Nordic activation policies in flux: A study of Swedish and Danish labour market
policies
Mattias Bengtsson
Department of Sociology and Work Science
University of Gothenburg
P.O. Box 720
405 30 Göteborg
Sweden
Email: Mattias.Bengtsson@socav.gu.se
Introduction
Organisations as the OECD and the EU have promoted active labour market policies
(ALMPs) as part of an ‘activation turn’ from passive to active measures – in order to develop
‘activating states’ or an ‘active society’ (Lødemel and Trickey 2001; Barbier 2005). 1 The
Nordic countries have been defined as ‘an active corner’ in Europe. Sweden pioneered
ALMPs at the mid-20th century, an important pillar of the so-called Rehn-Meidner model.
However, with the economic recession in the 1990s a situation of de facto full employment
transformed into a post-full-employment situation.2 Instead of the Swedish model, Denmark
has overtaken the role as the European labour market/welfare model of the 21st century. The
Danish ‘job miracle’ is often explained as a result of Danish flexicurity (Larsen 2004).3 If the
Swedish ‘Rehn-Meidner model’ could be seen as a 20th century mid-way between capitalism
and socialism, the Danish ‘flexicurity model’ has been launched as a third way besides more
deregulated Anglo-Saxon countries and stricter employment protection regulation in southern
European countries (OECD 2004: 97).
Danish flexicurity has much in common with the Rehn-Meidner model: ‘Swedish policy
makers formulated a remarkably coherent combination of policies which amounted to nothing
less than an integrated ‘flexicurity’ plan avant la lettre’ (van den Berg 2008: 4). As in Danish
flexicurity, ALMPs in the Rehn-Meidner model acted as a mechanism to socialize labour
market risks through employment security in times of manpower redundancies. How has
institutional change in a Swedish post-Rehn-Meidner era and in a Danish flexicurity era
affected the direction of labour market policies? The aim of this paper is to analyze Swedish
and Danish labour market policies from the 1990s and onwards.
1
Active and passive measures are well-established descriptive terms, but also normative ditto in
describing social security claimants as being ‘passive’. Neither does ‘passive’ account for peoples’
involvement in activities and work outside the formal labour market (cf. van Berkel and Hornemann
Møller 2002: 47) or that income security enhance peoples’ risk-taking activities, as understood in the
Rehn-Meidner model and in the Danish flexicurity model.
2
In Sweden there was overall a situation of full employment from the 1960s to the early 1990s. In the
first half of the 1990s unemployment accelerated. Thereafter, it has fluctuated around 5-8 percent.
3
The Danish unemployment rate that reached close to 11 percent in 1993 declined to 3.4 percent in
2008 (OECD 2012).
1
Activation approaches
Welfare state restructuring in Europe could be analyzed through ‘an activation lens’ (Barbier
2005: 417). Activation is here defined as a ‘broad range of policies and measures targeted at
people receiving public income support or in danger of becoming permanently excluded from
the labour market’ (Drøpping, Hvinden and Vik 1999: 134). Activation could be seen as a
more comprehensive concept than ALMPs, but the boundaries between the concepts are
vague.
Activation is ‘Janus-faced’ as public policy approaches dealing with the question of
unemployment have emphasized both the prevention of negative individual consequences of
joblessness and social exclusion through skill and personal development and the
strengthening of work incentives and the increase of labour supply in the remaking of social
security, through a reduced ‘dependency’ on social transfers and a restoration of civic duties
and social discipline (e.g. Lødemel and Trickey 2001; Barbier 2004; Bonoli 2009, 2010,
2012).
One of several classifications that stress the Janus faced character of activation is ‘offensive’
and ‘defensive’ strategies (Torfing 1999). An offensive activation strategy is said to
characterize universalistic welfare states where the unemployed is seen as an active being
with a civic right to receive support in order to develop his or hers resources and skills. An
offensive activation strategy is a characteristic of the social democratic welfare regime (cf.
Esping-Andersen 1990). A defensive strategy is characteristic for residual welfare states
based on a perception that too generous benefits will form a culture of dependency and that
it is the individual’s obligation to be activated, even through enforcement.
In one activation approach, ‘paternalism optimists’ and ‘activation optimists’ share a basic
understanding of the importance of active state intervention to employ people, but there are
main differences as well (van Berkel and Hornemann Møller 2002: 51-56; Johansson and
Hornemann Møller 2009: 19-23). According to the paternalism optimists, state intervention is
important to activate the ‘passive’ benefit recipients. ‘Enforced emancipation’ captures the
essence of this approach, whereby force and social discipline are in the interests of both the
individual and of the society. If the paternalism optimists stress individual responsibilities, the
activation optimists emphasize state responsibility to support the individual to re-enter the
labour market, and that activities should take into account prerequisites and desires of the
individual. This corresponds to an offensive activation strategy (cf. Torfing 1999). An
unemployed is perceived as willing to be activated, not as an idle human being that needs to
be socially disciplined.
As well, a typology of four types of ALMPs is used (Bonoli 2010: 440-41, 2012: 182-86). Two
dimensions are the basis for this typology. The first dimension, the ‘pro-market employment
orientation’, concerns whether there exists weak or strong policy objectives to ‘put people
back into real jobs’. The second dimension concerns the level of human capital investment,
from none (as in ‘passive’ economic benefits) to strong. The first type is labeled ‘occupation’,
with a weak pro-market employment orientation and a weak investment in human capital.
Instead of human capital development it consists of measures as job creation schemes and
shorter training courses to keep people active and to prevent human capital reduction. The
second type, ‘incentive reinforcement’, is characterized by a strong pro-market employment
orientation and none human capital investment. The aim is to strengthen work incentives
2
through a reduced dependency on welfare benefits. 4 The third type is ‘employment
assistance’ with a strong pro-market employment orientation and weak human capital
investment. It consists of measures as placement services, job subsidies, counseling and job
search programmes that aim to match job seekers with potential employers. The weak
human capital investment refers mainly to the improvement of ‘soft skills’. Finally, the fourth
type is ‘upskilling’ with a strong pro-market employment orientation as well as strong human
capital investment. Jobless people receive job-related vocational training – a hallmark of the
so-called ‘social investment approach’ (cf. de la Porte and Jacobsson 2012).
The Swedish case
Sweden and Denmark are usually classified in a social democratic welfare regime,
characterized by relatively generous and comprehensive public social insurances as well as
a clear link to previous labour market earnings (Esping-Andersen 1990). A main policy
principle in the Swedish welfare state is the work strategy. That is, social rights in form of
compensations from the social insurance system are tied to the role as a worker (Garsten
and Jacobsson 2004: 9; Junestav 2011).
Post-war policies were to a significant degree governed by the Rehn-Meidner model, aiming
to solve as diverse problems as economic growth, price stability, full employment, wage
levelling and union solidarity. As flexicurity is supposed to deliver institutional
complementarity at the Danish labour market (Klindt 2008, 2009), the policies designed by
Gösta Rehn and Rudolf Meidner, economists working at LO – the blue-collar union
federation – were supposed to lead to ‘that flexibility could be had as a result of security’ (van
den Berg 2008: 6). The social contract meant employee acceptance of rapid structural labour
market rationalizations, aimed to strengthen the competitiveness of the small, exportdependent Swedish economy, in exchange for income and employment security. ‘The
security of the wings’ was the motto (Rehn 1988: 214). An important condition for the policies
to be accepted was that LO could work as an ‘encompassing organization’ for the whole
working class (van den Berg 2008: 7-8).
Both fiscal and wage policy were components as the model combined fiscal restraint and a
‘solidaristic’ wage policy. Instead of relying on wage dispersion as a mechanism of allocating
labour, extensive investments in ALMPs should enhance occupational and geographical
mobility. Job security was substituted by employment security, i.e. shifting focus from the
protection of a specific job to ensure people to get any kind of job. Training, retraining,
relocation, and mobility allowances were main instruments. As well, supply-side policies were
a response to the problem of labour shortage during a period of rapid economic growth.
Another pillar of the model was a generous income security, which not only worked as social
protection but also as a work incentive, as the individual should participate in ALMPs and
even be ready to resettle in regions with strong labour demand (Rehn 1988; EspingAndersen 1990; van den Berg, Furåker and Johansson 1997; van den Berg 2008; Junestav
2011).
4
It could be contested whether the tools mentioned under incentive reinforcement should be classified
as ALMPs as these are not services or programmes but concerns the intensity of work incentives in
national LMPs and financial instruments as tax credits. I see them as part of the wider concept of
activation – indicators of the intensity of work incentives (cf. Bonoli 2009: 63; Venn 2012) to be
analyzed in parallel with the other types of ALMPs (cf. van Berkel and Hornemann Møller 2002: 49).
3
Consequently, the mix of policies with generous unemployment insurance (UI) and an
extensive use of ALMPs was a Swedish trade mark. However, the policy of substitution of job
security with employment security was to some extent undermined by stricter job protection
legislation in the 1970s. Together with massive state subsidies to enterprises following the
economic recession in the mid-1970s, it meant a significant retreat from the policies of the
model (van den Berg 2008). The content of ALMPs shifted in the 1970-80s in a period of
economic stagnation and industrial restructuring towards more demand-side interventions
such as job creation schemes in the public sector (Bonoli 2012). In addition to this, the
political climate changed – a mode of policymaking based on monetarist economic theory
followed (Ryner 2002). However, despite the policy shift massive state interventions were still
put into fighting unemployment.
The Danish case
In the 1960s, labour market policy became an independent policy field in Denmark. A by an
international standard very generous unemployment support system was established in an
era of economic growth and full employment, but from the 1970s to the early 1990s mass
unemployment prevailed (Larsen 2004). Even though a heavy burden was placed on public
expenditures, the UI continued to be very generous as the ‘historical compromise struck in
1960s to balance labour market flexibility against security protection for wage earners was
upheld’ (Larsen 2004: 145).
Two of the three main pillars of the Danish flexicurity model are a relatively limited job
security and generous unemployment benefits. According to Danish flexicurity, a developed
welfare state with high tax and benefit levels do not oppose a more flexible labour market.
The generosity of the UI has not only been defended by the unions but also by the employer
side, it has seen this as ‘a reasonable price to pay’ in exchange for the continuation of the
managerial prerogative to hire and fire personnel following the less strict employment
protection legislation (van den Berg 2008: 10). As mentioned in the case of Sweden, an
important condition for reaching these agreements has been that the social partners have
acted like encompassing organizations.
The Danish employment schemes have traditionally been of a largely passive nature.
However, the third pillar of the flexicurity model, which was set in place in the 1990s, is
ALMPs. These are supposed to have two main effects. First, a qualification effect that will
follow from upskilling. Second, a motivation effect to be enhanced among the unemployed
approaching the date when they need to participate in active measures, i.e. as they probably
will see this as something negative they will increase their job search (Madsen 2010: 64-65).
To conclude, the flexicurity model is seen as a compromise between reduced job security in
exchange for high employment security and income security – a compromise that has been
reached in an industrial relations regime where strong trade unions and employers have had
considerable freedom to regulate working conditions through collective bargaining.
Results
In this part, government bills, labour market statistics and research reports are studied to
analyze the development of Swedish and Danish LMPs. The main data analyzed is OECD
4
data on public expenditures on labour market programs from 1985, the first year with data, to
2009 (OECD 2012). The OECD reports public expenditure as a percentage of GDP of labour
market programs. Besides this, Figure 1 and Figure 3 are standardized for unemployment
level through dividing the public expenditure in passive and active measures of the GDP with
the unemployment level (e.g. van den Berg et al. 1997; Bengtsson and Berglund 2012).
Changes in the Swedish unemployment insurance
The UI consists of a basic insurance and an income-related insurance. The basic insurance
is valid for the person that satisfies the work condition but that is not a member in an
unemployment insurance fund or has not been a member long enough. The income-related
insurance demands that the person both satisfies the work condition and the membership
condition; that is, having been a member in a fund for at least 12 months.
In 1989, to fulfil the work condition the person needed to have worked, during the last 12
months before unemployment, at least 75 days over a four-month period. The compensation
level was 90 percent of previous earnings and the duration was 300 days; thereafter a new
work condition had to be fulfilled by the individual to be able to receive a new benefit period.
Older employees were entitled to a slightly longer period (SO 2008). Before 1989 there had
been a waiting period of normally six days but it was abolished that year. In 1993, during the
economic recession, the work condition was strengthened, five waiting days were
reintroduced and the compensation level was lowered to 80 percent.
In the new century, the unemployment benefit has increasingly been seen as a
compensation for the activity of job seeking; it has transformed into ‘activity compensation’,
i.e. it is the activity, not unemployment itself, which is the basis for the compensation. A sign
of a strengthened pro-market employment orientation through ‘incentive reinforcement’ (cf.
Bonoli 2010, 2012) is stricter eligibility criteria. In 2001, demands on activation were
strengthened. An unemployed shall actively collaborate with the case worker in creating an
individual action plan. The definition of ‘appropriate job’ changed, the unemployed should
during the first 100 days search for jobs in their area of occupation and close to their
residence, but after that at the whole labour market. Also, participation in active measures
did no longer qualify for a new benefit period. Furthermore, the entitlement to a somewhat
longer duration among older workers was abolished (Sörnsen 2009: 243).
During the current centre-right government, to fulfil the work condition a person needs to
have worked, during the last 12 months before unemployment, at least 80 hours a month in
at least six months or, alternatively, 480 hours during six consecutive months and then at
least 50 hours per month. The duration is 300 days. An exception is parents with children in
the household under 18, who can receive 450 days. The compensation level for the first 200
days is 80 percent, thereafter 70 percent until day 300. The waiting period is seven days. If
the applicant during the benefit period once again meets the work condition, he/she can
receive a new benefit period of 300 days. The compensation will then be calculated on the
applicant’s income during the last 12 months. If it is preferable, the applicant can instead be
granted a day-wage at 65 percent of the daily earnings that the benefit was based on in the
earlier benefit period (Government Bill 2007/08:136, p. 46; Sibbmark 2010: 12-13). Besides
this, the unemployed should be prepared to accept a job at the whole labour market from the
first day of being registered at the local employment office.
5
The net compensation rate of an average industrial worker decreased with as much as 13
percent between 2005 and 2010. It is currently below 60 percent, which is below the OECD
average. This could be compared to the early 1990s when it was around 85 percent
(Ferrarini, Nelson, Sjöberg and Palme 2012: 11, 29, 33). One important factor is that the
nominal income ceiling has not been raised since 2002 – there is no automatic indexing, that
is, the value of the compensations during the last decade has strongly decreased relative to
the wage development. This change has increased the demand for private insurances
(Ferrarini et al. 2012: 33-36, 60-62).
Fewer and fewer full-time unemployed persons receive any unemployment benefit
whatsoever – the proportion fell from 70 percent in January 2006 to 36 percent in November
2011 (Swedish Public Employment Service 2012: 43). Also, the monthly membership fees to
the UI were dramatically raised. At the end of 2010, well over 1.4 million individuals, nearly a
third of the Swedish labour force, were not part of the UI, compared to around 700,000
individuals in 2006 before the current government took office. They did not, therefore, meet
the membership condition in case of unemployment (Kjellberg 2012: 6).
Swedish expenditures on active and passive measures
During 1985-2009, the highest relative levels spent on active measures are in the period of
1985-91. By studying Figure 1 we can see that Swedish expenditures during the whole
period follow a pattern. With the exception of 2004, the expenditure level has been higher for
active than passive measures. Of the total expenditure in 1985-90 as much as 70 percent
was invested in active measures, which can be compared to 57 percent in the final years of
2004-09 (OECD 2012). However, the conditions for LMPs changed dramatically with the
economic recession in the early 1990s. Mass unemployment and strict austerity measures
followed. In 1992, public expenditures on LMPs peaked at 5.76 percent of the GDP, but for
each percentage point unemployed it dropped considerably (Figure 1). If ALMPs had been
developed to handle unemployment on the margin, the effectiveness and quality dramatically
weakened as the volume of participants nearly four-doubled 1990-94. The legitimacy of state
ALMPs and of the public employment services was challenged (Lindvall 2011). Despite mass
unemployment in the mid-1990, the expenditures on LMPs fell steeply both during the
centre-right government (1991-94) and during the succeeding social democratic
governments (1994-2006) (Figure 1).
At the turn of the century guarantees were introduced, which is a combination of measures
as intensified guidance, individual action plans, training and supported employment. These
are usually introduced in a specific order and at fixed dates. Participants are selected based
on unemployment duration and participation is obligatory. The first program was the 1998
Youth Guarantee. In 2000, the Activity Guarantee was introduced – a fulltime, umbrella
programme with no definite time limit for persons aged 20 or older that was, or risked
becoming, long-term unemployed. These should have stable fulltime activities until finding a
job or an education to counteract unemployment cultures, hinder abuse of the insurance and
moonlighting as well as activating those hit either by structural problems in sparselypopulated areas or in larger segregated urban areas. Also, a place in the guarantee was
seen as a solution to individuals whose period of unemployment benefit had expired
(Government Bill 1999/2000:98, p. 57-8; Johansson 2006: 37ff). Those who declined job
offers or to participate in activities, misbehaved or did not participate in the constitution of an
6
individual action plan could be dispelled from the guarantee and, thus, from the right to
compensation (Johansson 2006: 46ff.).
3.5
3
2.5
2
Active
Passive
1.5
Active (s)
Passive (s)
1
0.5
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
0
Figure 1 Total expenditures in Sweden 1985-2009 on labour market programs of share of GDP as
well as expenditures on share of GDP for percentage point unemployed (percent).
Source: Own calculations based on OECD (2012)
The introduction of guarantees has meant a contractualisation of the relationship between
the citizen and the state. The guarantee takes the form of a ‘client contract’ in which the
unemployed person and the local employment officer signs an agreement. However, the
contract consists of an asymmetric relation between the state and the citizen as it works as
an instrument to test whether the unemployed actively stands at the disposal at the labour
market. Following this, citizenship rights have become contractual as well as more
conditional (Johansson 2006: 48-9).
In 2007, the Youth Guarantee was replaced by the Youth Job Programme and the Activity
Guarantee was replaced by the Job and Development Programme (JOB), currently the most
extensive programme. The target group of the first programme is youths (16-24 years) and is
first and foremost directed to those that have been unemployed during a coherent period of
three months but also for people on parole and for those that have worked less than their
whole labour supply (Sibbmark 2010: 16-23).
JOB is divided into three phases. First, in-depth assessment, job seeking activities with
coaching and preparatory measures (for 150 days). Second, work placement, practical
training or different forms of work training (for an additional 300 days). Third, supported
employment for maximum two years is supplied in ‘artificial jobs’; that is, the participants
shall perform tasks otherwise not performed at the workplace (SOU 2011:11, p. 214). This
7
third phase has been heavily criticized for consisting of rather meaningless work, tasks that
in earlier research have been described as ‘participation in futile work-for-the-sake-ofworking projects’ (Torfing 1999: 18). Persons that have exhausted the benefit period in the UI
are normally offered to participate in JOB, with significantly lower economic compensations,
often forcing the job seeker to apply for support from the social services.
An important part of ‘incentive reinforcement’ is in-work benefits. During the current centreright wing government there have been a number of tax reforms. The job tax deduction is the
most significant change in direct taxation of work incomes. The government emphasizes that,
together with the changes in the unemployment and sickness insurances, the deduction will
increase the work incentives and lower the thresholds for labour market entry, as well as
increase the hours worked (Government Bill 2010/11:1, p. 26, 51, 59).
Between early 1990s and early 2000s there were major cuts in public expenditures on LMPs,
specifically for out-of-work income maintenance and employment support (Figure 2). The
pattern for labour market training is of specific interest here. In 1986, training as a
qualification for a new unemployment benefit period was introduced. In 1991, as many as 66
percent of participants in various LMPs participated in vocational training (Swedish Public
Employment Service 2010). However, in the early 1990s expenditures on training steeply
decreased. However, in the final years of the 1990s the unemployment rate decreased and
the social democratic government launched a policy on lifelong learning and human capital
development – the work strategy was complemented with a ‘competence strategy’. The
investments in training increased, but after a couple of years they once again decreased
before making a halt in 2003-06 at a level of 0.31-0.36 percent of the GDP (OECD 2012).
3
2.5
2
1.5
Public
Employment
Service
Training
1
Employment
support
0.5
Out-of-work
income
maintenance
0
1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009
Figure 2 Total expenditures in Sweden 1985-2009 on four areas of labour market programs of share
of GDP. Source: OECD (2012)
8
With the introduction of new instruments as JOB, training has significantly declined and the
emphasis has shifted towards matching, job counseling and job coaching. The percent of
GDP for LMPs was 1.79 percent in 2007, 1.45 percent in 2008 and 1.85 percent in 2009, the
three years with the lowest shares for the whole period of 1985-2009. The year with the
fourth lowest expenditure was 1989 with 2.16 percent, but the unemployment rate of 1.6
percent was much lower than the 6-8 percent in 2007-09 (OECD 2012). During the centreright government the expenditure on training decreased to a very low level – as percentage
of GDP it was only 0.06 percent in 2009 and fallen below the OECD average (0.18 percent).
However, the expenditure on ALMPs in Sweden that year, 1.13 percent, was significantly
above the OECD average, 0.62 percent. Simultaneously, the expenditure on passive
measures, 0.72 percent, was below the OECD average of 1.04 percent (OECD 2012).
Changes in the Danish unemployment insurance
Denmark has for a long time been a deviant case with its extended public expenditures on
LMPs, specifically due to a very generous UI. In 1989-95 the expenditure on passive
measures as percentage of GDP never fell below 4 percent. Today the total expenditure on
LMPs is still very high by international standards – of the OECD countries Denmark was only
surpassed by Spain and Belgium in 2009 (OECD 2012). Denmark, as well as Sweden, has
voluntary, union-administered unemployment insurance funds, the so-called Ghent system.
The compensation level amounts to a maximum of 90 percent of the previous wage, and
there are no waiting days. Following this, both the compensation level and waiting days has
not changed since the end of the 1960s (Klindt 2009: 65).
The unemployment benefit consists of both an initial ‘passive’ and a subsequent ‘active’
period. To be entitled one must have been a member for at least 12 months. In 1985, the
maximum benefit duration was set to 9 years. In 1994, the social democratic government
restructured LMPs towards an activating state – described as a shift from income support to
employment (Klindt 2009). This was made in a situation with decades of high unemployment.
The unemployment rate reached close to 11 percent in 1993 (OECD 2012). In 1983-90 the
long-term unemployment rate (6 months or more) as percentage of the labour force was 35.5 percent – the corresponding figures in Sweden were below 1 percent during the period
(Furåker 2003: 179-80).
With the labour market reform act of 1994, the benefit period was reduced to seven years,
and the unemployed should be activated faster. Stricter eligibility criteria and compulsory
activation was implemented. Re-qualification for a new compensation period through
participation in active measures was disallowed and the work condition was expanded from
26 to 52 weeks. At the end of the 1990s, UI duration was first reduced to five years, and then
to four years. In 2001, a liberal-conservative government was elected and with the reform
‘More people to work’ (‘Flere i arbejde’) stricter eligibility criteria and stronger economic
incentives was implemented, e.g. an unemployed person had to accept every job offer from
the first day on which he or she was to be activated. In 2010, a crisis package was passed
and the benefit period was lowered from four to two years (Jørgensen and Schulze 2011).
9
Danish expenditures on active and passive measures
By comparing Figures 1 and 3, the relative proportion of active and passive measures has
taken very different forms in Sweden and Denmark. While major investments in active
measures have been a Swedish trade mark, passive measures have been a main
characteristic of Danish LMPs – with 2.5-4 percentage points more invested in passive than
in active measures from the mid-1980s to the mid-1990s. However, in the 1990s the
proportion of active measures of overall spending on LMPs strongly increased – a major
reorientation in combination with major cuts on passive measures since the highest level in
1993. In 2009, the total expenditures on passive and active measures were more or less
equal (Figure 3).
Danish unemployment policies in the 1970-80s were characterized by a low degree of work
incentives. Activation was weakly developed with a main commitment to provide welfare and
to maintain the living standard (Torfing 1999: 13). However, by international standards,
activation policies were introduced early (1989-90) on municipal level in the fight against
youth unemployment. In 1994, the labour market reform act emphasized activation policies
and reforms of the benefit systems, as mentioned above. Larsen (2004: 147) underlines that
‘a giant leap was made to an expansive fiscal policy’. Besides more individual and tailormade measures, ALMPs were directed towards upskilling. Education and job training were
the main measures being offered – a ‘learn-fare’ approach (Larsen 2004; Jørgensen 2009).
6
5.5
5
4.5
4
3.5
Active
3
Passive
2.5
Active (s)
Passive (s)
2
1.5
1
0.5
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
0
Figure 3 Total expenditures in Denmark 1986-2009 on labour market programs of share of GDP as
well as expenditures on share of GDP for percentage point unemployed (percent).
Source: Own calculations based on OECD (2012)
10
Studying Figure 4 the reorientation in the mid-1990s towards active policies and training
measures are evident. As unemployment quickly decreased and the expenditures on out-ofwork income maintenance steeply declined during the final six-seven years of the 1990s
(approximately three percent of GDP), the investments in training simultaneously increased.
There was also a minor increase in public employment services, while expenditures in
employment support decreased in the mid-1990s, but has thereafter been on a rather
consistent level.
In the new century a rather dramatic decrease in expenditures on programs as share of the
GDP has been seen, with the exception for 2009 due to the effects of the global financial
crisis. The labour market reform of 2003 meant less upskilling and stronger emphasis on
measures as guidance and contact meetings and subsidized job training. There was also a
greater emphasis on social disciplining and workfare policies (Larsen 2004; Klindt 2009).
Part of this was a change from more qualitative and attractive offers to more low-cost, lowquality offers to encourage unemployed persons to find a job of whatever sort – i.e. a
strengthening of the ‘motivation effect’ of active measures (Jørgensen 2009).
5.5
5
4.5
4
3.5
Public
Employment
Service
3
2.5
Training
2
1.5
Employment
support
1
Out-of-work
income
maintenance
0.5
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
0
Figure 4 Total expenditures in Denmark 1986-2009 on four areas of labour market programs of share
of GDP. Source: OECD (2012)
It could be stressed that unemployment fell during the period in question to around 3-4
percent (before the major rise in 2009) which affected the total public expenditures. Still, it is
evident that there has occurred a policy change. Expenditure cuts on out-of-work income
maintenance were extensive in 2004-08 and from the turn of the century there were major
11
cuts in spending on training – in 2008 it was at the same level as in the end of the 1980s and
early 1990s. The decline has not been as steep as in Sweden, but it is a similar pattern. At
the second half of the 1990s, upskilling was stressed in both countries, but in the last decade
there have been successive cuts in training expenditures. The investments in public
employment services have increased from a very low level but that have surpassed the
expenditures on training (Figure 4). This also happened in Sweden (cf. Figure 2).
De-corporatism and the introduction of ‘other actors’
In a discussion on the cuts in Swedish expenditures of ALMPs, Johannes Lindvall (2011: 4143) emphasizes two factors. Besides badly adapted policies to the massive demand created
by the economic recession in the 1990s,5 another main factor, is the retreat of corporatism in
1990 when the employers abandoned most social partnership institutions and all formal
tripartite cooperation (Minas 2011: 205). Research supports that there is a relationship
between the degree of corporatism and investments in ALMPs as training and similar
measures, as they presuppose active participation of the social partners. For example, the
increased Danish investments in ALMPs during the 1990s were part of wider reform
packages due to tripartite deliberations between the state and the social partners (Lindvall
2011: 42-43; see also Madsen 2010: 65). The social partners participated in both policy
formation and implementation of the state-run public employment service, as legitimate
representatives in boards and councils and as actors in regional and local implementation.
Corporatist structures were considerably strengthened, which has been seen as a key factor
in the Danish success story (Larsen 2004: 141, 151). However, during the rule of the liberalconservative government corporatism has severely weakened. In 2007, the state-run
employment service merged with local government labour market administrations into 91
jobcenters and formal corporatism was dismissed – regional and local labour market councils
transformed into local employment councils. Advisors and monitoring agents were the new
roles given to the social partners. Finally, in 2009, the municipalities became formally
responsible for labour market policy (Klindt 2009; Bredgaard 2011; Jørgensen and Schulze
2011). According to Thomas Bredgaard (2011: 768), one important motive for decentralizing
the operational responsibility for activation policies to the municipalities ‘was to reduce the
former influence of the social partners on policy implementation’ (cf. Jørgensen and Schulze
2011: 216). This means a transition from a proactive role to more of a reactive, controlling
role (Jørgensen 2008: 13).
Another element in a new mode of governance is an increased influence of market
mechanisms following the widespread use of both New Public Management instruments and
external service providers (Jørgensen 2009; Bredgaard 2011; Minas 2011). Public tendering
of private service providers on quasi-markets aims to create more efficient services and more
freedom of choice of services. This has been described in terms of a transition from a welfare
state to a welfare society where private providers have received a considerable role in
implementation of welfare services (cf. Hartman 2011). Contracting-out of public employment
services was part of the 2003 labour market reform in Denmark. Through the involvement of
5
The massive increase of participants in various ALMPs during the deep economic recession lowered
the quality. As measures to a very low degree resulted in employment they were increasingly
perceived as meaningless (Lindvall 2011: 41-43).
12
private, or ‘other actors’, as labeled by the government, the aim was to ‘create the conditions
for a large-scale market for employment services’ (Bredgaard 2011: 770). In Sweden, the
Public employment service has from 2007 onwards been assigned to purchase employment
services from external service providers, so-called ‘complementing actors’. Early on these
actors were mainly responsible for applicants to JOB and the Youth Job Programme.
However, at the end of 2009 there was a wide expansion of applicants to other services of
complementing actors, e.g. job coaching for new and short-term unemployed persons
(Swedish Public Employment Service 2011: 45-7). Following this, institutional change can
occur through a process of ‘differential growth’. This means that the faster growth of the
implementation of market-mechanisms and private solutions may shrink the public domain
and drain off political support for public sector services, which may possibly stagnate
(Streeck and Thelen 2005: 23-24).
A new policy landscape
Both in Sweden and in Denmark, the social democratic rule during the 1990s and early
2000s strengthened the pro-market employment orientation through the reinforcement of
work incentives. However, during the current Swedish centre-right government the ‘work first
approach’ has been increasingly emphasized. Major cuts in ALMPs and in economic
compensations, as well as tightened eligibility criteria, altogether work as activation policies
to create stronger work incentives. Following the activation turn in OECD countries since the
mid-1990s, a new role has been set for ALMPs. Instead of meeting the shortage of skilled
labour through upskilling, public policies are reoriented towards ‘employment assistance’
(e.g. job subsidies, counselling, job search programmes) and ‘incentive reinforcement’ (e.g.
tax credits, benefit reductions, benefit conditionality) (Bonoli 2012: 196). As well, in Sweden
and Denmark, this reorientation towards less costly forms of activation is evident (cf. de la
Porte and Jacobsson 2012: 130). Regarding the net compensation rate of the Swedish
unemployment benefit it has decreased towards a Danish rate, where the majority of fulltime
employed only has been offered a basic security (Ferrarini et al. 2012: 61). Reductions in
compensation rates from the social insurances act as incentive for complementing private
insurances. Tommy Ferrarini et al. (2012: 30, my translation) stress that the high levels of
income protection in Swedish social security systems ‘seems largely to be part of history’.
Besides these institutional changes to former comprehensive publicly funded social security
systems, in-work benefits and subsidised employment for low-wage service jobs are used to
reach the main policy goal – to ‘improve wage setting’ (Government Bill 2009/10:1, p. 38).
That is, reforms that will pressure the reservation wages downwards.
Today, the three pillars of the traditional Rehn-Meidner model have been abandoned – first,
through an increased employment protection legislation in the 1970s; second, through
decreased income security since the 1990s; third, through decreased investment in ALMPs
since the 1990s, specifically a major decrease in labour market training. Denmark does by
international standards invest extensive public resources in LMPs. For example, in 2009, of
total expenditures in LMPs, the Danish level was 80 percent higher than the Swedish level.
However, the increase in unemployment since the economic crisis, and sociopolitical reforms
during the new century, has challenged the basic principles of the Danish flexicurity model.6
First, stronger demands on greater job protection have been put forward. Second, stronger
6
In 2008-10 around 175.000 jobs were lost in Denmark (Bredgaard and Daemmrich 2012: 13).
13
demands on income security, as there have been a relative decline in unemployment
generosity since the 1990s. Third, declining employment security following changes in
ALMPs, from a human capital approach towards a stronger focus on the motivation effects of
active measures. Fourth, mutual trust between the social partners and the government has
declined (Bredgaard and Daemmrich 2012: 13-15).
If we focus on point three above, ‘learn-fare’ has been weakened. According to Klindt (2008:
30-31), the mobility political aspects that was at the core of the flexicurity model has
generally been eroded and replaced with an employment policy stressing that all should work
no matter the character of the job or the qualifications of people. This is in line with the British
work first approach that ‘any job is better than no job’ (Lindsey, Mcquaid and Dutton 2007:
541). As in Sweden, current governmental rationalities aim to mobilize additional labour
supply through policies of ‘work first’ and ‘making work pay’ (Jørgensen 2009; Bredgaard
2011). Decreasing unemployment and lack of qualified labour before the global financial
crisis made the work first policy ‘appear successful’ (Bredgaard 2011: 772-73). However, the
poor outcome of employment policies since the crisis has meant that the work first approach
has become unpopular (Bredgaard and Daemmrich 2012: 14).
From employability to standby-ability
Institutional change in the Nordic countries under study, and specifically with regards to
Sweden, can be seen as a development from employability to standby-ability. The Janus
faced character of activation policies has a counterpart in discourses on employability.
Activation through upskilling could be seen as being equivalent to discourses of employability
where the individual needs to show adaptability to changing requirements on the labour
market through learning (Garsten and Jacobsson 2004: 1). It is a preventive, social
investment approach – to invest ‘in people in order to equip the labour force with the
necessary skills to face change’ (de la Porte and Jacobsson 2012: 118). The other side of
activation policies focuses on individuals’ moral and social qualities. In discourses on
employability this means to have the ability to show ‘the “right” attitudes (initiative, flexibility,
availability)’ (Garsten and Jacobsson 2004: 8). In Scandinavia, employability has primarily
been seen as a collective responsibility – that is, risks have largely been socialized.
However, there has been an increased emphasis of activation in form of incentive
reinforcement and an individualization of risks. One can see this as a shift from an
intervention paradigm based on the provision on welfare to one that is based on the provision
of activation (Serrano Pascual 2007). In this latter paradigm, the location of the problem is
the individual’s behavior, motivation and attitudes. The status as a benefit recipient has more
and more turned into a moral contract – social citizenship rights should be deserved,
conditional on responsible behavior: ‘The result is a shift in the nature of the problem being
tackled: rather than being a fight against poverty, it is now, above all, a fight against (welfare)
dependency’ (Serrano Pascual 2007: 15).
Swedish and Danish activation policies have increasingly retreated from the approach that
Rik van Berkel and Iver Hornemann Møller (2002: 54) have called the ‘activation optimists’,
whereby ‘policies should assist people, when and where necessary, in developing forms of
participation’. A defensive workfare strategy (Torfing 1999) has increased in importance.
Bengtsson and Berglund (2012: 102) has described the development in terms of a ‘strategy
to discipline and develop a “standby-ability” among the unemployed to maintain motivation
14
for job searching, as skill enhancing education, training and human capital accumulation do
not longer seem to be prerogatives in state regulation of the Swedish labour market’. This
means a re-orientation from an offensive activation approach – from a social investment
approach in order to further ‘the security of the wings’ and the adaptability needed to face
labour market change – to stronger focus on individual persons’ moral and social qualities
through social discipline, as well as demands to work on their ‘soft skills’ via support from
employment services as job counseling and job coaching. Institutional change resulting in
both a new set of demands on persons in socially marginalized labour market positions and a
retreat from postwar principles of welfare and social citizenship is here conceptualized as the
rise of standby-ability – including stricter work incentives, contractualization of citizenships
rights, less generous unemployment benefits and less costly forms of activation – which
altogether further the development of Nordic residual workfare states. Consequently, we, as
social scientists, need to reevaluate existing typologies of western welfare states and market
economies and develop new conceptual tools in a continuing era of defensive activation
strategies.
References
Barbier, J.-C. (2004) ‘Systems of Social Protection in Europe. Two Contrasted Paths to Activation, and
Maybe a Third’. In J. Lind, H. Knudsen and H. Jørgensen (eds) Labour and Employment
Regulation in Europe. Brussels: P.I.E. Peter Lang.
Barbier, J.-C. (2005) ‘The European Employment Strategy, a Channel for Activating Social
Protection?’ In J. Zeitlin and P. Pochet (eds) The Open Method in Co-ordination in Action. The
European Employment and Social Inclusion Strategies. Brussels: P.I.E.-PETER LANG.
Bengtsson, M. and Berglund, T. (2012) ‘Labour Market Policies in Transition: From Social Engineering
to Standby-Ability’. In B. Larsson, M. Letell and H. Thörn (eds) Transformations of the Swedish
Welfare State: From Social Engineering to Governance?, pp. 86-103. Houndmills, Basingstoke:
Palgrave Macmillan.
Bonoli, G. (2009) ‘Varieties of Social Investment in Labour Market Policy’. In N. Morel, B. Palier and J.
Palme (eds) What Future for Social Investment? Stockholm: Institute for Futures Studies.
Bonoli, G. (2010) ‘The Political Economy of Active Labor-Market Policy’, Politics & Society 38(4): 43557.
Bonoli, G. (2012) ‘Active labour market policy and social investment: a changing relationship’. In N.
Morel, B. Palier and J. Palme (eds) Towards a Social Investment Welfare State. Ideas, policies and
challenges. Bristol: The Policy Press.
Bredgaard, T. (2011) ‘When the Government Governs: Closing Compliance Gaps in Danish
Employment Policies’, International Journal of Public Administration 34: 764-774.
Bredgaard, T. and Daemmrich, A. (2012) ‘The Welfare State as an Investment Strategy: Denmark’s
Flexicurity Policies’. Paper presented at the ILERA 2012 Congress in Philadelphia, USA, 2-5
August 2012.
de la Porte, C. and Jacobsson, K. (2012) ‘Social investment or recommodification? Assessing the
employment policies of the EU member states’. In N. Morel, B. Palier and J. Palme (eds) Towards
a Social Investment Welfare State? Ideas, Policies and Challenges. Bristol: The Policy Press.
Drøpping, J.A., Hvinden, B. & Vik, K. (1999) ‘Activation policies in the Nordic countries’. In M. Kautto
(ed.) Nordic Social Policy: Changing Welfare States. London: Routledge.
Esping-Andersen, G. (1990) The Three Worlds of Welfare Capitalism. Cambridge: Polity.
Ferrarini, T., Nelson, K., Sjöberg, O. and Palme, J. (2012) Sveriges socialförsäkringar i jämförande
perspektiv. En institutionell analys av sjuk-, arbetsskade- och arbetslöshetsförsäkringarna i 18
OECD-länder 1930 till 2010. Parlamentariska socialförsäkringsutredningen (S 2010:04).
Stockholm: Statens Offentliga Utredningar.
15
Furåker, B. (2003) ‘Patterns of unemployment’. In T.P. Boje and B. Furåker (eds) Post-industrial
Labour Markets. Profiles of North America and Scandinavia. London/New York: Routledge.
Garsten, C. and Jacobsson, K. (2004) ‘Learning To Be Employable: An Introduction’. In C. Garsten
and K. Jacobsson (eds) Learning To Be Employable. New Agendas on Work, Responsibility and
Learning in a Globalizing World. Houndmills, Basingstoke: Palgrave Macmillan.
Government Bill (Proposition) 1999/2000:98. Förnyad arbetsmarknadspolitik för delaktighet och
tillväxt. Stockholm: Näringsdepartementet.
Government Bill (Proposition) 2007/08:136. En reformerad sjukskrivningsprocess för ökad återgång i
arbete. Stockholm: Socialdepartementet.
Government Bill (Proposition) 2009/10:1. Budgetproposition för 2010. Utgiftsområde 14:
Arbetsmarknad och arbetsliv. Stockholm: Finansdepartementet.
Government Bill (Proposition) 2010/11:1. Budgetpropositionen för 2011. Förslag till statsbudget för
2011, finansplan och skattefrågor m.m. Stockholm: Finansdepartementet.
Hartman, L. (eds) (2011) Konkurrensens konsekvenser. Vad händer med svensk välfärd? Stockholm:
SNS.
Johansson, H. (2006) Svensk aktiveringspolitik i nordisk belysning. Stockholm: Finansdepartementet.
Johansson, H. and Hornemann Møller, I. (2009) ’Vad menar vi med aktivering?’ In H. Johansson and
I. Hornemann Møller (eds) Aktivering – arbetsmarknadspolitik och socialt arbete i förändring.
Malmö: Liber.
Junestav, M. (2011) ‘Promoting Employment or Employability? The Move from Active Labour Market
Policy to Workfare’. In A. Thörnquist and Å.-K. Engstrand (eds) Precarious Employment in
Perspective. Old and new challenges to working conditions in Sweden. Bruxelles/New York: P.I.E.
Peter Lang.
Jørgensen, H. (2008) ’Fra arbejdsmarkedspolitik til beskæftigelsepolitik – kosmetiske eller
indholdsmæssige forskelle?’, Tidsskrift for Arbejdsliv 10(3): 8-23.
Jørgensen, H. (2009) ‘From a Beautiful Swan to an Ugly Duckling: The Renewal of Danish Activation
Policy since 2003’, European Journal of Social Security 11(4): 337-67.
Jørgensen, H. and Schulze, M. (2011) ‘Leaving the Nordic Path? The Changing Role of Danish Trade
Unions in the Welfare Reform Process’, Social Policy & Administration 45(2): 206-19.
Kjellberg, A. (2012) Vilka ”hoppade av” a-kassan eller avstod från att gå med? En studie av akassornas medlemsras. Research Reports 2010:3. Uppdaterad 25 maj 2012. Lund: Department of
Sociology, Lund University.
Klindt, M.P. (2008) ’Arbejdsmarkedspolitik og institutionel komplementaritet’, Samfundsøkonomen 6
(December): 26-31.
Klindt, M.P. (2009) Institutionel komplementaritet mellem velfærdsstat og arbejdsmarked. Aalborg:
Institut for Økonomi, Politik og Forvaltning, Aalborg Universitet.
Larsen, F. (2004) ‘The Importance of Institutional Regimes for Active Labour Market Policies – The
Case of Denmark’, European Journal of Social Security 6(2): 137-53.
Lindsey, C., McQuaid, R.W. and Dutton, M. (2007) ‘New Approaches to Employability in the UK:
Combining ‘Human Capital Development’ and ‘Work First’ Strategies?’, Journal of Social Policy
36(4): 539-60.
Lindvall, J. (2011) ’Vad hände med den aktiva arbetsmarknadspolitiken?’, Ekonomisk Debatt 39(3):
38-45.
Lødemel, I. and Trickey, H. (2001) ‘A new contract for social assistance’. In I. Lødemel and H. Trickey
(eds) ‘An offer you can’t refuse’. Workfare in international perspective, pp. 1-39. Bristol: Policy.
Madsen Kongshøj, P. (2010) ‘Denmark’. In P. de Beer and T. Schils (eds) The Labour Market
Triangle. Employment Protection, Unemployment Compensation and Activation in Europe.
Cheltenham: Edward Elgar Publ.
Minas, R. (2011) ‘Decentralization and Back to Centralization: the Swedish case’. In R. van Berkel, W.
de Graaf and T. Sirovátka (eds) The Governance of Active Welfare States in Europe. Houndmills,
Basingstoke: Palgrave Macmillan.
OECD (2004) Employment Outlook 2004. Paris: OECD.
16
OECD (2012) ‘Labour market programmes: expenditure and participants’, in OECD Employment and
Labour Market Statistics (database), doi: 10.1787/data-00312-en, Accessed 10 February 2012.
Rehn, G. (1988) Full sysselsättning utan inflation. Skrifter i urval. Stockholm: Tidens förlag.
Ryner, J.M. (2002) Capitalist Restructuring, Globalisation and the Third Way: Lessons from the
Swedish model. London: Routledge.
Serrano Pascual, A. (2007) ‘Reshaping Welfare States: Activation Regimes in Europe’. In A. Serrano
Pascual and L. Magnusson (eds) Reshaping Welfare States and Activation Regimes in Europe.
Brussels: P.I.E. Peter Lang.
Sibbmark, K. (2010) Arbetsmarknadspolitisk översikt 2009. Rapport 2010:18. Uppsala: IFAU.
SO (2008): Historik över Arbetslöshetsförsäkringen från 1885 . Stockholm: Arbetslöshetskassornas
Samorganisation.
SOU 2011:11. Långtidsutredningen 2011. Huvudbetänkande. Stockholm: Fritzes.
Streeck, W. and Thelen, K. (2005) ‘Introduction: Institutional Change in Advanced Political
Economies’. In W. Streeck and K. Thelen (eds) Beyond Continuity. Institutional Change in
Advanced Political Economies. Oxford: Oxford University Press.
Swedish Public Employment Service (Arbetsförmedlingen) (2010) Arbetsmarknadsrapport 2010.
Stockholm: Arbetsförmedlingen.
Swedish Public Employment Service (Arbetsförmedlingen) (2011) Arbetsmarknadsrapport 2011.
Stockholm: Arbetsförmedlingen.
Swedish Public Employment Service (Arbetsförmedlingen) (2012) Arbetsförmedlingens Årsredovisning
2011. Stockholm: Arbetsförmedlingen.
Sörnsen, L. (2009) ’De olika världarna av aktivering – ett komparativt perspektiv’. In H. Johansson and
I. Hornemann Møller (eds) Aktivering – arbetsmarknadspolitik och socialt arbete i förändring.
Malmö: Liber.
Torfing, J. (1999) ‘Workfare With Welfare: Recent Reforms of the Danish Welfare State’, Journal of
European Social Policy 9(1): 5-28.
van Berkel, R. and Hornemann Møller, I. (2002) ‘The concept of activation’. In R. van Berkel and I.
Hornemann Møller (eds) Active Social Policies in the EU. Inclusion through participation? Bristol:
The Policy Press.
van den Berg, A. (2008) Flexicurity: Theory, Practice or Rhetoric? Paper presented at the 25th
Anniversary Conference of the Centre for Labour Market Research (CARMA) at Aalborg University,
October 9-10, 2008, Aalborg, Denmark.
van den Berg, A., Furåker, B. and Johansson, L. (1997) Labour Market Regimes and Patterns of
Flexibility. A Sweden – Canada Comparison. Lund: Arkiv.
Venn, D. (2012) ‘Eligibility Criteria for Unemployment Benefits: Quantitative Indicators for OECD and
EU countries’, OECD Social, Employment and Migration Working Papers, No. 131, OECD
Publishing.
17