Activation in Scandinavian welfare policy: Denmark and Norway in a comparative perspective Open Access

As in most other Western European countries, employment or labour market programmes and policies were redesigned in Denmark and Norway in the early 1990s. This redesigning was referred to as ‘activation’ (Danish: aktivering) or ‘the work line’ (Norwegian: arbeidslinjen) in the political discourse and rhetoric. This activation ‘turn’ in policies has been analysed as a changing relationship between rights and obligations, and the re-balancing of rights and obligations has been interpreted as a measure used to discipline the recipients of welfare benefits (e.g. Torfing 1999; Kildal 2001,; Kvist 2002; Oorshot and Abrahamson 2003). For instance, Torfing (1999, 15) argues that a reduction in the duration of unemployment benefits in Denmark in 1993 from 9 to 7 years was an ‘aggressive’ attempt to get people to return to the labour market. As such, activation in the Nordic countries has been interpreted as a new qualitative path formation, epitomised as a decline in citizenship.

The relationship between rights and obligations has surely changed in Denmark and Norway in the 1990s; however, a decline in social rights does not necessarily have any practical impact or effects. This may be the case if a decrease in the duration of unemployment benefits is preconditioned by a decrease in the problem pressure (size and structure of unemployment). In Denmark, for instance, between 1993 and 1999, the duration of benefits fell from 9 to 4 years, while overall unemployment fell from 9.6 to 4.3 per cent between 1993 and 2001. That is, a decline in the duration of benefits may have had only limited consequences, because there were fewer unemployed people and the labour market was more open to them. At least, the effects of a decline in citizenship (e.g. duration of benefits) are difficult to assess and certainly complex and potentially ambiguous. Therefore, if we are to move beyond moral judgements about activation, we ought to cease in analysing the changing relationship between rights and obligations independently of the historical and social conditions that constitute them in their specificity at a given moment in time. Accordingly, the aim of this article is, first, to analyse how changes in the programmatic structure of welfare policies are connected to changes in the problem pressure; and second, to shed light on the implementation and effects of the activation policies in the two Nordic social democratic welfare states (Esping-Andersen 1990).

In order to avoid the pitfalls arising from a narrow focus on rights and obligations in one or another welfare program, we shall employ a broad analytical approach in our analysis of activation in Denmark and Norway. As an analytical notion, activation may refer to the interconnectedness between social protection, labour market participation and labour market programmes (cf. Barbier and Ludwig-Mayerhofer 2004). Or to put it differently, activation may be analysed as to how interlinkages between welfare policies and employment/labour market programmes enhance labour force participation (Barbier 2004). This notion of activation has clear connotations to the concept of the ‘active society’ (e.g. Walters 1997), as an active society strives to render all citizens workers. In an active society inclusion, integration and full citizenship are achieved through labour market participation.

As a consequence, a central aim in an active society is to cultivate the labour force. Two options are open: first, to stimulate flows into the labour force, for instance by recruiting new segments, e.g. women, into the labour market. Policies stimulating flows into the labour market include relatively extensive welfare programmes, e.g. taxation policies, provision of day and elderly care institutions, etc. Second, to counteract flows out of the labour market by means of policies aiming to retain workers in the workforce. Such policies could be active or compensatory. Active policies include a wide range of active labour market policies (ALMP), such as support to occupational and geographical mobility, job-creation programmes such as wage subsidisation and compensation schemes for employers, support schemes for distressed firms, regional policies and so on. The compensatory mode refers to generous income maintenance schemes that potentially render it more attractive to remain oriented towards the labour market and define oneself as unemployed and, in this sense, remain within the labour force rather than leaving it.

For decades, a central aim in Norway and Denmark has been to have most segments of the population in paid employment. As such, the ideal of an active society has been a central hallmark and a common bottom line for policies and policy development since the late 1950s/early 1960s. In other words, an active society has been promoted for more than half a century in the Nordic countries. On this background, the activation ‘turn’ in the early 1990s can hardly be regarded as a path-breaking innovation; rather, it must be seen as an adjustment, i.e. a ‘first-order’ change (Hall 1993: 278) within the framework of an active society, which has existed for a considerable period of time in the Nordic countries. Thus, new expressions emerged in the political rhetoric, i.e. aktivering and arbeidslinjen, but the aims and means of welfare policies remained more or less unchanged. In other words, the reforms in the 1990s must be understood on the background of the previous tradition of an active society.

As indicated, activation has – in the analytical sense – deep historical roots in the Nordic countries. Nevertheless, social protection and labour market programmes have continuously been modified in accordance with new challenges and opportunities. In particular, changes in the problem pressure have contributed to the restructuring of welfare policies. Accordingly, this article is sub-divided into three sections. In ⁰⁰⁰² we discuss the formative period of the active society between the late 1950s to the mid 1970s. The formative period was characterised by full employment and attempts to stimulate flows into the labour force. The period extending from the mid 1970s to the early/mid 1990s was marked by high and enduring unemployment, and even more so in Denmark than in Norway. In this period, there was a predominance of policies that sought to retain workers in the workforce, primarily by means of compensatory policies. This is discussed in the third section. As will be shown in ⁰⁰⁰⁴, unemployment fell as of the mid 1990s in Denmark; i.e. when ‘activation’ became a part of the political rhetoric. A near-full employment situation was attained by 2000 and activation policies were adjusted accordingly. Policies aimed at retaining workers in the workforce thus changed from being primarily compensatory to being primarily active.

In general, activation policies since the early 1960s have attempted to include all segments in the labour force; i.e. the Nordic countries have historically followed a path of inclusion through participation (e.g. Esping-Andersen and Korpi 1987; Berkel and Møller 2002). Policies stimulating flows out of the labour market by means of, e.g. early retirement schemes have, however, also been compatible with the active society in the Nordic countries. Especially during times of high and enduring unemployment, it has been considered cost-effective to stimulate flows out of the labour market. This is partly because a person who has left the labour market, whether temporarily or permanently, will receive an inferior welfare benefit as compared with those who remain available to the labour market. This is also partly because it has been assumed, for instance, that early retirement schemes contribute to the rejuvenation of the labour force, which renders the labour force more efficient and active. In other words, policies stimulating flows out of the labour market may ultimately contribute to the active society being sustainable in the long-term in the face of heavy problem pressure.

Before turning to the discussion, however, we shall briefly mention two factors that have functioned as a framework or preconditions for Scandinavian activation policies:

First, the active society ideal has been supported by a vast majority of the population. For instance, the proportion of the population that in 1997 could agree with the statement ‘I would enjoy having a job even if I did not need the money’ was about 81 per cent in Denmark, 74 per cent in Norway, 73 per cent in Germany (East), 55.8 per cent in Great Britain, 56 per cent in Italy, 52 per cent in France and 43 per cent in Slovenia (ISSP 1997). A very strong work ethic and work-orientation can thus be found in Scandinavia. Additionally, the commitment to employment among women is equal to men in Scandinavia (Svallfors et al. 2001).

Second, the state has historically played a central role in regulating the life of the individual citizen. The role of the state has rarely ever been questioned; rather, it has been taken for granted that the state ought to provide for its citizens. The state has been conceived of as a ‘fellow player’ rather than an ‘adversary’. Thus, in Scandinavia there is a high degree of trust in public institutions (Rothstein and Stolle 2003). These high levels of trust most likely derive from the fact that the state–citizen relationship is structured as a reciprocal relationship. For instance, rights and obligation to activation also mean that the bodies of public authority are expected to provide an activation offer.

In Denmark and Norway, the ‘active society’ emerged in the late 1950s and early 1960s as a response to the new challenges of full employment. Between the late 1950s and early 1970s, Denmark and Norway experienced a long period of near-full employment, which helped to bring into focus insufficiencies with respect to the size and composition of the workforce. In response, the two Scandinavian countries comprehensively changed the ‘tax, benefit and service’ systems to promote an increase in the national labour supply.

To stimulate the flows into the labour force, a wide range of policies were employed. For example, to increase the female labour supply, both countries formally moved from joint household taxation to the introduction of individual taxation, and both countries sought to introduce equal treatment of men and women in terms of wages and working conditions. Such legislation served abstract ideas of equality and equal opportunities for men and women. Moreover, it also served down-to-earth labour-market purposes. Individual taxation and high relative wages contributed to the creation of an incentive structure rendering it economically advantageous to increase the labour supply from families. Thus, in economic theory, for instance, women's relative wages are considered the most important parameter for the decision of women to enter the labour market (e.g. Blau and Ferber 1986).

As of the late 1950s women also became the target for policies aimed at retaining workers in the workforce. Qualitative changes in the maternity leave schemes in the 1960s and 1970s functioned as an economic and symbolic incentive encouraging mothers to remain in the labour market (e.g. Leira 1992). Childcare institutions were another measure aimed at overcoming the contradiction between the obligations associated with motherhood and participation in the labour market. The formation of childcare institutions, however, grew much faster in Denmark than in Norway (Blom-Hansen 1998). On the other hand, the maternity leave scheme in Norway has been more generous than in Denmark. Arguably the maternity leave scheme in Norway and the childcare institutions in Denmark could be interpreted as ‘functional alternatives’ (Merton 1968) in terms of labour supply policies. Both measures have served to retain women with children in the labour market.

Older workers and workers who had suffered from diminished working capacity were also encouraged to remain in the labour market. Both countries sought to institutionalise a relatively high pensionable age. Moreover, earnings-related supplementary pensions were introduced in the two countries in the 1960s. The design of these supplementary pensions in Scandinavia made it economically advantageous to postpone retirement. In the late 1950s and the 1960s, both countries introduced policies designed to retain burned-out workers and workers who had become disabled in the labour market by means of retraining, support for occupational mobility and/or medical assistance (Hvinden 1994; Jensen 2003). The purpose was to prevent people who had not fully lost their working capacity from withdrawing from the labour market.

Despite the similarities mentioned above, Denmark and Norway exhibited different policy packages to some extent. While income compensation and supply-side oriented ALMP played a major role in Denmark, industrial and regional policies played a more prominent role in Norway.

In Denmark, ALMP have become an integrated part of the general political economy since the early 1960s (Kirstein 1974). Danish labour market policies have had two components: income compensation and active labour market policies. By means of educational and geographical mobility, the purpose of ALMP has been, first, to adjust the qualifications and locations of the workforce to the needs and demands of employers, and second, to counteract individual workers being forced to leave the labour market due to outdated skills, wrong location and so on (Jørgensen 2002: 171ff). The underlying purpose of ALMP has been to improve the quality and size of the labour force. For this purpose ALMP have interacted with compensatory policies, as these policies serve several purposes: first, they improve the living conditions of the unemployed sectors of the population, and thus to counteract poverty, inequality and so on; second, they improve the functioning of labour markets, since the recipients of benefits must meet the programme requirement that they are available for and actively seeking work; and third, they cultivate the size of the labour force, as generous benefits make it attractive for unemployed persons to remain in the labour market rather than leave it. Between 1960 and 1973 the replacement rate for unemployment benefits rose from 45 to 75 per cent (Lind 1992).

As to income compensation, Norway experienced a somewhat different development. Overall unemployment was exceptionally low. In effect, as of the late 1950s eligibility for unemployment benefits was tightened substantially (White paper 1958). As in Denmark, the state provided assistance for geographical mobility and vocational training in Norway, but the measures were relatively modest. Rather than providing generous welfare benefits combined with active labour market policies, Norway committed itself to full employment by means of regional policies and state funding to industries in financial difficulties (Therborn 1986). The state utilised credit policy to channel resources into priority sectors (Mjøset 1993). These types of measures continued to increase during the 1960s and 1970s and were associated with a stronger state control of the economy compared to Denmark.

The policies in the 1960s and 1970s served to change the practices and identities of citizens in the two Nordic countries. Between 1960 and 1974, the total labour force as a percentage of the population between 15 and 64 years of age grew from 71.2 to 76.6 in Denmark, and from 64.3 to 68.6 in Norway. In comparison, the labour force decreased on an average from 69.8 to 67.4 in the European OECD countries. At the same time female labour force participation grew from 43.5 to 63.2 in Denmark, and from 36.3 to 50.0 in Norway. In comparison, it grew on average from 44.3 to 46.4 in the OECD countries in Europe (OECD 1995). Overall, these changes in the Scandinavian countries had a huge impact on society as a whole. Women's informal work was formalised, epitomised as a development from a ‘traditional male breadwinner/female carer model’ to a ‘dual breadwinner/state carer model’ (cf. Pfau-Effinger 1998).

Another policy effect was that while regional policies in Norway promoted the formation of stable and regional labour markets, the ALMP in Denmark promoted a highly flexible national labour market. Brüniche-Olsen (1987), for instance, has shown that out of a labour force of 2.5 million, over 1 million persons changed jobs one or more times during the period from November 1980 to 1981 in Denmark (Brüniche-Olsen 1987: 323). Compensatory policies such as unemployment benefits supported the increased flexibility of the labour market, as compensatory policies were highly marked by mechanisms to reward affirmative labour market behaviour, and to punish inadequate behaviour. In effect, policies in Denmark were criticised for resembling ‘work-fare’. At least it was argued that policies treated the unemployed arbitrarily or as though they were ‘outlaws’ (Posborg and Scocozza 1974).

Unemployment began to rise in the mid 1970s, especially in Denmark as compared to Norway. However, the rise in unemployment did not lead to the abandonment of the ideal of an active society in Denmark; rather, the contrary is true. Between 1974 and 1990, labour force participation grew from 76.6 to 84.1 per cent in Denmark and from 68.6 to 78.0 per cent in Norway, while labour force participation in OECD Europe fell from 67.4 to 67.2 per cent (OECD 1995). To a large extent, the growth in labour force participation rates in the two Nordic countries in the 1970s and 1980s was the outcome of a continuation of and putting into force the policies developed in the formative period of the active society. Nonetheless, to counteract social exclusion resulting from a rise in unemployment, new policies were developed alongside the active society paradigm. First, compensatory policies were expanded, for instance, by softening the eligibility and duration of unemployment benefits. Second, job creation programmes and the like were developed and expanded. Third, new policies were established that stimulated flows out of the labour market. These policies targeted older and disabled workers, i.e. those who were perceived to be the ‘weakest’ segments of the work force. Between the mid 1970s and the mid 1990s, policies stimulating flows out of the labour market were more pronounced in Denmark than in Norway.

When unemployment in Denmark began to rise in the mid 1970s, politicians in Denmark were convinced that unemployment would merely represent a temporary phenomenon. Therefore, they did not hesitate to establish high quality compensatory policies. By the mid 1970s, the eligibility criterion for unemployment benefits was 26 weeks of work within the last 3 years, and the replacement rate was 90 per cent of previous earnings, though with a ceiling. The duration of the benefits was 1 year. However, under the impression of high levels of unemployment, the politicians decided to prolong the duration of benefits to 2.5 years from the mid 1970s. The aim of this prolongation was to ensure that the unemployed did not forfeit their rights to unemployment benefits.

As unemployment in the late 1970s proved to be more persistent than initially presumed, it became a priority in Denmark to reintegrate unemployed persons into employment, first and foremost by means of job creation programmes. A law was passed in 1977 obliging municipalities and counties to earmark a certain amount of money per inhabitant to prevent or inhibit unemployment among welfare benefit recipients. These measures included subsidised employment in the private sector, training courses, education, etc. Until 1994, participation in municipal employment programmes was regarded as ‘work’ with respect to the qualifying conditions for unemployment insurance. The welfare beneficiaries could thus qualify for unemployment benefits by participating in municipal job creation programmes.

In 1978, a ‘job-offer’ or job-training scheme was introduced. Long-term unemployed at risk of losing their entitlement to unemployment benefits were offered subsidised jobs. The salary in the ‘job-offer’ scheme was equivalent to wages set by collective bargaining agreements between employers and the trade unions, and the duration of the job was 7–9 months. Participation in a ‘job-offer’ scheme was thus an offer of at least 26 weeks of employment, which in turn allowed the unemployed person to re-establish eligibility for unemployment benefits. A major purpose of the job-offer scheme was thus to prevent unemployed persons from becoming disconnected from the unemployment benefit system. The job-offer scheme provided opportunity to endlessly ‘commute’ between unemployment benefits and subsidised employment. By the end of the 1980s, the number of job offers that could be made to the same individual was reduced to two, i.e. the effective maximum duration of unemployment benefits in the early 1990s was roughly 9 years. According to the law, the job-offer was a right – not an obligation. Nonetheless, the ‘right’ was indisputably a ‘silent obligation’. If a person in risk of losing his/her right to unemployment benefits did not make use of this right to a job-offer, he/she would forfeit his/her right to unemployment benefits.

As it became increasingly clear that unemployment did not represent a temporary phenomenon, policies stimulating flows out of the labour market were introduced in Denmark from the late 1970s. In 1979, for example, an early retirement scheme was established (efterløn). This scheme allowed those members of an unemployment insurance fund aged 60–66 years of age to freely choose to retire from the labour market prior to the ordinary retirement age (Jensen 2004). Subsequently, the disability scheme was reformed in 1984 (Jensen 2003). Under the impression of high and enduring unemployment, a major reform made it possible to award a disability pension (førtidspension) for non-medical reasons to people over 50 years of age. That is, for instance, if their prospects on the labour market appeared particularly bad.

In Norway, unemployment remained much lower compared to Denmark. At the beginning of the 1980s, the maximum duration of unemployment benefits was 40 weeks within 1 year; however, when unemployment increased in the 1980s, more unemployed people came at risk to forfeit their rights to unemployment benefits. The Conservative minority government (1981–86) was therefore pushed to increase the duration of the allowance to 80 weeks during a 2-year period. In other words, unemployment benefits became more of a long-term entitlement during the 1980s. Entitlement to unemployment benefits did not have a time limit. When an unemployed person had received unemployment benefits for 80 weeks; however, a so-called waiting period was in force. The waiting period had a duration of 12 weeks, during which time the claimant would have to rely on other types of welfare benefits. To increase the incentives to seek and accept paid work, however, the Conservative government insisted on increasing the waiting period from 12 to 26 weeks. At the same time the Norwegian Parliament decided to lower the ceiling for unemployment benefits (White paper 1983–84). In other words, concerns about risks of social exclusion were balanced against the concern to provide incentives to seek gainful employment. The Social Democratic government later reduced the waiting period again to 13 weeks in 1991.

While the compensation level of welfare benefits for the unemployed was lowered, new schemes were introduced for the disabled. For instance, in 1981 a supplementary pension was introduced for people born with disabilities or who had become impaired before they had the opportunity to attain paid employment. All in all, this emerges as a period with the expansion of social rights and increases in public expenditures for income maintenance schemes in Norway. As opposed to Denmark, the official focus was more on retaining workers in the labour force. At the same time, the actual implementation of the policy came to provide more attention to income maintenance than to transferring or retaining workers in the labour market. In other words, some of the original policy objectives were pushed into the background or forgotten in the implementation process during this period of time (Hvinden 1994).

During the 1970s and 1980s, Denmark and Norway succeeded in increasing overall labour force participation rate in the face of high and enduring unemployment (OECD 2002: Table 19), despite the fact that new types of policies were established with the aim of stimulating flows out of the labour market. By 1990, the total labour force in Denmark was 2.9 million persons, while 259,000 received a disability pension and 93,000 early retirement benefits (Danmarks Statistik 1998). In Norway, the total labour force was 2.1 million people, while 234,000 received a disability pension (SSB 1991: Table 134, 161). In 1990 the standardised unemployment rate was 7.2 per cent in Denmark and 5.3 per cent in Norway (OECD 2002: Table A). In both countries the early retirement benefit has been relatively generous. In effect, relatively few early retired persons experience economic hardship (Saurama 2005, forthcoming). Nevertheless, in Denmark and Norway the risk of poverty and social exclusion among the early retired is higher among unmarried as opposed to married retirees.

In the early 1990s, the ‘active society’ took a discursive turn in the Nordic countries. ‘Activation’ (aktivering) and ‘the work line’ (arbeidslinjen) emerged as catch phrases in the political debate in terms of how to combat unemployment in Denmark and Norway, respectively. These new catch phrases, however, emerged under quite different circumstances in the two countries. The problem pressure in Norway was much lower as compared to Denmark. In Norway, the standardised unemployment rate declined from 6.5 per cent in 1993 to 3.6 per cent in 2001. By contrast, unemployment grew dramatically in Denmark between the late 1980s and into the early 1990s. By the mid 1990s, however, unemployment started to fall. Between 1993 and 2001, the standardised unemployment rate in Denmark fell from 9.6 to 4.3 per cent (OECD 2002: Table A, 2004). In accordance with these differences in the problem pressure, Denmark and Norway underwent somewhat different policy developments during the 1990s. While policies in Norway were marked by coherence and continuity, policies in Denmark were continuously adjusted in accordance with fluctuations in the problem pressure.

In Denmark, there is no consensus as to what the term activation (aktivering) actually refers to. It is observable, however, that the term aktivering functioned as a common banner for three quite different discourses in the early/mid 1990s. First, a ‘combat unemployment’ discourse, part of the campaign that brought the Social Democratic Party into power in January 1993, when they promised that they would fight unemployment. Second, a ‘structural unemployment discourse’ emphasising that the labour market suffered from structural imbalances. Third, a discourse on care for and reintegration of those who were totally excluded from the labour market due to abuse of alcohol, etc. In the ‘care discourse’ leading Social Democrats argued that welfare benefits may be characterised as ‘shut up’ money (i.e. we give you some money and in return you don't make any demands or bother us). Each of the three discourses had an impact on the design and effects of activation policies.

Immediately after assuming office, the Social Democrats translated the ‘combat unemployment’ discourse into a strengthening of policies stimulating flows out of the labour market. Those who did not fit in an active society were given the opportunity to leave the labour market permanently by means of early retirement schemes. In order to reduce official unemployment further, new schemes were established that encouraged all members of the work force to leave the labour market temporarily by means of the so-called leave-of-absence schemes. The leave of absence schemes were designed in such a way that they should have a positive effect on the efficiency of the workforce as a whole. In contrast to policies stimulating flows out of the labour market, a common feature of the ‘care’ and the ‘structural unemployment’ discourses is that activation alludes to a process in which people on unemployment and welfare benefits are transferred into education, job-training or ordinary employment. In effect, labour market policies – active as well as compensatory – were strengthened as of the early 1990s. As mentioned, however, policies were continuously adjusted to changes in the problem pressure. So when unemployment started to fall in the mid 1990s, policies stimulating flows out of the labour market were phased out, while ‘ordinary’ labour market policies became the predominant policy measure.

As mentioned, after assuming office, one of the first acts of the Social Democratic Government was to expand the arsenal of policies stimulating flows out of the labour market. Already in 1992 the so-called transition allowance scheme (overgangsydelse) extended the rights to early retirement for long-term unemployed aged 55–59. That is, the older long-term unemployed, who did not have a strong orientation towards the labour market, could freely choose to leave the labour force. In 1994 the government further extended the early retirement opportunities. As of 1994, the long-term unemployed aged 50–59 were eligible for transition allowance benefits. The transition allowance was cost-effective. The allowance amounted to 80–82 per cent of the unemployment benefits.

Similarly, as of January 1994, new legislation came into effect and expanded the leave-of-absence schemes introduced in 1992 (Jensen 2002). The 1994 act encouraged workers, and even the unemployed, to take educational leave, sabbatical leave and childcare leave. Persons on educational leave received an allowance corresponding to the level of unemployment benefits, while those on sabbatical or childcare leave would be compensated at a rate of 80 per cent of unemployment benefits. The general purpose of the leave-of-absence schemes was to reduce unemployment by means of job-rotation and work sharing in the labour market, i.e. to redistribute the burdens of unemployment, hoping that vacancies generated by the schemes would be filled with unemployed people. Another purpose was to improve the qualifications of the work force. This especially applied to the scheme for educational leave that would improve opportunities for life-long learning. However, the schemes for sabbatical and childcare leave also include an educational dimension. That is, in as much as the schemes create job-openings, they will render it possible for unemployed people to improve their skills by means of ‘on-the-job-training’ in ordinary jobs. In effect, these people will become more employable. Finally, the purpose was, at the individual and family levels, to improve the welfare of the individual by relieving the time pressure on families with small children (Andersen and Larsen 1995: 83).

As of the mid 1990s, unemployment began to decline. It was now argued that the liberal opportunities for leaving the labour market were counterproductive for the overall economic performance of the nation. It was made less attractive to take leave in 1995 and 1996, and as of 1999, the sabbatical and educational leave schemes were phased out in order to increase the labour supply. Accordingly, the early retirement opportunities were restricted. From January 1996 the transition-allowance scheme was abolished. In 1999 it was made favourable to postpone early retirement. If a person does not make use of his/her rights to early retirement, he/she will be economically rewarded. Contributions and benefits taken into account, it has been calculated that the tax exemption award will provide a net gain of DKK 45,000 (about €6,000) for those postponing retirement until the age of 65 (Jensen 2004). Finally, in 2002, the childcare leave scheme was abandoned in return for an extension made to the maternity leave scheme to 52 weeks.

The disability benefit scheme was also tightened. In 2000 the major political parties in the Danish parliament agreed to reform the disability pension system. It was more strongly emphasised that a disability pension should only be awarded if it was impossible to remain in or re-enter the labour market on normal conditions or in subsidised work. ‘Working capacity’ therefore replaced the old criteria of ‘loss of employability’ when considering whether to grant disability pension. The underlying idea is that by focussing on working capacity, attention will be directed towards the resources and development potentials of citizens rather than toward their problems and limitations. Throughout 2001, all social workers in Denmark have undergone training and re-education in order to become familiar with the new concept and its application. The new direction in the handling of disability problems was triggered in the mid 1990s, when the Ministry of Social Affairs launched a campaign – ‘It concerns all of us’ – in an attempt to create an all-inclusive labour market.

The notion of an all-inclusive labour market emerged when unemployment started to decline in the mid 1990s. Simultaneously, rights and obligations in relation to unemployment benefits were changed in several respects: (1) between 1993 and 1999 the maximum duration of benefits was cut from 9 to 4 years; (2) the conditions for qualification were tightened, e.g. it is no longer possible to extend one's rights to benefits by participating in labour market policy programs; (3) the type of job an unemployed person could reasonably be expected to accept was tightened, especially with respect to the distance of a reasonable commute and the duty to accept a job outside his or her prior occupation; (4) activation (the ‘job-offer’) should no longer necessarily be in accordance with collective agreements – a change that was backed by the trade union movement; (5) the right (and obligation) to a job or an educational opportunity was advanced, i.e. the unemployed became eligible for activation after 1 instead of 2.5 years of unemployment. Nonetheless, in Denmark activation is not a ‘work first’ strategy. In 2002, for instance, among those enrolled in activation, about 47 per cent were subject to some manner of job training, while 53 per cent were enrolled in educational measures.

The right and obligation to activation was already introduced in 1990 with regard to 18–19-year-old claimants of welfare benefits. The age limit was subsequently extended between 1993 and 1998. From 1998 onwards, all welfare claimants have the right and obligation to participate in activation measures. Contrary to the activation of beneficiaries of unemployment benefits, the aim of activation of welfare beneficiaries may be to stabilise and improve the general life situation of the individual in the event that employment or education is not a realistic goal. Activities that help to organise the time structure of the beneficiaries in a similar vein as the time structure of wage earners are regarded to be a virtue unto itself. The Ministry of Social Affairs has thus stated that activation should be considered a goal in itself. Activation provides for ‘people entering into meaningful (work) communities … i.e. participation in communities in this context is a goal in itself, as it is regarded as being good for the individual – even if it fails to lead to self-maintenance’ (Socialministeriet 2000, own translation). Only about 14 per cent of all activated beneficiaries of welfare benefits are enrolled in educational measures (Beskæftigelsesministeriet 2002). This is due to the fact that the municipalities prioritise social integration over the improvement of qualifications and integration into the labour market (Larsen et al. 2001).

Metaphorically, one may argue that activation (aktivering) in Denmark refers to an ‘active society’ and ‘active citizenship’ being fully institutionalised by the turn of the century. In any case, it has become increasingly clear that enrolment in the activation machine is considered to be the best alternative to social exclusion.

In Norway, activation was from the outset attributed a narrower meaning as compared to the Danish experience. An underlying concern has been that a liberalisation of the policy or the opening up for secondary objectives for those with lesser chances of achieving gainful employment would weaken the political priority of maximum participation in the labour market. The emphasis has been on the curtailment of the eligibility rules, requirements concerning participation in activation measures were reinforced, and the conditions for continued receipt were made stricter. Overall spending on ALMP was fairly stable during the 1990s, though Norway spent a smaller percentage of GDP on supply-side oriented labour market policies as compared to Denmark (OECD 2003: SNA93). One could have argued that this ought to be understood on the background of the lower unemployment in Norway; however, even if we hypothetically calculate what the two countries would have spent if they both had had a registered unemployment rate of ten percent, Denmark still had a higher level of relative spending on ALMP than Norway since the early 1990s. This arguably reflects some differences in the political priorities and in the meaning of ‘activation’ between the two countries.

Since the late 1980s and early 1990s, Norwegian authorities have emphasised claimants’ co-responsibility to gain employment and qualify for labour-market participation, rather than solely relying on help from others. The government has sought to develop new policy instruments to achieve a higher degree of involvement and participation from larger sections of the population in the labour market. It has been focused to a greater degree on how the level and design of the welfare system may stimulate the individual to choose to remain in or return to paid work (Hvinden et al. 2001), except for the older age cohorts and to some extent parents with small children. The central government has reduced the duration of several welfare benefits, increased the demands of prior income from participation in the labour market to qualify for the benefits, and stressed and enforced to a larger extent demands of geographical and professional mobility and acceptance of job offerings or participation in labour market measures, training or re-education. For instance: the sick-leave allowance scheme was first modified in 1988, when it was required that a doctor ought to develop an independent report on the reasons for sickness after 8 weeks. In 1993 it was emphasised that difficult life situations or circumstances not caused by sickness should not qualify for sick leave (White paper 1993). There have been efforts to reduce the use of sick leave in Norway through dialogue with the social partners, stricter enforcement of medical criteria for sick-leave and the obligation of doctors to examine the working capacity of those claiming sick-leave allowance (White paper 2004).

The conditions for access to disability benefits without time limits have been enjoined. In 1988 it was emphasised that re-education and vocational training should be tested before claimants less than 35 years of age and with drug problems or socio-psychological problems were granted disability benefits. Demands pertaining to geographical and vocational mobility were increased in 1990. It should also be considered whether commuting or moving house could improve the prospects of finding work (White paper 1990). In 1991 it was stated that the medical causes for disability should be more clear-cut. One should be stricter in cases of vague diagnoses (White paper 1991). Attempt has also been made to achieve better exploitation of people's capacity to work part-time combined with partial receipt of disability benefits, or provide improved opportunities to explore the capacity to work without loosing the right to a predictable income in the form of a welfare benefit. Since 2003, a time-limited disability benefit scheme has been introduced to ensure the automatic re-evaluation of working capacity after 1–4 years (White paper 2002).

The unemployment benefit was limited to a maximum of 3 years in 1998 and later to 2 years (Drøpping et al. 1999). As a rule, the single-parent allowance was limited to 3 years from the birth of the youngest child in 1998.

Finally, the central government in Norway made the expectations and obligations to participate in labour market measures more visible in connection with the introduction of the 1991 Social Services Act on the provision of welfare benefit and other types of assistance from the municipalities, particularly for people 18–35 years of age. A larger number of beneficiaries have been faced with demands regarding participation in work, whether in the regular labour market or in some kind of quasi-labour market measure, in return for the receipt of means-tested welfare benefit (Kjellevold 1995; Vik-Mo and Nervik 1999).

What then, have been the effects of all changes? Clearly, these effects should be assessed more thoroughly, but preliminary data suggests that the reforms may have had less dramatic effects than is sometimes suggested in the analysis focusing solely on the policy reforms as such, rather than taking the implementation and their consequences into account:

• Labour force participation has been stabilised at a relatively high level. Since the mid 1990s, the rates have been around 80 per cent in both countries (OECD 2001: Table 21, 2003: Table 19).

• Overall, there are reasons to doubt whether economic inequalities have increased during the 1990s (Goul Andersen 2002). The compensation rate for unemployment benefits remained unchanged in both countries during this period of time. In 1985–95, relative poverty (50 per cent of median income or less) decreased by 2 per cent in Denmark to 5 per cent of the population and increased in Norway by 1 per cent to 8 per cent of the population (Goul Andersen 2004: Table 2.6). Admittedly, more detailed or in-depth statistics may disclose more complex patterns of development.

• The available statistics for social insurance schemes suggest that relatively few beneficiaries were punished economically. Estimates have been made in Denmark that less than 1 per cent of all recipients forfeited their right to unemployment benefits during the second half of the 1990s, while the number of cases increased from around 2.6 to 4.8 per cent in Norway from 1994 to 2000 (Arbejdsløshedskassernes Samvirke 2002; White paper 1995: 26, 2001: 192). Danish data suggests that the recipients of means-tested welfare benefit were rarely deprived of their benefits (Ebsen et al. 1999: 254; Boll and Christensen 2002). In comparison, Norway appears to demonstrate more disparate practices or local variation.

• The reasonability of the requirements ought to be considered while keeping the opportunities to achieve gainful employment in mind. It is therefore worth noting that evaluations of the activation efforts point to the fact that the greatest employment effect is ‘prior’ to activation (e.g. Jensen et al. 2002). Information about an approaching activation programme thus increased the number of welfare beneficiaries who exited the unemployment benefit scheme because they found paid employment. This indicates that there are realistic employment opportunities in the ‘ordinary’ market.

• More systematic data is required to systematically evaluate the content of the activation programmes and vocational training, as well as how they are experienced by the beneficiaries of welfare benefits. However, preliminary Danish reports suggest that a relatively large proportion of the beneficiaries experienced that their life quality improved and they were part of society to a larger extent than was the case prior to ‘activation’ (e.g. Hansen 2001). On average, the degree of self-sufficiency increased by about eighteen percent upon the conclusion of activation programmes in Denmark (Beskæftigelsesministeriet 2002).

• A particular incentive structure tempts municipalities to initiate particular forms of activation vis-à-vis welfare benefit claimants (Jepsen et al. 2002). Both in Denmark and Norway, the municipalities finance 100 per cent of the means-tested welfare benefit and none of the unemployment benefits. The municipalities will therefore be inclined to offer welfare benefit recipients ordinary wage work, making it possible for the beneficiary to attain or re-attain the right to unemployment benefits. The extent to which the jobs have substantial content and a requirement pertaining to return performance is likely to vary.

The relative positive assessments of ‘activation’ in Denmark and Norway may be due to the fact that activation must not occur in the abstract. Both Denmark and Norway have increasingly emphasised that welfare beneficiaries ought to receive an individual action plan prior to activation. Such individual action plans are intended to outline the objectives of participation in activation measures and the means by which the goals are expected to be achieved. Individual action plans can be interpreted as efforts by the public authorities and frontline staff in welfare-state services to lend an ear to the recipients of those services. If co-operation and user involvement proceed under the threat of negative sanctions, e.g. the withdrawal of services and social benefits, then the efforts by street level bureaucrats to motivate the users of welfare-state services may thus come to be perceived as representing social control (Rønning and Solheim 1997). It remains an unanswered question, however, whether there are systematic differences between the two countries with respect to the extent to which the needs of the unemployed are met in the individual action plan. As to the beneficiaries of unemployment benefits in Denmark, findings indicate that two-thirds of all unemployment benefit recipients who were activated within the first year of unemployment reported that they had a decisive influence on their activation, compared to 56 per cent of those who were activated during the second year of unemployment (Langager 1997). In the field of means-tested welfare benefits, preliminary data suggests that Denmark may consider the needs and wishes of the welfare beneficiaries more clearly as compared to Norway (Andenæs 2003; Ketcher 2002). Accordingly, Danish municipalities have granted priority to social integration over the improvement of labour market qualification and integration through paid work (Larsen et al. 2001). Norwegian data suggest that there have been significant variations in municipal practices (Vik-Mo and Nervik 1999). Overall, it would appear as though Norwegian municipalities have been under greater pressure to promote inclusion in the labour market, while participation in non-paid activities has been less legitimate.

Denmark and Norway appear to be relatively coherent ‘active societies’. The social protection system contributes to the mobilisation of most segments of the population in relation to the labour market. The value system prescribes that it is beneficial for both society as a whole and the individual that as many people as possible are participating in the labour market, and labour market participation is a value orientation acclaimed and supported by a majority of citizens.

The active societies in Denmark and Norway are clearly connected to what may be termed a universalistic activation regime (e.g. Barbier 2004). A trait of a universalistic activation regime is that welfare benefits and services are generous, whereas sanctions are rare. Activation is not enforced and it does not occur in the abstract. Rather, activation presupposes that an action plan is elaborated for each person individually. Thus, in principle, activation is tailored to individual needs and capacities. The aim of activation is to provide equal opportunities for everybody in the labour market and in society in general. The poverty rate and the number of working poor are low.

As a political catch phrase, however, activation is a relatively new phenomenon in the Scandinavian countries. It emerged in the early 1990s under the headings of aktivering (Denmark) and arbeidslinjen (Norway). Using the concept of an ‘active society’ as a lens of observation, however, the ‘newness’ of activation may be questioned. Since the late 1950s/early 1960s, it has actually been an overall goal to include the entire adult population in the labour market, and welfare policies have continuously sought to increase the size and the quality of the labour force. Thus, structurally activation as of the early 1990s may be perceived of as a trait of continuity, rather than change.

By contrast, quite a number of scholars have argued that activation in Scandinavia as of the early 1990s represents a dramatic change in welfare policy and citizenship. To some extent, however, this continuity–change controversy derives from the fact that different social science traditions tend to employ different analytical perspectives. A state-centred approach tends to identify profound changes by observing changes in the political discourse, programmatic changes in welfare policies, etc., while a society-centred approach emphasizes continuity by focussing on implementation deficits and the effects of policies on the structure of inequalities and the societal organisation of society.

Whatever the case, additionally one may argue that changes in the problem pressure may have a spin-over effect on the relationship between rights and obligations in welfare policies. A characteristic feature in Scandinavia has thus been that the relationship between rights and obligations has never been constant or naturally given. On the contrary, the relationship between rights and obligations has continuously been adjusted in accordance with the needs of the labour market. The relationship between rights and obligations has especially fluctuated alongside fluctuations in unemployment. For instance, in Denmark the conditions for receiving welfare benefits have been relatively relaxed in periods of high and enduring unemployment, and relatively tight in periods of near-full employment. Similarly, flows out of the labour market have been regarded as a result of failure or lack of success in national policy. Nevertheless, flows out of the labour market have been stimulated when unemployment was high.

Observing the emergence of aktivering and arbeidslinjen from the point of view of an active society helps to clarify, first, that the activation ‘turn’ in the Scandinavian countries in the early 1990s must largely be understood to be a trait of continuity, rather than a dramatic change; second, that the re-balancing of rights and obligations since the early/mid 1990s must be understood on the background of changes in the problem pressure. A decline in the problem pressure thus provided opportunity to return to an old tradition with an active society in the Scandinavian countries. It is therefore highly questionable whether the activation ‘turn’ from the early/mid 1990s can be interpreted as a decline in citizenship. Much evidence indicates that it should rather be understood as a restoration or enforcement of active citizenship in an active society.

Rune Halvorsen, dr. polit. in sociology, is a research fellow in the Department of Sociology and Political Science, Norwegian University of Science and Technology, Trondheim, Norway. Social marginality has represented the common denominator in his research on family policy, social movements and welfare policy.

Per H. Jensen, Ph.D. in sociology, is associate professor at the Centre for Comparative Welfare Studies, Aalborg University, Denmark, and Director of the Danish National Research School on Welfare State and Difference. From 1998 to 2003 he served as the chair of COST A13, Changing Labour Markets, Welfare Policies and Citizenship. He has conducted theoretical and empirical research within the fields of gender, early exit/early retirement, and labour market and welfare state conditions from a comparative perspective.