
No
No
No
Yes
Yes
Yes
No
Only for widows
No
Yes
No
Supplement for single parents
Finland
No
16 to 19
Country
Sweden
No
TABLE B3. FAMILY CASH BENEFITS IN EUROPEAN COUNTRIES (situation on January 2001)
United Kingdom
(1) Benefits are paid until the child reaches this age limit, which is extended for children in training or higher education.
Source: European Commission Missoc 2001.
Luxembourg
Italy
Ireland
France
Spain
Greece
Denmark
Belgium
Housing benefit
Housing Supplement for those under RMI
No direct benefit but social housing is available
Housing Supplement for those under Social
Welfare Allowance
Housing benefit
No general housing benefit but some regions have
introduced one.
No direct benefit but a tax allowance for house
rents is available.
• General housing benefit (open to everybody)
• Special housing benefit (for those receiving
income support)
Housing benefit
No direct benefit but social housing is available
Means tested
Varies across regions
• Means test
• Receiving RMI
Means tested
• Means test
• Receiving Social Welfare Allowance
-
• Means test
• Children in the household
• Married for less than 5 years
-
• Means test
• Occurrence of a negative “ social event” that
affects housing.
Means test
-
Depending on income
Varies across regions
Depending on income and rent
Depending on Income
Depending on income, rent and family composition
-
Income related
-
Calculation of benefit
Netherlands
Housing Supplement for those under Social
Assistance (but it varies across regions)
Depending on income, family composition and
town of residence
Qualifying conditions
Austria
Housing Supplement for those under RMI
Means tested
Depending on income, rent and family composition
Type of Benefit
Portugal
• Housing benefit for low income households
• Housing benefit for pensioners
• Housing benefit for students
Means tested
Country
TABLE B4. HOUSING BENEFITS IN EUROPE (situation in 1999)
Finland
• Housing benefit for low income households
• Housing supplement for social assistance
recipients
• Housing benefit for pensioners
Means tested
Depending on Income
-
Depending on income, family composition and rent
Sweden
Housing benefit
Germany
United Kingdom
Depending on income and rent. Special
supplements for those under Income Support,
young and old households.
Source: OECD, Benefit System ad Work Incentives (2002).
Denomination
TABLE B5. LOW-INCOME BENEFITS IN EUROPE (situation on January 2001)
Qualifying conditions
Country
Willingness to work
requirement1
Yes
Duration
Unlimited
Yes
-
3 to 12 months
12 months
-
Yes
Yes
Yes
-
Minimum de Moyens d’Existence
(MIMEX)
Unlimited
Greece
Renta Minima
Unlimited
Varies across towns
Belgium
• Nationals and refugees
• Resident in the country
• Aged 18y.o. and above
All persons
Spain
Revenu Minimum d’Insertion
(RMI)
Varies across towns
Yes
Sozial Bistand
France
Supplementary Welfare Allowance
Unlimited
Yes
Denmark
Ireland
Minimo Vitale/Reddito Minimo
• Resident in the country
• Aged 18y.o. and above
Unlimited
Yes
Unlimited
Italy2
Revenu Minimum Garanti
Unlimited
Yes
Sozialhilfe
Luxembourg
Algemene Bijstand
•
•
•
•
12 months, extendible
Yes
Germany
Netherlands
Sozialhilfe
• Resident in the country
• Aged 18y.o. and above
Unlimited
Yes
• National and refugees
• Resident in the country
Austria
Rendimento Minimo Garantido
None
Unlimited
-
Portugal3
Toimeentulotuki
None
Nationals and refugees
Resident in the country
Aged 18y.o. and above
Resident in the country
• Nationals and refugees
• Resident in the country
• Aged 18y.o. and above
Vary across towns
Resident in the country
Aged between 25 and 65y.o.
Resident in the country
Aged 25y.o. and above
Finland
Social Bidrag
•
•
•
•
Sweden
United Kingdom
Income Support
Unlimited
Yes
• Nationals
• Resident in the country
• Aged 18y.o. and above
(1) For those who are able to work.
(2) There is no national legal framework for income support. Many towns, however, have introduced a minimum income scheme but rules
vary widely across the nation. In 1998 an experiment was run in 39 towns with the objective of introducing a national minimum income
scheme by the year 2000. The experiment has been extended to 2002 and then abandoned.
(3) Introduced in 1997.
Source: European Commission Missoc 2001.