PROMOTING ACTIVE WELFARE STATES IN THE EUROPEAN
Lecture by Frank Vandenbroucke
Minister for Employment and Pensions in the
Belgian Federal Government
1. The idea of the “active welfare state” revisited
I don’t think personal
histories are very important, yet there is some personal history involved in
the expression “the active welfare state”. When preparing a press conference in
The term "active
welfare state" actually combines three ideas. First, it refers to an intelligently
active state, i.e. to the way by which government should conduct its social
policy. Referring to an “active state” signals that government, or, “public
authorities” continue to play a key role in this conception of welfare. Yet, it
also signals that the active welfare state is about new approaches to
governance, new approaches at the local level, the national level and the EU
level. Second, the notion refers to an objective, viz. a society of active
people. We want to enable all citizens to participate in the mainstream of
social and economic life. Last but not least, it is not at all a coincidence
that the objective is still called a welfare state: the traditional
ambition of providing adequate social protection is preserved. The second and
third ideas refer to objectives of policy, the first to the mode of governance underpinning
social policy. I will first elaborate somewhat on the objectives, and then
return very briefly to aspects of governance.
Before I do so, let me
emphasize that I do not claim any credit for inventing an expression. I know at
least one paper, published before 1999 by an Oxford-based political
philosopher, Stuart White, where you can find the expression “active welfare
state”. And what is more important, substantively the idea of the active
welfare state built upon some widely shared convictions with regard to social
policy that came to be accepted during the 1990s in most European countries.
These shared convictions boil down to four main points.
SLIDE 1: SHARED CONVICTIONS
First, welfare policy cannot be
reduced to questions of employment and unemployment, but employment is
nonetheless the key issue in welfare reform. Moreover, policy makers agreed
that the fundamental objective had to be increasing the employment rate, rather
than decreasing unemployment rates. In fact, the nature of the employment
objective had changed. Full employment as it was conceived in the past in most
European countries, was employment for men. The new social challenge was employment
for men and women. This made it necessary to rethink the architecture of the
welfare state.
Second, during the nineties there was growing consensus that active labour
market policies should be higher on the agenda and made more effective, both in
quantity and in quality, by tailoring them more effectively to individual
needs. Active labour market policies presuppose a correct balance between
opportunities, obligations and incentives for the people involved. Hence, taxes
and benefits must not lead to a situation in which the poor (or their families)
face very high marginal tax rates when their hours of work or their wages
increase, or when they take up a job. More generally, mechanisms in the current
social security systems that discourage people from being active should be
discarded as much as possible.
Third, the welfare state should not only cover traditionally defined social
risks (unemployment, illness, disability and old age), but also new social
risks, such as single parenthood or a lack of skills causing long-term
unemployment or inferior employment. And it should respond to new social needs,
such as the reconciliation of work, family life and education, and the need to
be able to negotiate changes within both family and workplace over one’s entire
life cycle. Let me put is this way: yes, the traditional ambitions of welfare
policy still remain and are subject to greater pressures than ever, as a
result, for example, of demographic change; but now they co-exist with new
challenges. And let me add as a personal thought, that it should therefore be
obvious that rolling back the welfare state is the worst conceivable option: we
don’t need less, but more welfare.
This brings me to the fourth point that is now accepted by most policy makers.
The traditional welfare state is, in a sense, predominantly a passive
institution. It is only once an undesirable outcome has occurred that the
safety net is spread. It is surely much more sensible for an active state to
respond to old and new risks and needs by prevention.
An intelligently active state recognises the individual vulnerability that is
the result of social dependency. But it recognises at the same time that this
dependency can often be avoided. Increasing dependency is no law of nature but
the result of socio-economic changes. What is needed, in addition to social
spending, is social investment, such as investment in education and training.
For me, the
active welfare state was a deliberate departure from the Bismarckian conception
of the welfare state, which social-democratic parties and trade unions in
countries like
For that reason, I did prefer
the term “active welfare state” to the notion of a “social investment state”,
which you find for instance in Anthony Giddens’ writing.
What I have
said, so far, about the active welfare state is probably not controversial.
Certainly, most people would agree that “getting people back to work” is an
important ambition. Let me nevertheless ask a question that, at first sight,
seems redundant: why does this matter? Why should we aim at more “active
participation” in our society?
SLIDE 2: WHY
There are
two complementary answers to that question.
The first
answer “why” active participation should increase points to the ageing of our
population. Many European countries have a large reservoir of people in working
age who are not in employment; raising employment rates can counteract to a
significant degree the impact of demographic ageing on the ratio of non-working
to working members of the population.
This first answer is
right, but it is somewhat limited in scope. What if there were no such
demographic shift; would a low activity rate then no longer be a problem? There
is a second argument as to why participation matters. However, this second
argument should be used properly.
Encouraging
active participation, it is said, is the best weapon against poverty and the
best guarantee for a fair income distribution. However, put in this way, this
argument is not really adequate. Obviously, for many individuals access to the
labour market is crucial to escape from poverty. Yet, a cross-country
comparison shows that promoting labour market participation is no substitute
for income redistribution and the fight against poverty: more work does not
necessarily mean less poverty in a cross-country comparison. In the
SLIDE 3: FIGURE
1
[Figure 1 is, for the poverty rate, based on Förster, M. (2000), Trends
and driving factors in income distribution in the OECD area, Labour Market and Social Policy Occasional
Paper n. 42,
In Figure
1, the poverty rate is the percentage of working age individuals from 15 to 64
years old living in households with incomes below 50% of median adjusted
disposable income of the entire population. Hence, in this figure poverty is
defined with reference to the national context, that is, on a relative basis.
Moreover, I would insist that such a measure gives you an indication of the
proportion of individuals who are at risk of being poor, but not necessarily
poor. Whether or not they are really poor depends also on other factors, for
instance, the costs they have to incur for housing, or for their health care,
etc.
The
relation between the prevalence of “risk of poverty” and “employment rates” is
not mutually exclusive, it is not an “either – or” relationship; but, at the
other hand, the relationship is not “if – then”: higher employment rates do not
entail less risk of poverty. The relation can and should be “and – and”.
Let me now come back
to my answer to the “why?” question. My defence of participation is not based on the belief that we have
discovered a substitute for social protection. What justifies fundamentally the
importance I attach to participation is a conception of equality.
Since participation in social life is crucial for gaining respect from others
and self-respect, the opportunity to participate actively in society is one of
the basic opportunities that should be the right of everyone. Borrowing some
terminology from John Rawls, the argument may be developed as follows. Rawls
considers self-respect as perhaps the most important of his “social primary
goods” (social primary goods are what social justice should distribute fairly).
Rawls takes it that in a well-ordered society self-respect is secured by the
public affirmation of the status of equal citizenship for all. But, he adds
that self-respect requires that there should be “for each person at least one
community of shared interests to which he belongs and where he finds his
endeavours confirmed by his associates.” In Rawls’ conception of social justice
“the social bases of self-respect” are to be distributed equally. If I
am right to include the opportunity to participate in the social bases of
self-respect, then justice requires that the opportunities to participate be
shared equally too. Or, at least, an unequal distribution can only be just when
it works to the greatest benefit of the least advantaged.
My argument
is an egalitarian argument, yet the egalitarianism of the active welfare state,
as I see it, is responsibility-sensitive. My egalitarianism is based on the
conviction that it is unfair that individuals should be put at a disadvantage
by characteristics or circumstances for which they cannot be held responsible.
Viewed in this way, the pursuit of equality is not a question of equal
outcomes, such as equal pay, independent of personal choice and effort. Yet it
requires more than what is usually understood by “equal opportunities”.
Individual choices are not determined solely by formal social institutions, but
also by natural gifts and talents which individuals received at birth and
through their early childhood. Those who want to give everyone the same
opportunities but ignore such differences hold a narrow, meritocratic view of
individual responsibility. In the view I am committed to, people cannot be held
responsible for differences in talents, only for what they do with those
talents. Here again, our philosophical stance has clear policy implications:
even in a situation of formally equal opportunities (an objective that has so
far not been attained) and full employment, there is still a case for income
redistribution.
In my introduction, I
said that the idea of an active welfare state refers both to objectives of
social policy and to approaches to governance at local, national and EU level.
Putting it very succinctly, one may say that an intelligently active welfare
state does not command but rather delegates. It is not simply an instructing
authority, but one that guides autonomous – and often better informed –
individuals and institutions by encouraging them to take what they see as the
best measures. Encouraging institutions to take what they see as the best
measures, implies allowing or even stimulating policy experimentation.
If one compares the
active welfare state with the traditional Bismarckian approach to social
policy, it should certainly rely relatively more on services, instead of solely
on cash benefits. One may also say that it needs to rely to a larger degree on customised,
individually tailored measures to achieve its aims. Such a tailor-made approach
means that the target groups and objectives must be carefully identified. It often
entails individual coaching. The need for individual coaching has become
generally accepted in the domain of active labour market policies. Yet, it also
applies to health care. Before turning to governance at EU level, I would like
to give you two examples of policy measures I’ve taken in
Wrongly,
everybody thinks the active welfare state is merely about “more jobs”. To
correct this misunderstanding, my examples relate to child benefits and health
care.
Example 1:
family policy
Disabled children get supplementary child benefit in
Example 2
(health care)
The Belgian health care system exemplifies the Bismarckian legacy in our
social security system, with some advantages, but also drawbacks. Health care
is covered with a quasi-universal insurance system, though with one scheme for
salaried workers and another – less generous – for the self-employed. Health
care costs are partially reimbursed; simplifying the system, one may say that
patients have to pay co-payments for each medical treatment, each unit of consumption
of drugs, etc.
The level of co-payment traditionally depended on the social category to
which a person belongs: co-payment was lower for so-called “VIPO”, that is,
widows, disabled persons, pensioners and orphans”. This clearly reflects a
“risk profile” that was maybe adequate for the industrial welfare state of the
1950s, but that is not adequate today. Indeed, a single mother with two
children, who works part-time, may have less income than a pensioner.
Nonetheless, she is not considered a “VIPO”. As a matter of fact, the VIPO
classification has become less relevant. Moreover, the dominant social problem
today is with people who are chronically ill, and who are confronted with an
accumulation of co-payments (in the 1950s many of those who today are
chronically ill, would have died).
Hence, my strategic choice during the last government was to opt for
true universalism (with, ideally, no distinction between salaried workers and
self-employed), and new forms of selectivity, not based on the
VIPO-classification, but on the household budget. I introduced the “maximum
invoice”, which is an income-selective cap on the total amount of co-payment
the household is confronted with during one year.
If the “maximum invoice” entails adequate protection for the household’s
yearly budget
against
the accumulation of co-payments, than co-payments may be used legimately as a
steering mechanism for patients’ behaviour within the health care system,
stimulating some trajectories through the system, via financial incentives or
disincentives, rather than other trajectories. Such co-payments can usefully be
linked to a discourse on “responsibility of patients.” In other words, the
“maximum invoice” also admits a discourse on “responsibility of patients”
without having perverse social effects on the poorer strata of the population.
Within
this new universalism, I also laid stronger emphasis on “tailor-made care”. In
many of cases, patients need and meet different healthcare providers at
different places. This ‘continuum of care’ is not reflected in the existing
financing structures, in the sense that Belgian health insurance provides
separate financing for the various components of such a continuum. Thus, the
structural environment by no means fosters the kind of well-integrated
co-operation that is so crucial from the patient’s point of view. Our task here
is to identify cases where integrated care is most needed (e.g. chronic
illnesses, autism), and to set up new and more appropriate structures that are
work to deal with the patient’s own needs, rather than the administrative convenience
of the provider and the insurance system.
In these examples,
functional recalibration is linked to new patterns of risks in the domain of
health care (e.g. chronic illness), and normative recalibration is linked to
new conceptions of equality and responsibility (with regard to disabled
children; in the health care sector).
Admittedly,
these examples are not very controversial. Functional and normative
recalibration is much more complex and controversial when it comes to pension
policy and early exit from the labour market.
Example 2
illustrates that, at least in Belgium, my defence of the “active welfare state”
encompasses a deliberate departure
from part of the Bismarckian legacy (we have no real universalism in health
care, but separate insurance systems for salaried workers on the one hand and
the self-employed on the other hand). The Bismarckian legacy is problematic
(cf. also the distinction that still exists in
2. The active welfare state as a political
objective for the EU
Putting the
active welfare state on the European agenda was a political statement, both on the objectives of the EU and on its
governance. I strongly believe, with many others, that the European project
needs a social ambition, hence, that the EU should be much more than an integrated
market. Notwithstanding the diversity of social systems and policies within the
EU, so my argument went, the concept of the active welfare state can describe a
common ambition, given common challenges. Moreover, that common ambition can be
specified by means of more or less tangible common goals and objectives, not
just with regard to employment but also with regard to social inclusion and
social protection.
ð
return
to Figure 1
Figure 1
illustrates that it is possible but also necessary, for our policies to aim
simultaneously at both high levels of employment and social protection,
that is, to try simultaneaously to move to the right on the X-axis and
downwards on the Y-axis. Moving to the right on the X-axis is the
quintessence of the “European Employment Strategy”, launched in 1997. According
to the EES,
“Member States should consider setting national targets for raising the
rate of employment, in order to contribute to the overall European objectives
of:
-
reaching by January 2005 an overall employment
rate of 67% and an employment rate of 57% for women;
-
reaching by 2010 an overall employment rate of
70% and an employment rate of more than 60% for women.”
In 1999, at
that press conference where I coined the expression “the active welfare state”,
I argued that the EU should also develop a social protection strategy,
focussing first of all on social inclusion and poverty. With reference to
Figure 1, the argument was that the EU should not only consider the X-axis but also
the Y-axis. In 1999 also, the European Commission tabled a proposal for EU
co-operation on the modernization of social protection. Thanks to the
Portuguese government, these arguments were vindicated at the Lisbon European
Summit of March 2000, which initiated a process of co-operation on social
inclusion, with a view to eradicating poverty.
The
European Employment Strategy, which tries to push the Member States to the
right on the X-axis, is based on a specific form of policy co-ordination, that
differs radically from a top-down, rule-based, centralized approach to
governance often used in social policy. Following two American scholars, who
are close observers of what we are doing, Jonathan Zeitlin and Sabel one may
summarize the essential elements of that new method as follows:
-
Joint
definition by the EU member states of initial objectives, guidelines, and
indicators;
-
National
reports or “action plans” which assess performance in light of the objectives
and metrics, and propose reforms accordingly
-
Peer
review of these plans, including mutual criticism and exchange of good
practices, baced up by recommendations to individual member states in some
cases
-
Re-elaboration
of the individual plans and, at less frequent intervals, of the broader
objectives and metrics in light of the experience gained in their
implementation
The 1999
argument was that in the field of social inclusion, the EU should apply a
broadly similar mode of governance, albeit somewhat looser than the method
applied in the EES. This methodological idea was also vindicated at the Lisbon
Summit, which coined the expression “open method of co-ordination” as a generic
label for this mode of governance.
In a
remarkable contribution to our European debate, which you may find in the
collection of papers for tomorrow’s workshop, David Trubek and Louise Trubek
argue that the open method of co-ordination is a natural methodological
corollary of the new social policy vision which I call the active welfare
state. This new vision, so they write, presents immense problems for a governance
approach based on the traditional methods of the EU. “It takes the EU into
areas that have traditionally been the exclusive province of the
In
tomorrow’s workshop, we will assess the OMC. I will argue that, for me,
launching the Open Method of Co-ordination on social inclusion (and then on
pensions, possibly on health care) was much more than launching a “policy
learning process”. The OMC should be a creative
political process, in which the EU Member States specify what the idea of a
“European Social Model” really means. It should be creative at the local level,
in fostering better policies, and creative at the EU level, in fostering a
common understanding of the overarching social ambition in which these local
policies fit.
Yet it also
follows from my view on the active welfare state that the renovation of EU
social policy cannot be confined, with regard to governance, to the OMC (cf. the
emphasis on patient mobility and the “hard” legal architecture of the EU in my
Max Planck lecture on the European Convention, see www.vandenbroucke.com).
But today,
I would like to comment briefly on some substantive results and problems.
3. The active welfare state as EU policy
Quoting
Trubek and Trubek, I said that the policies we want to pursue deal with areas
of policy in which there are great differences among the EU Member States. The
differences are not just parametric, they are structurally embedded in the
Member States’ social and political history.
Building
upon the typology of welfare states pioneered by Gosta Esping-Andersen, one can,
with Ferrera and Hemerijck, distinguish four types of welfare regimes in
ACTIVATE
SLIDE 3 A SECOND TIME [=> Anglo-Saxon cluster of welfare is emphasised]
“The
Anglo-Saxon welfare cluster is characterised by a bias towards targeted,
needs-based entitlements; low replacement rates in transfer programs; general
revenue financing; underdeveloped public social services beyond health and
education; poor family services; low levels of employment protection, largely
confined to ensure fair contracts, and no legacy of active labour market
policy, nor vocational training and education; uncoordinated industrial
relations with moderately strong unions, decentralized wage bargaining, and low
levels of collective bargaining coverage.
ACTIVATE
SLIDE 3 A THIRD TIME [=> Continental European cluster of welfare is
emphasised]
The
Continental European Welfare State, historically influenced by a mix of
etatiste, corporatist and familialist traditions is characterised by
occupationally distinct employment-related social insurance; very unequal
levels of generosity in transfer programmes, combining generally very high pension
replacement rates; a contribution-biased revenue dependency; very modes levels
of public social services beyond health and education and often a considerable
reliance on “third sector” and private delivery; passive family policies
premised on the conventional male breadwinner family; generally strict levels
of employment protection that is meant to protect, once again, the male
breadwinner combined with passive labour market policies, but comprehensive
systems of vocational training and education, especially in Germany, Austria,
and the Netherlands; strong social partnership that extends into the
administration of social insurance; and co-ordinated industrial relations, with
a predominance of sectoral wage bargaining, with high levels of bargaining coverage
and moderately strong unions.
ACTIVATE
SLIDE 3 A FOURTH TIME [=> Southern European cluster of welfare is
emphasised]
The
Southern European welfare family institutionally resembles the continental
welfare states, but provides more chequered and unequal coverage in terms of
public services and social insurance, with disproportionately high expenditure
on retirement. Holes in the social safety net are patched by family members in
countries such as
ACTIVATE
SLIDE 3 A FIFTH TIME [=> Scandinavian
cluster of welfare is emphasised]
The
comprehensive Scandinavian welfare states are characterised by
citizenship-based universal entitlements; generous replacement rates in
transfer programs; general revenue financing; a broad supply of social services
beyond health and education, active family policy encouraging gender
egalitarianism and women’s integration in the labour market; low (in Denmark)
to high (in Sweden) levels of employment protection, with a strong emphasis on
active policies and training programs linked to general education; and
corporatist industrial relations with peak level bargaining, strong unions and
high levels of collective bargaining coverage.”
Figure 1
illustrates that the four regimes perform very differently with regard to the
twin dimensions of employment and poverty. Hence, it suggests that improving
performance does entail “regime changes” (at least partial regime changes).
This explains the difficulty of the exercise we are engaged in.
But, when it
comes to regime changes, one should not jump to conclusions, Should we all
become Nordic? The Nordic model has definite advantages, I believe, in
comparison to the Continental model, but the
Figure 2
illustrates that from the mid 1980’s to the mid 1990’s some important changes
took place in the performance of OECD countries with regard to employment and
poverty:
SLIDE 4: FIGURE
2
Figure 4 compares the mid 1980’s with the mid 1990’s
[data based on Förster and OECD Employment Outlook]
In
tomorrow’s seminar, we will assess the Open Method of Co-ordination. I will
suggest that we should specify carefully the focus of our assessment: do we
want to assess…
- the
extent to which the OMC contributed to “local” policy change?
- the
extent to which the OMC contributed to the development of a common framework at
EU level, specifying the meaning of the “European Social Model”?
- the
extent to which the OMC contributed to change in the balance of political
forces between various formations of the Council?
I will also
argue that one should distinguish long-term and short-term effects, which may
be of a different nature.
And, when
it comes to evaluating the short-term impact on “local” policy change, I will
explain that an evaluation can either look at facts (such as employment rates),
or at policy formulation, or, screen really existing and effective mechanisms
of influence.
Simply
looking at the facts may be rather deceptive, since one still must prove
causality. This being said, a very simple presentation of the most stylized
facts is provided in my next and last figure, Figure 3:
Figure 3
answers the following question: what happened after launching the European
Employment Strategy?
SLIDE 5: FIGURE
3
[Poverty rate = most recently available, mid 1990’s; my guess is that
poverty rates did not change very much in this period. For employment rates I
take averages, in order to eliminate short term effects: Employment rate for
1998 = average of 1997, 1998, 1999; Employment rate for 2002 is average of
2001, 2002]
At first
sight, Figure 3 shows progress with regard to employment rates, but no
convergence between the various welfare clusters.
At
tomorrow’s seminar, I will argue that a weakness in the implementation of the
European Employment Strategy, so far, has been that Member States “pick and
choose” some measure, but do not necessarily adopt the comprehensive package,
i.e. do not necessarily adopt the underlying strategy. Hence, positive
measures, inspired by the European Employment Strategy, are countered by other
measures which are totally at odds with the strategy. In order to adopt the
European Employment Strategy as a “comprehensive package”, all the relevant actors must me convinced by the strategy.
4. The difficulty of the active welfare state
Earlier I
said: “Wrongly, everybody thinks the active welfare state is merely about more
jobs.” Yet, it certainly is about
more jobs. The
Let me, to
launch the discussion, point to two problems which I have to overcome if I want
to achieve those objectives.
The first
problem concerns the incentives for unemployed people to look for a job. Here
is a kind of trilemma. The active welfare state, as I see it, postulates three
requirements:
-
no
one who is involuntary unemployed, should live in poverty;
-
to
avoid “inactivity traps”, the difference between the income a person can earn
via a job and the unemployment benefit should constitute an incentive for that
person to take that job; if necessary we have to top up the job income by
mechanisms such as EITC, WFT, mechanisms which we also develop in Belgium;
-
to
avoid “salary improvement traps”, when a person increases his working hours, or
when his boss improves his level of pay, a positive impact on net earnings
should also be felt by that person.
It is in
fact impossible to solve this equation, if agencies who deal with unemployed
persons solely rely on financial incentives. There must also be some kind of
administrative pressure on people to continue looking for jobs, and
administrative elimination of people who don’t.
ð
In
practice, as long as there is involuntary unemployment, we need to develop
procedures for monitoring the effort unemployed people make to get out of their
situation, even if there is no job on offer with which we can test “on the
spot” their willingness to work. This is easier said than done, and raises
complicated issues of justice.
The second
problem concerns the incentives for older people to continue working. In
Continental Europe, people will have to work longer. That is certainly the case
in
In a study
for the Belgian Presidency of the EU, John Myles points out the following
difficulty:
When
retirement ages were falling, the social welfare ‘gains’ in additional leisure
and free time tended to go disproportionately to the least well off. An
additional year of retirement, for example, represents a larger proportional
gain for someone with a 7-year life expectancy after retirement than for
someone with a 12-year life expectancy. But the reverse is also true: an
additional year of employment represents a proportionately greater loss for
those with shorter life expectancies. Raising the retirement age for public
pension benefits has the largest effect on those without sufficient means to
finance early retirement on their own and the least impact on those who do.
Since health and wealth tend to be correlated, the equity problem is
compounded.
Belgium,
despite health care system: if you divide men at the age of 25 in quintiles
according to their level of education, you get the following results: life
expectancy for a man belonging to the lowest education quintile is 27 years in
good health and 16 years in poor health; life expectancy for a man belonging to
the highest quintile is 42 years in good health and 4 years in poor health. The
highly educated live longer, and, moreover, live much longer in better health
conditions. Hence, raising the retirement age is, comparatively, not a big loss
for someone Frank Vandenbroucke, but it certainly is a big loss for the blue
collar worker to who I send that message. There are no easy solutions to this
problem, but it is a problem which can be handled: flexibility in retirement
age, and a sufficiently robust basic pension guarantee for those who want to
leave early. It should not deter us from moving forward with reforms. But we
should however not be blind to the deep problems of justice posed by a policy
of raising the effective retirement age. The lowly qualified blue collar worker,
who approaches retirement, should know that
EU policy makers are aware of his condition and worries.
CONCLUSION
I said
launching the OMC was not just a policy, but a political statement. On the
level of policy, the OMC should contribute to “functional and normative
recalibration”, which is an urgent necessity in an enlarging and ageing EU.
Yet, that recalibration may meet a lot of resistance, for very fundamental
reasons. OMC and the active welfare state risk becoming scapegoats, to which
national policy makers only refer to explain the unpopular measures they have
to take, rather than mobilising projects. That would discredit the idea of
“social