Introduction

1Swedish labour market policy has for decades relied on activation programmes in order to facilitate structural changes in the economy. People made redundant in no-longer profitable sectors should, by the use of active labour market measures, be passed over to expanding sectors. The bulk here has traditionally consisted of labour market education and various forms of training. This was the dominant content of labour market policy in Sweden from the 1950s to the beginning of the 1990s – a period chiefly characterized by very low unemployment figures and high levels of labour market participation (Johannesson, 1991). When the deep economic recession and massive increase in unemployment hit Sweden in the early 1990s, the preconditions for national labour market policy were fundamentally altered. The extent of the economic downturn had been unmatched since the 1930s, and the development that followed was, in important respects, unfamiliar to a nation that had experienced almost full employment for more than half a century. Between 1990 and 1993, employment in Sweden declined by over half a million people, or around 13 per cent of the workforce, and unemployment on the open market rose from 1.7 to 8.3 per cent. During the decade, 1.8 million individuals (out of a population of approximately 9 million) experienced spells of unemployment varying in length (Korpi&Stenberg, 2001).

2As a result of development, the welfare of individuals, as well as the resources available to the welfare programmes, were affected negatively. Living conditions deteriorated for almost every group in society, although already disadvantaged groups like young adults, immigrants and single mothers were hit harder than others (Palme et al., 2002a). The downturn in employment also had a strong impact on public finances as it resulted in drastically eroded revenues as well as in a substantial increase in public expenditure. The rapidly accelerating budget deficit that followed was more or less unavoidable and in three years, between 1990 and 1993, the national budget went from a yearly surplus to a deficit that represented more than 13 per cent of the Swedish GDP (Bergmark & Palme, 2003).

3The recession at the beginning of the 1990s was widespread and affected most Western European countries. Yet Sweden was, in several respects, hit harder. Employment figures fell more dramatically than in other advanced industrial nations (with the exception of Finland) and employment problems persisted more or less throughout the entire decade. In Western Europe, many national governments developed policies that comprised an ambition to shift from passive to active labour market measures [1] – a path that followed recommendations formulated by the OECD and EU. Concurrently, many countries reduced compensation levels in unemployment benefits and social assistance, reduced the maximum length of benefit periods, tightened eligibility criteria and shifted the emphasis from non-means-tested to means-tested programmes (Hvinden et al., 2001). The explicit goal of these changes was not only to counteract unemployment and welfare dependency, but also to reduce social expenditure.
In this article, I shall describe how activation policies changed in Sweden during the 1990s as a response to the unemployment crisis, but also under the influence of international trends. Arguably, it is possible to depict the decade as a period of transition in aims as well as content of Swedish labour market policy. As to proportions, it is incorrect to speak of a move from passive to active measures. The transition essentially took place within the active labour market policy, where new forms emerged and responsibility for the measures was increasingly decentralized from state to municipal level.

The development of traditional labour market policy: growth, diversity and dilution

4From an international viewpoint, Sweden has, along with the other Scandinavian countries, served as an archetype of the modern or advanced welfare state (Esping-Andersen, 1985, 1996). The features normally emphasized are universality, comprehensive programmes, income redistribution, high taxation and state-based structures of finance, provision and regulation. The objective is to compensate for needs and inequalities that originate from vulnerable life-phases or from unfavourable positions on the market. The principle of universality implies that common goods are provided on the basis of need, regardless of age, income or place of residence. These principles are relevant to all parts of the Swedish welfare system, i.e. not only to income-maintenance programmes, but also to the social service sector (Anttonen & Sipilä, 1996).

5Another salient feature of Swedish welfare policy is extensive state intervention in the market sector designed to redirect production and sustain full employment. Where structural change in the labour market has been unavoidable, active measures have served to facilitate the transition of individuals from one sector to another. Activist and employment-oriented social policies have, for decades, been a central part of the Swedish system. The “workline” concept (sometimes referred to as “work option”) has been introduced to accentuate the emphasis on providing the unemployed with training and education rather than with passive financial hand-outs (the latter consequently referred to as the “benefit line”, Björklund et al., 1998). Another attribute of the Swedish workline is that eligibility and compensation levels in the social insurance system are closely linked to labour market participation, which means that groups holding a marginal position on the labour market also hold a less favourable position with respect to social security. [2] Those who do not match the prerequisites of regularity and permanence in work are also relegated to less generous and mostly means-tested systems such as social assistance. This departure from the otherwise universalistic principles of Swedish welfare provision is by no means accidental or a reflection of a situation where some areas are still in the process of development. On the contrary, these “dual welfare” patterns (Marklund & Svallfors, 1987) are in all essentials deliberate and, as such, are exponents of a number of moral precepts underlying Swedish social policy.

6The macroeconomic policy on which the Swedish welfare state was founded is often referred to as the Rehn-Meidner model. The model represented an integrated and coherent approach to three objectives: low inflation through restrictive monetary and fiscal policies, a “solidaristic” wage policy aimed at improving conditions for under compensated workers [3] and an “active” labour market policy (Milner & Wadensjö, 2001). Despite low unemployment during most of the post-war period, active labour market measures remained an important part of national policy and incorporated a wide variety of programmes.
When massive unemployment hit Sweden at the beginning of the 1990s, there were four activation programmes available: vocational training, temporary employment, recruitment support and youth measures – only one of which survived the entire decade. [4] The remaining three successively disappeared to be replaced by a number of other programmes and, at the end of the 1990s, the total range of options was considerably extended. Table 1 shows an overview of this development as well as the number of participants registered in the various programmes.

Table 1

Number of participants in active labour market programmes, in thousands, 1990-1999^a

Year Vocational Training Temporary Employment Measures for the Disabled Recruitment Support Youth Measures Training Substitutes Work Skill Development Computer Centres Workplace Introduction 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 38.6 58.6 86.3 53.2 59.5 54.6 45.6 36.9 41.9 45.0 8.1 10.8 15.8 14.0 16.7 14.5 8.3 7.1 1.1 - 87.7 86.0 84.8 84.0 87.2 92.0 85.9 85.2 89.6 89.5 2.3 4.8 13.5 9.2 25.5 21.0 12.3 3.7 -- 4.8 13.2 34.6 67.7 56.9 20.4 ---- --8.3 9.8 12.7 11.2 9.8 3.5 0.3 - ---35.1 44.5 41.3 52.3 52.5 38.9 8.0 -----5.9 11.9 14.0 11.5 7.3 -----21.5 32.4 34.3 19.4 2.3 Start-Your-Own Municipal Youth Programmes Temporary Public Employment (OTA) Individual Employment Support Support Work in Public Service Project Work with Un-employment Benefits IT under-takings Development Fund Guarantee Workplace training 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 ----10.2 10.1 10.3 12.3 12.5 10.8 -----1.7 12.6 14.2 12.2 7.3 -------5.4 7.8 7.8 --------10.6 9.7 --------4.1 2.9 --------0.3 0.6 --------0.7 3.0 --------2.8 5.6 ---------22.5 a The measures for the disabled include protected workplaces, salary contributions, labour market institute and other publicly protected employment. Youth measures include special and contracted introductory positions. As of 1 July 1995 Workplace Introduction replaced programmes for university graduates, immigrants and young people. Source: Regnér (2000) and the Labour Market Board (AMS) Statistics Service.

Number of participants in active labour market programmes, in thousands, 1990-1999^a

7Diversity increased over the entire period although many of the programmes introduced later were comparatively small. In terms of the number of participants, vocational training remains the largest, although it has declined proportionally. Vocational training was offered to people who needed to increase their competence in order to compete on the labour market or when they changed to a new occupation. Participants were entitled to an education grant with a compensation level that varied according to family composition and if the person was a member of an unemployment insurance fund. Economically, it was often more beneficial to be in vocational training than to attend more regular educational classes. Temporary employment, which was offered until 1997, gave unemployed persons work for six to twelve months in publicly funded activities. Participants were over 25 years of age and wages were set according to the level decided in collective agreements on the open labour market. In 1998, the programme was replaced by individual employment support. Recruitment support was a system of subsidized employment in the public and private sectors. The proportion of subsidies varied between 50 and 70 per cent of the total wage and employers could receive subsidies for six to twelve months, depending on the qualifications of the individual hired. Recruitment support was replaced by individual employment support in 1998.

8The first of the new programmes to be introduced during the 1990s were training subsidies, in 1992, which meant that unemployed individuals replaced people who were absent from their occupation for long periods for further education. The companies that hired people were allowed tax deductions for the replacement as well as for the expenses for the employees’ education. In 1998, this measure came to an end. In 1993, the work skills development programme started, which gave unemployed people entitled to unemployment benefits the chance to embark upon projects which were not part of a company’s regular activities. Projects were limited to six months and the predominant sectors in which they were conducted were repair and maintenance work, forestry and landscape architecture and informative work to various institutions (schools, organizations, etc.) (Arbetsmarknadsstyrelsen, 1999).

9Among the other new programmes, workplace introduction, launched in 1995, was one of the more important. In content, it was quite similar to work skill development, but one difference was that the company had to pay a certain sum to the authorities every month if it wanted to enjoy the contribution made by the participant. Other new programmes had a distinctly more educational profile, for example the computer centres that offered training in computer use and computer programming on a half-time basis and the IT schemes that offered advanced labour market education for IT occupations with a shortage of skilled labour. The start-your-own-programme was introduced in 1994 in order to give financial aid to people who wanted to start their own business. Temporary public employment (OTA) was launched in 1997 to give the long-term unemployed over 55 a chance to work in the public sector at the same time as receiving unemployment benefit.

10A salient feature of Swedish labour market policy is that participation in many of the programmes counts as regular work in assessing the work requirement (see Footnote above) when people apply for unemployment benefit. This is, for example, the case in vocational training, temporary employment, individual development and work skill development programmes. This represents the possibility of re-qualifying when the benefit period ends (normally after 14 months), which, in some cases, may result in virtually never-ending periods of compensation (Clasen et al., 2001). There are obvious incentives for Swedish municipalities to keep the unemployed on unemployment benefit instead of social assistance, since the latter is a municipal (financial) responsibility while the former is not. In a situation where the municipalities have increasingly gained influence over all activation measures (see below), this has created a situation where certain municipalities have been accused of abusing the system.

11The aggregate expenditure for the active labour market programmes (excluding those for the disabled) rose from just below 9,000 to more than 20,000 million SEK (Swedish kronor) [5] between 1990 and 1999, which is equal to an increase of 88 per cent in fixed prices. But since the number of participants increased even more, the resources spent on each individual in the programmes decreased. Hence, active labour market measures is one of many areas where the Swedish government managed to meet an increased demand by using dilution as an allocative strategy (see Bergmark, 1997). From a welfare point of view, the questions of the quality of programmes is, of course, crucial for the individuals enrolled. This regards both the long-term performance on the labour market and the benefits during the programmes. Here, the diminished resources provided in the middle of the decade can be seen as problematic.
The increased diversity in the measures offered meant that the unemployment agencies’ scope for elaborating individual solutions improved during the 1990s. On the other hand, the increased number of cases did put them under strong pressure. The proportion of the workforce that participated in the active labour market programmes increased from 1.2 per cent at the beginning of the decade to 5.3 per cent at the peak in 1996. In 1999, the figures were down to 3.1 per cent (Palme et al., 2002b). Despite these high volumes, there are very few systematic evaluations of what effects the interventions have had. However, one effect that has been proved in macro studies is that programmes which included subsidized work of any kind seem to have had displacement effects, in that they reduced the number of available jobs on the open market (Calmfors & Skedinger, 1995; Dahlberg & Forslund, 1999). Available results on a micro level, i.e. how a programme affects the likelihood of an individual getting a job, are disparate, but do, on the whole, show no or very modest (in some cases even negative) effects of participating in active measures (Regnér, 2000; Ackum Agell&Lundin, 2001). [6]
In absolute figures, there is no doubt that the decade represents a period of increased importance of activation measures. In relative terms, however, the pattern is less clear-cut. The dramatic development on the Swedish labour market did not only affect the magnitude of activation programmes, but no doubt also accentuated the importance of cash benefits. Figure 1 illustrates how the number of unemployed with active, passive and no labour market measures at all developed between 1990 and 1999. Two separate curves for active measures are presented: one with measures for the disabled included and one without. The reasons for this is that there is no consensus as to whether this programme should be defined as an active measure alongside the others and that its inclusion (due to its magnitude and, in this case, stability) makes a big difference in any assessment of relative changes.

Figure 1

Number with and without unemployment benefit and in active labour market programmes (measures for the disabled respectively included and not included) (index 1990 = 100)

Number with and without unemployment benefit and in active labour market programmes (measures for the disabled respectively included and not included) (index 1990 = 100)

12We can see here that, in relative terms, the number receiving unemployment benefit increased most markedly during the period in question and was over five times greater when it peaked in 1996-1997. The number lacking compensation altogether also displays a significant increase, but not as massive, which means that the relative proportion of unemployed without any kind of benefit actually decreased during the 1990s. Both curves for active programmes show growth, but the evolution where measures for the disabled are included is considerably less dramatic than the one where it is omitted. However, no matter which measure for active programmes we decide to consider relevant, the relative increase in passive labour market measures is bigger. This means that summing up the 1990s as a period principally characterized by an increase in active labour market measures in Sweden is not entirely correct. The fact that passive measures actually increased more has also been shown in previous comparisons of labour market expenditure (Regnér, ibid.; Hvinden et al., ibid.). [7]

The evolvement of workfare arrangements: conditionality and decentralization

13Parallel to the development of traditional active labour market policy, two other trends were visible during the 1990s in Sweden: an introduction of activation measures where eligibility for means-tested benefits was tied to participation in the programmes and local authorities’ increased responsibility for activation in that respect and in general. Although none of these trends were completely new, it was not until the 1990s that they gained some proportions and acquired a formal institutional setting.

14Like many other modern welfare states, Sweden developed policies that tied social assistance recipiency to work or to work-related activities (Lødemel & Trickey, 2001). Traditionally, Sweden, like other Scandinavian countries, has established “work-test” guidelines for those claiming unemployment benefit or social assistance. Normally, this requires being registered as unemployed at the unemployment office and demonstrating availability for work through intensive job-seeking activity, the character of jobs applied for and readiness for geographic mobility. Proving that one is “at the disposal of the labour market”, in this sense, has long served as a central eligibility criterion in provision of benefit and has functioned as a gateway to compensation. Refusal to accept offers of jobs that have been defined as suitable by the unemployment agency or a low level of activity in job-seeking might lead to suspension of or reduced compensation. In international comparisons, work conditionality in Sweden has been described as strict – both for social assistance and unemployment insurance schemes (Eardly et al., 1996; Clasen et al., 2001).

15At the beginning of the 1990s, the Swedish authorities (through the National Board of Health and Welfare) specified that “being at the disposal of the labour market” for social assistance claimants should include the obligation to a) seek full-time work, b) participate in national government activation programmes and c) take whatever work they were assigned. At that time, a number of local authorities had already started to develop models that aimed to increase job-seeking activities or require recipients to take part in work-training schemes or similar work-related activities if they wanted to remain eligible for social assistance (Johansson, 2000; Palme et al., 2002b). [8] The National Board of Health and Welfare reacted negatively to this and declared in its National Guidelines for 1992 that no-one should be denied social assistance on the grounds that he or she refused to accept employment (or employment-training), where wages were not agreement-linked or where normal labour market insurance was not provided (National Board of Health and Welfare, 1992). Meanwhile, local policies were moving in another direction, and, in the mid-1990s, the issue reached the Supreme Administrative Court, which delivered a ruling whereby social assistance recipients were required to participate not only in government activation programmes but also in programmes arranged by the municipalities.

16As the formal legal obstacles had been removed and with rapidly escalating social assistance costs, an increasing number of municipalities either launched activation programmes for social assistance recipients and/or intensified the demands for job-seeking. This was part of a general development where local authorities made explicit efforts to keep costs down and the trend, in most respects, was towards less generosity and tougher requirements. In a study from 1996, the National Board of Health and Welfare reported that 16 out of the 22 municipalities examined had imposed more stringent demands on what it meant to “be at the disposal of the labour market” between 1993 and 1995. These increased demands had, at that time, found a professional context in the so-called Uppsala model (named after the municipality where it was developed), which was based on an explicit goal to replace “passive financial hand-outs” with increased demands aiming at making the recipient “more responsible” for his or her situation – by requiring social assistance applicants to develop individual career plans in consultation with welfare clerks and to apply for a certain number of jobs each month (Bergmark, 2000).

17The linking of eligibility for social assistance to participation in activation measures has had a long tradition in the United States and is commonly referred to as “workfare” or “welfare to work” (Bane & Ellwood, 1994). Introduction of workfare-like programmes was a salient feature of most Western European labour market policy in the 1990s and the Swedish development was, in many respects, influenced by international examples. Here, Denmark unarguably served as the most important role model for Swedish decision-makers. Denmark has, like Sweden, long had a labour market policy characterized by strong work-conditionality (Lødemel, 1998). During the 1990s, this trait became more manifest and, in 1994, Danish legislation was changed in a manner that increased the possibility of enforcing participation in activation programmes for (especially young) people claiming social assistance (Abrahamson, 1998). Reports of decreasing numbers of unemployed and social assistance recipients in Denmark in the following years did unarguably have an impact on Swedish decision-makers (National Board of Health and Welfare, 1997) and, in 1998, the Swedish Social Services Act was revised in a manner that gave municipalities the option of making participation in activation programmes mandatory for social assistance recipients (and other long-term unemployed young adults) between 20 and 24 years of age. Required activities were to be offered within 90 days of the beginning of the unemployment period and should be “competence enhancing”, determined by the individual’s own needs and interests, and provide the client with an “improved prospect to become self-sufficient through employment”. No additional payment on top of the social assistance compensation was offered. If recipients refused to participate, their social assistance could be withheld or reduced. Young adults receiving social assistance were thereby given additional obligations and, in this respect, treated differently from other claimants. [9]
Since the financial and executive responsibility for social assistance in Sweden lies with the municipalities, and not with the state, the increased conditionality of social assistance did represent a decentralization of the active labour market policy in itself. Coupling “welfare-to-work” programmes with increased local responsibility was, however, not unique to Sweden but a trend that was visible in many European Union and OECD countries (Finn, 2000). The national government in Sweden voiced an explicit aim in the mid-1990s to increase the role of the municipalities and thereby make better use of local knowledge (Johansson, 2000). In 1995, the political responsibility for unemployed individuals under 20 shifted completely from the state to the municipalities, and activation measures were subsequently introduced within the so-called “Municipal Development Programme”. Other decentralizing measures were to include the municipalities on the boards of the local public employment services, which had traditionally been tripartite bodies with representatives from the state, the employers’ organization and the trade unions. In 1996, the municipalities were given the majority on these boards. After this, it became increasingly common for municipalities to establish special units with responsibility for labour market policy. By and by, the municipalities have also been given wider responsibility as providers of some of the activation programmes introduced in the later part of the decade (see Figure 1), for example, IT schemes.

Discussion

18Labour market policy has, for many decades, been a vital part of the Swedish welfare state. The macroeconomic objectives have been to promote economic growth alongside low unemployment and low inflation, while the welfare objectives have been to offer economic protection to the unemployed and to increase their chances of finding a job. These general goals are, in all essentials, still the same today, but the unemployment crisis of the 1990s and the influence of international trends have, in many respects, changed the content and the legal and institutional structures of the policy.

19To what extent these changes represent a shift in kind, or if they should be perceived as mere adaptations to new challenges, is, however, an open question. The debate on this issue has, in Sweden, as well as internationally, been very intense and polarized. Roughly speaking, one could say that most policy-makers have been in favour of active labour market measures in general and of welfare-to-work programmes – stressing them as means of empowering the poor and unemployed and helping them to attain autonomy and dignity through work. Scholars, on the other hand, have, in general, been more hesitant or negative – maybe less so in relation to traditional active labour measures (or the transition of them) and more so in relation to workfare arrangements.

20Beside the more explicit motives given for activation programmes as empowering and counterweights to passive financial hand-outs, there are a number of objectives that are either likely to be of significance for their existence or that have been ascribed to policy-makers by critics. The existence of hidden objectives is, in fact, a dominant theme in much of the criticism that has been directed towards the programmes. An ever-present argument in the Swedish debate – voiced mainly by the political opposition addressing the Social Democratic government – is that directing large numbers of unemployed people into activation is a way of concealing actual unemployment levels. The measure most referred to in the public debate in Sweden is open unemployment, where people in active labour market programmes are excluded. An explicit goal for the national government has, since the unemployment shock at the beginning of the 1990s, been to pressure down open unemployment below the four per cent level. This goal was reached in 2001, but with the unemployed in activation programmes included, the figure should have been somewhere between six and seven per cent of the workforce.

21A more prevailing theme in the academic debate deals with the moral precepts underlying the introduction of workfare programmes. In contrast to the formal policy discourse where activation is depicted as an empowering approach, it is construed as a coercive measure violating basic civic rights (Lødemel & Trickey, 2001; Van Oorschot, 2002). This incorporates a view where the true motives for the activation measures are not fully accounted for, and where there is an underlying concern for the moral standards of people and perceived needs for arrangements that prevent indolent or not truly needy citizens from receiving benefits. The concept of moral hazard is central to this notion. Moral hazard refers to a phenomenon arising when individual awareness of a forthcoming compensation in a needy situation make such situations more likely to occur or to prevail (Elster, 1992). Recipients of social assistance, for example, should possess a number of prerequisites in order not to be suspected of intentionally living on benefit (Midré, 1990). These are: good morals, limited personal resources and restricted possibilities on the labour market. “Good morals”, here, refers to the recipients’ genuine desire to cope without any help from society and implies an explicit intention of “not being a charge to the public”. “Limited resources” bears reference to individual capacity to reach and maintain a position where state support is not needed. Disability, advanced age or illness (i.e. well-established diagnoses) are traditionally legitimate resource problems, while repeated inability to keep a job or abstract personality-based problems are not. The possibility provided by external conditions (e.g. the labour market) principally set the standards for what might be regarded as inadequate resources.

22From this perspective, the introduction of welfare-to-work programmes are founded on a morally based standpoint in which the state (and the municipalities) underlines individual moral autonomy and some reciprocal traits of the welfare system. In the process of connecting eligibility for social assistance to activation programmes and work tests, one could, on the one hand, detect the common sense imperative that “those who don’t try hard enough have no right to enjoy the outcome of collective efforts”, and, on the other, see the inclination to identify individual capability. The basic idea is that society, in setting the standards for eligibility, should not compensate for low levels of ambition, but rather strive towards a system that is “ambition-sensitive” yet not “endowment-sensitive” (Dworkin, 1985). Against the rationale of a system based on universal rights as the principal way of addressing individual social and economic shortcomings, a picture is drawn of decreasing individual responsibility and people choosing welfare over work.

23It has been suggested that the motives for introducing workfare arrangements may vary according to the ideological foundation of the government and the strategies may be placed on a progressive (Social Democratic)-Conservative (laissez-faire Liberalism) axis (Torfing, 1998). The former, also the neo-statist approach, should mainly be found in Scandinavian countries, emphasizing positive incentives to seek employment rather than enhancement of work-morale. The latter, the neo-liberal approach associated with Anglo-American nations, relies more on sanctions and coercive measures. However, looking at the development in Sweden, this distinction seems to draw too sharp a line between the various motives and to deny the fact that elements of work-morale, sanctions and coercions have played a role in the rapid growth of welfare-to-work programmes. A stereotype image of Sweden as always being a proponent of the modern welfare state, or universalist values, denies the fact that political intention in these matters is very difficult to grasp and the fact that the increased influence of local political bodies (with a mix of Conservative and Socialist majorities) easily make such generalizations misleading.

24The movement towards reinforced reciprocity – meaning that social assistance is made conditional on the efforts of the recipients – is not an isolated Swedish phenomenon but rather a tendency experienced all over Europe (OECD, 1999). Stressing the motivational element also means an increased emphasis on individual personality traits, rather than formal competence or the mismatch between the demands of the labour market and the qualifications of the unemployed. In this respect, the 1990s introduced new elements in Swedish labour market policy, which make it reasonable to speak of a converging trend leading Sweden closer to the rest of Western Europe – and hence also closer to a US conservatist model (Gilbert, 2002).

25Since Sweden has long had a tradition of active labour market policy, it is inaccurate – looking at the broader picture – to speak of a transition from passive to active policies. Nor is it correct, given the fact that the number receiving cash benefits was greater than the number in activation measures during the 1990s, a pattern Sweden actually shared with a number of other European countries (Hvinden et al., ibid.). To speak of a transformation from passive to active policies could, however, be partly valid in relation to means-tested benefits where preconditions, especially for younger people, were actually altered in that direction.
Although the dividing line between traditional active labour market programmes and workfare arrangements may be arbitrary with respect to content and – at least some would argue – objectives, most scholars in Sweden and elsewhere treat them as representing totally different principles of welfare policy. This is also essentially correct if we limit the comparison to eligibility criteria and compensation levels. The active labour market measures nevertheless underwent a period of transformation during the 1990s, with an increased variety of programmes and a decrease in resources spent on each participant. Furthermore, we have seen how they changed character as they moved from being an instrument for matching the workforce to the demands of the labour market, to a general tool for fighting mass unemployment. They were also increasingly put under the local government domain, since the municipalities were included on the boards of the local public employment services. The general trend towards decentralization meant that national state control was eroded and that the aims of the programmes increasingly shifted towards the interests of the 289 Swedish municipalities. Hence, added to the diversity that followed the introduction of a number of new programmes during the 1990s, enhanced diversity can now also be expected from the fact that local political influence will create increased variations as to how the programmes are implemented.
A ongoing theme in the criticism against activation measures is that there is no clear-cut evidence that the programmes increased the likelihood of participants getting a job. [10] Information on the effects is rather scarce, and the few studies that have been conducted show disparate results. This is, of course, a disturbing fact, given that increased employment opportunities are ultimately what legitimize the existence of a large-scale active labour market policy and the introduction of potentially stigmatizing – and heavily youth-biased – workfare schemes. The more activation has become an integral element of national welfare policy and local income maintenance programmes, the more the question regarding the results come to the fore.