paper to be presented to the conference

European University Institute

Florence

15-16 April 1999

dr. Wim van Oorschot

Richard Engelfriet

1. Introduction

2. Developments in Dutch (un)employment

3. Types of participation policy

4. Main trends in Dutch social policies 1970-2000: social security and work

5. A closer look at Dutch participation policies

6. Dutch public opinion on the activation of unemployed people

7. Conclusions

Welfare states are under reconstruction throughout Europe; some as long as from the early 1980s (e.g. UK), others from the beginning of the 1990s onwards (e.g. Nordic countries), and still others have just made the first steps (e.g. Germany). A number of recent comparative studies show that national reconstruction processes share some common tendencies (see e.g. George and Taylor-Gooby 1996, Ploug and Kvist 1996, Daly 1997, Ferge 1997, Clasen 1997). To differing degrees, access to universal protection schemes has been limited and solidarity ties in social insurance systems (between good and bad risks and higher and lower income groups) have been reduced. At the same time, the role of means-testing has increased and measures that reduced the level and duration of benefits have been implemented. European workers and citizens who are confronted with a social risk and/or dependent on cash support now have to fulfil stricter qualifying conditions.

In addition to increased conditionality, a second trend of ‘participation’ can be identified, of a new emphasis on (mainly paid) employment as the prime basis of social protection. Under populist policy slogans like ‘welfare to work’ (UK) , ‘active line’ (Denmark) or ‘work, work, work’ (The Netherlands) those measures are becoming central elements in welfare reform programmes that are aimed at stimulating labour-market participation of unemployed and (partially) disabled social security claimants (e.g. Gilbert 1992, Scarpetta 1996, Thornton et al. 1997, Hvinden 1998, OECD 1996,1998). The trend manifests a fundamental change in social protection philosophy, holding that the provision of income from work is seen as a first and better means than providing income from benefit. In this philosophy benefit dependency should be prevented, and, if unavoidable, beneficiaries should be stimulated and supported in re-entering the labour market. If paid employment is not possible on a short term, long-term unemployed should be stimulated towards other socially valuable activities, in order to avoid them to become socially isolated and excluded. Perhaps with the exception of Sweden only, in which the ‘arbetslinjen’-approach has always been central in social protection policy, the participation trend in most countries constitutes a break from former times’ ‘passive doling out of money’.

The Netherlands is no exception to these general trends in Europe. Elsewhere we have described the trend towards increased conditionality in Dutch social security (van Oorschot 1998a). This paper’s aim it is to sketch a picture of the labour market participation trend as it developed in Dutch social protection during the last decades. For this we will first describe the main developments in Dutch (un)employment, in order to show the size and character of the social problem at issue. For heuristic reasons we will, secondly, discuss different types of (approaches to) participation policies. This will facilitate us in discerning certain trends in actual policy developments. Such developments will be sketched in broad terms in a fourth section, focusing on changing philosophies regarding work and social protection. A more detailed discussion of policies and measures then follows in section five. We distinguish between participation measures proper on the one hand, and the increased ‘work-relatedness’ of social security eligibility and entitlement criteria on the other. Measures stimulating the participation of disabled people will be discussed separately. In the sixth session a brief review of public opinion on the activation of unemployed people will show that Dutch activation policies have a sufficient societal legitimacy base. The paper ends with some general conclusions.

Figure 1 shows the development of unemployment in The Netherlands from 1970 onwards. Due to changes over time in measurement methods and definitions of unemployment the figures should not be taken for their absolute value¹. However, the trends are clear (figure 1). In the first oil-crisis of around 1973 unemployment started to rise from about 50.000 to 200.000. A number denoted as ‘unacceptable’ by the then government. As a result of the second oil-crisis of 1979 unemployment figures tripled within a few years, from about 200.000 in 1978 to more than 600.000 in 1982. Two years later an all time high was reached at more than 800.000 people. From then the numbers dropped gradually, but a third economic crisis in the beginning of the 1990s resulted in a rise again. In the last few years the number of unemployed decreases again.

Source: SCP 1998, SZW 1998

The figure 1 also shows that soon long-term unemployment became a central part of total unemployment. From the first peak in the 1980s the proportion of people unemployed for a year or more has been about half, while in all those years about a third to half of the long-term unemployed is unemployed for even three years or more. Throughout the years unemployment concentrated in specific groups. Unemployment rates have been higher than average for women, young people, ethnic minorities and people with a lower educational level. In 1997, for instance, the overall unemployment rate was 6.4%, but among women it was 9.1%, 10.1% among those between 15 and 24 years of age, 16% among ethnic minorities and 14% among people with only primary school.

Suggesting that unemployment developments only follow economic ups and downs and the resulting variations in the demand for labour would be unjust. Certainly in the Netherlands as important have been developments in the labour force, i.e. in the supply of labour. Figure 2 confronts labour demand and labour supply for the period between 1970 and 1996. It shows that employment has risen permanently in the Netherlands, with the exception of the recession years of the 1980s. A steep rise in job growth occurs after it and manifests the ‘Dutch job-machine’ or

Source: SCP 1998

‘Dutch miracle’ (Visser and Hemerijck 1997). The ‘miracle’ has not resulted in vanishing unemployment, however, because the labour force grew even quicker than the number of jobs. For a large part this is due to the massive growth in female labour market participation, which started by the end of the 1970s, stabilised in the beginning of the 1980s, but accelerated after the major economic crisis. The male labour force increased with 20% between 1970 and 1997, the female labour force with 125% (SCP 1998). In the same period the participation rate of men dropped from around 80% to 74%, while participation by women grew from about 20% to 47% (SZW 1998). Overall participation increased in this period from 50% to 65%.

Would the female labour participation have stayed at its low 1960s and 1970s levels, The Netherlands would not have had a high degree of long-term unemployment and it might have regained a situation of near full employment by now. Since this is not the case, the fight against (especially long-term) unemployment has been high on the agendas of successive governments throughout the last decades.

In recent years the Dutch labour market has become more flexible (table 1): the proportion of part-time employment increased, among men and women, as well as the proportion of temporary jobs.

The strong rise in labour participation of women mainly consists of part time jobs. Among couples the modern work pattern is one of ‘one-and-a-half job’, the man working (nearly) full time, the women part time. Finally, as a whole the numbers of hours worked annually decreased gradually from 1724 hours in 1973 and 1530 hours in 1983 to 1372 hours in 1996, which is the lowest in Europe (OECD Employment Outlook 1997). According to some this single fact has contributed substantially to the success of the Dutch jobs-machine (SCP 1998), next to the structural low-wage policy of government and social partners (see later).

To conclude, the main Dutch (un)employment trends of the last decades are: 1) structural long-term unemployment, 2) escape of older workers (disability, early retirement), 3) strong increase in female labour participation, 4) strong job growth in the last decade, and 5) flexibilisation of labour (part time and temporary jobs).

On principle a wide variety of measures is available for influencing the labour market participation of a country’s labour force. Distinguishing some general types is helpful in ordering the complexity and can be of help in describing, comparing and understanding developments and trends. Seeing the problematic of labour market participation as the (mis)match between demand and supply, a first distinction between measures directed at the demand side of the labour market and measures directed at the supply side imposes itself immediately. The first aim at influencing the amount and quality of work available, the second at influencing the size and quality of the work force. A second distinction generally made is that between policies directed at the macro or micro level (see e.g. OECD 1996, Goul Andersen 1999). Macro level participation policies aim at creating favourable wider economic, social and institutional conditions for participation, while micro policies try to influence directly the qualifications, behaviour and choices of individual actors. A combination of both dimensions leads to four main types portrayed in schema 1, each illustrated with some exemplary policies.

The ‘macro-demand’ approach is mentioned by the OECD as one making the difference between those countries who in the 1990s have been successful in decreasing the structural unemployment rate, and those who were not. The successful countries (The Netherlands, United Kingdom, Ireland and New Zealand) established an appropriate framework for macroeconomic stability and employment growth, by means of control over public finances and inflation, while the other, unsuccessful OECD-countries did not (OECD 1998). The ‘macro-supply’-approach may be manifested in national education policies, e.g. aimed at matching the future labour force’s qualifications to technological developments. In this approach it may also be tried to influence the size of the labour force. E.g. if unemployment is high through widening nationally possibilities for the early exit of older workers or through increasing the legal minimum working age. In other times, when employment is high, supply conditions may be improved e.g. through the widening of care facilities.

Policies at the micro-level come closest to what is generally understood by participation or activation policy. Such policies can be directed at employers, in profit or non-

profit sector, trying to stimulate (or sometimes enforce) them to employ unemployed or disabled workers. Wage subsidies, tax credits and direct job-creation programmes, for example, are widely used throughout Europe. Other micro policies are directed at workers. Here a wide range of instruments is available and practised, ranging from training, counselling, job search stimulation, job brokerage and the like, to individual incentives and sanctions.

Concentrating further on micro-level participation policies a number of other distinctions can be made, of which we think that those based on differences in general philosophy and on differences in the way in which participation policies are related to the income protection system are most relevant for understanding Dutch trends.

Participation policies might differ according to their basic philosophy regarding motivations and abilities of (unemployed) people. Such philosophies relate to preferences for certain types of instrument (see also Hvinden 1998).

In the enforcing and disciplining approach to the labour market participation of unemployed people unemployed are seen basically as people not willing to work, and not possessing the right kind of moral standards. Forcing them into work makes them ‘earn’ their income protection and is a way of disciplining them into working life. Next to this, force and discipline might be seen as a punishment for showing the wrong behaviour of idleness, laziness, taking advantage of society etc. Nowadays the approach mainly takes the form of strict work requirements for claimants of income support schemes. It means that benefits are nor granted, reduced or withdrawn if claimants are not willing to participate in some sort of work or training programme. Workfare is the term often used to refer to this type of policy. The enforcing approach might be applied with regard to the demand side of the labour market too, were it can take the general form of legal obligations for employers to offer employment to certain categories of workers, like through quota schemes for disabled people or ethnic minorities.

The incentives approach is based on a rational conception of men. Utility maximising unemployed people apparently prefer their status of unemployment above one of work. In this view unemployment can be prevented and reduced by introducing disincentives, making the status of unemployment less advantageous, and /or by introducing incentives that make work more attractive. Disincentive instruments might be to lower benefit levels, to shorten benefit duration, or make access to benefits more difficult generally. Incentive instruments are, for example, bonuses in case people find work, increase of wage levels of additional jobs, exempting part of earned income from means-testing, preventing unemployment- and poverty trap effects in tax- and income support systems (‘employment-friendly’ benefits and taxes: EC 1997) etc. Typically this approach is used also in participation policies directed at employers, like through wage subsidies and tax credits. Less popular, disincentive measures for them might be, for instance, penalties when firing disabled workers or linking social security contributions to unemployment or disability rates.

The guidance approach does not explicitly start from the assumption that unemployed are not or not sufficiently motivated to work, like the workfare and work incentive approach do, but from the idea that unemployed lack sufficient information on jobs available, and may lack sufficient qualifications and abilities to enter them. It tries to overcome difficulties of the unemployed in finding a job and to improve the functioning of the labour market more generally. In many countries there is a tradition of employment services or labour offices that implement public education and vocational training programmes, intercede between labour demand and supply, guide and counsel individual unemployed etc. Under the activation trend these activities mostly have gained more emphasis, and are referred to generally as Active Labour Market Policies (ALMP’s) (OECD 1996).

A recurrent reproof to the Dutch welfare state of the 1970s and 1980s is that it functioned mainly as a ‘benefit production plant’ (uitkeringsfabriek). In such a case social protection policy is limited to paying benefits to people who are confronted with social risks. At the same time the prevention of risks like unemployment and disability, or the recovery of self-sufficiency, e.g. through labour market re-insertion of disabled and unemployed workers, is grossly neglected. Active labour market policies may and often will exist, like training, schooling, job brokerage and guidance of job seekers, but they are applied mostly to short term unemployed with good qualifications. Furthermore, there will hardly be a link with the benefit system. Institutionally, because the administration of ALMPs is separated from that of benefits, as well as structurally, because there will be little or no interlinkage between benefit entitlements and participation in ALMPs. In this situation participation policy co-exists alongside income protection. Social protection as a whole is ‘passive’, since it focuses mainly on paying benefits.

In a more ‘active’ system of social protection matters are different. First, there will be an increasing degree of linking benefit entitlement to labour market participation. This can be done through a tightening and introduction of work related criteria regarding benefit eligibility, - level and – duration, as well as through obligations of unemployed beneficiaries to take part in active labour market programmes. And secondly, the overall approach to social protection will come closer to the originally Scandinavian ‘work line’, the Swedish ‘arbetslinjen’ being the prime example of it (see e.g. Wilensky 1992). Basically the idea of this is that participation policies precede public income support, in the sense that benefits are only paid after all suitable ALMP possibilities for (renewed) labour market participation have been tried. Self-sufficiency through employment is seen as a better means of social protection and support, than paying benefits. Not only from the perspective of society, which thus maximises the value of its labour force and limits social expenditure, but also from the perspective of the unemployed citizen, in whose interest it is to avoid dependency, social isolation and exclusion.

After the following sections we will be able to draw conclusions as to how in practice the approaches and philosophies directing practical policies changed over time in The Netherlands.

4. Main trends in Dutch social policies 1970-2000

After the prosperous 1960s, the first oil crisis of 1973 and the subsequent rise in unemployment to the ‘unacceptable’ number of 200.000 had a deep impact on Dutch socio-economic policy. The Dutch welfare state has been under permanent pressure from then onwards. The fight against unemployment and the restructuring of the social security system became increasingly interlinked. Despite this interlinkage we will, for reasons of clarity, sketch here broadly the main trends in changing philosophies and policies on both fields separately. (Based on information from Visser and Hemerijck 1997, van Oorschot 1998a, SCP 1998, SZW1998).

After the Second World War a generous and universal system of social security was built up in the Netherlands. It had its heyday between the late 1960s and the early 1980s. Since then it has been under permanent reconstruction.

Of first and crucial importance for this has been the steady rise in the number of claims for employees insurance and for social assistance in the course of the 1970s, followed by an alarmingly steep increase in unemployment and assistance dependency from 1978 to 1982. As table 2 shows, the number of people claiming unemployment benefits doubled from 1970 to 1978, and doubled again between 1978 and 1982, while the number of social assistance beneficiaries increased in those first eight years by 100.000, and by more than 250.000 in the next four years. The number of disability claims had a steadier, but by no means less meaningful, growth. Due to a broad definition of disability, based on post-war principles of universality and collective solidarity, the scheme had low access thresholds and attracted many older workers who otherwise, with more stringent entitlement criteria, would have been laid off. In other words, the number of beneficiaries of the disability scheme contains a large amount of ‘hidden’ unemployment. The number rose steadily from 215.000 in 1970 to 707.000 in 1982.

The economic crisis clearly learned that the system could be overloaded and would eventually collapse. The initial reaction was to try to keep social expenditures under control by lowering the duration and level of benefits. This reaction was known as "price" policy, because it was mainly directed at keeping the system affordable. However, by 1990 the number of employees insurance beneficiaries had increased by over 300.000 from 1982, which more than offset the decline in the number of social assistance beneficiaries during this period. Subsequently, the emphasis was put on "volume" policies which were aimed at reducing the accessibility of schemes and gaining control over the inflow of beneficiaries.

The reconstruction of the national insurances was not only the result of economic developments. It also reflected changes in Dutch society and culture. Revisions aimed to "modernise" the schemes by making them consistent with changing role patterns of men and women, particularly the increased participation of women in the labour force. This modernisation resulted in equal rights for men and women in all schemes. Where the modernisation of schemes implied a broadening of the entitled population, there was a conflict with the general aim of cutting back on social expenditures. The solution was that means tests were introduced to keep total expenditure under control. The ‘price’, ‘volume’ and modernisation measures taken subsequently include:

• a reduction of the earnings replacement ratio from 80% to 70% in unemployment, sickness and disability insurance,
• a sharpening of work history requirements for entitlement to unemployment insurance benefits,
• a limitation of the duration of earnings related disability and unemployment benefits, depending on age (disability) and work history (unemployment),
• a restriction of the concept of "disability", which in effect means that entitlement is reduced,
• a re-examination of the disability-status of 400.000 disabled workers according to the new concept, resulting in that 50% of them losing their benefits (and became entitled to unemployment benefit, which however has a much shorter duration),
• a ‘privatisation’ of the first 6 weeks of sickness, implying that the employer had to pay 70% of the wage during these weeks instead of the national sickness fund, followed by an extension of these 6 weeks to a full year (which de facto meant an abolishment of sickness insurance for almost all workers),
• excluding young people from the right to social assistance (instead they have to be offered a job by the municipality),
• a lowering of the basic assistance benefit for singles and single parents by 20% (a top up of 20% is possible if one can testify not to have a partner),
• introduction of means-tests in survivors and old-age pensions,
• a series of reductions of child benefits.

There is no doubt that these measures have contributed to putting a halt to the trend towards increasing numbers of beneficiaries and increasing expenditures. The system’s collapse was prevented. On the other hand, there is no prospect for a fast and substantial decrease in benefit demand and expenditures. The disability numbers nowadays come close to 900.000 and long-term unemployment still is a structural phenomenon.

Recognising the permanence of benefit dependency government gradually developed a new concept of social security, based on a fundamental critique on the built-in anonymity and irresponsibility in the model of collective solidarity. The national and collective nature of the system is supposed to undermine individual responsibility and to promote calculative behaviour by all actors involved, be it citizens, workers, employers, unions, or companies. Based on this diagnosis, market elements are introduced such as freedom of choice (for employers whether or not to take part in collective insurance) and risk differentiation (industries with higher levels of risk for disability paying higher contributions), which in essence are aimed at reintroducing individual responsibility, by way of confronting all actors more directly with the costs of social protection. The diagnosis also catalysed the trend towards activation, which comprises extended policies aimed at the (re-) insertion of beneficiaries into paid and even unpaid work (see next sections).

All in all, a shift has taken place from inclusive solidarity towards the direction of exclusive selectivity, from collective responsibility towards the direction of individual responsibility. With this shift the overall level of citizens’ social protection has declined. This loss, however, does not affect everybody to the same degree. Part of the decrease in protection offered by the collective system has been ‘repaired’ for workers in newly bargained collective labour contracts. The loss of collective social protection is also compensated at the household level as a result of the increased labour market participation of women, and the accompanying increase in double income households: more often the misfortunes of one partner can be compensated by the other partner’s means. Clearly, however, those who have lost most of their social protection are people with weaker or no ties to the market for paid labour. These include workers on flexible contracts, young workers, workers with repeated unemployment spells, and beneficiaries who have little chance of returning to the labour market, such as pensioners, disabled workers, long term unemployed and single parents.

In a first reaction to the initial increase in unemployment in the 1970s a Keynesian approach to stimulate the economy, and thus labour demand, was tried. Extra investments in e.g. housing were stimulated, unprofitable industries were supported and employment in the (semi-)government sector extended. When the budget deficit kept increasing and the measures had no direct impact on unemployment, soon a strategy directed at reducing labour supply was adapted. Early retirement of older workers was stimulated, older unemployed were no longer subjected to a work test and many redundant older workers were ‘shoved aside’ into the then easy accessible disability benefit WAO.

The measures taken, however, were not a match for the economic effects of the second oil crisis of 1979. In the early 1980s the idea took hold that the Keneysian approach was part of the problem. It had resulted in an increasing budget deficit and growing collective expenditures, and through this in high interest rates and wage costs, which both blocked economic recovery and employment growth. The period of ‘price’ policy, of major cutbacks on collective and social expenditures, was set in. For stimulating employment curtailing wage costs was regarded as the prime effective instrument. The benefit cutbacks had to result in lower contributions, and thus in lower wage costs, but the main emphasis was put on wage moderation. Under government’s threat of a national ‘wage stop’, social partners agreed in 1982 on a moderation of wage increases, which would recover the profitability of industries, and a reduction in working hours, in return for the unions. Notwithstanding its importance for the structural improvement of the competitiveness of the Dutch economy in future years, and thus for unemployment growth, the so-called Wassenaar Agreement had no immediate effect on unemployment. It was only with the improvement of the global economy from 1985 onwards that the numbers of unemployed started to drop. They dropped, however, slowly, and the problem of long-term unemployment persisted. This gave rise to the idea that the problem was not only based on a lack of jobs, but had to do also with characteristics and behaviour of unemployed and employers. Given the ongoing wage moderation (new agreements in 1992 and 1993), government’s attention shifted from a macro, demand directed approach, to a more micro directed approach. From now on labour market participation of unemployed people was on top of the agenda, and various means were applied. Some measures aimed at improving the ‘attractiveness’ of unemployed for employers, e.g. by improving their skills and qualifications, or by subsidising wage costs directly, or indirectly through the tax system. Other measures aimed at stimulating unemployed to search for and accept jobs, e.g. through personal ‘re-orientation’ interviews, a system of bonuses and penalties, lower benefit levels, and extra tax deductions for those in work. And a third type of measure was to create additional labour for specific categories, like for long-term unemployed (‘job-pools’) and young unemployed (Youth Work Guarantee Scheme, JWG). On the whole, the activation of beneficiaries and social security entitlements became more closely interlinked. When it became clear that long-term unemployment persisted, measures intensified. Especially in the beginning of the 1990s activating sticks and carrots were introduced in unemployment, sickness and disability insurance, as well as in social assistance. Work came to be seen as a better means of social protection than benefits. Again, however, despite the ongoing growth of employment, long-term unemployment persisted. When prime minister Kok came into power in 1994, the central slogan of his cabinet was ‘work, work, work’, with which it pointedly expressed its aim to intensify the approach already set in by the previous government. E.g. it extended additional labour by means of the so-called Melkert-jobs. To the 50.000 people employed in job-pools and Youth Work Guarantee Scheme another 50.000 were added. Furthermore, a reduction of taxes and contributions levied on low wages aimed at stimulating employment at the ‘bottom’ of the labour market were introduced, as well as a number of measures to stimulate flexible labour contracts.

Finally, although we do not want to pay much attention to matters of administration and implementation it has to be noted that with the linkage of benefit and participation policies the need for administrative co-ordination and co-operation increased. For this end the new institution of local and regional Centres for Work and Income (CWI’s) are implemented recently. They aim to function as one-door service organisations combining participation and benefit administration. The idea is that benefit eligibility and payment takes place only after possibilities for participation are assessed first. The CWI format should also promote participation in active labour market policies during benefit dependency.

Overlooking this brief history a few main conclusions can be drawn. First, wage moderation has been a central instrument in the Dutch fight against unemployment. It was and still is seen as a main macro economic condition for stimulating labour demand. Second, the initial emphasis on other macro-economic policies (like Keynesian investment or extending the public sector) has shifted towards micro directed policies aimed at influencing qualifications of unemployed, and behaviour and motivations of employees and employers. Third, labour market policy changed from passive to active. That is, emphasis shifted from paying social security benefits and reducing labour supply by older workers, to the activation of beneficiaries. Fourth, participation policies and social security have become more strongly interlinked, structurally as well as administratively.

Dutch active labour market policies concerning training, schooling, job brokerage and job search guidance have been designed and implemented already in the 1950s and 1960s. Even in times of shortages on the labour market these services are considered as necessary for the efficient matching of supply and demand. Until recently there were two institutions for trainign and schooling, each having about 25 district offices. The Centres for Professional Orientation and Practice (CBB) offered general training to about 4000 persons per year, especially ethnic minorities. The Centres for (Administrative) Vocational Training for Adults (C(A)VV) offered some 60 different courses in metal work, construction and business administration to about 10.000 persons a year. Both institutions have been merged for reasons of efficiency and co-ordination into the Centres for Vocational Training (CV). Unemployed people participate in CV courses after assessment of necessity and ability by labour offices, and employed can participate through company or branche projects aimed at preventing unemployment. For employers there are a number of subsidies stimulating them to train lower skilled employees.

Job brokerage and job search guidance is traditionally done by labour offices, which also act as the main gate keeper for training and schooling. Relevant to mention here is that according to central government labour offices were mostly creaming off the ‘best’ unemployed and failed in helping long-term unemployed. Since the beginning of the 1990s it tried to improve effectiveness. Social partners were given a say and responsibility in the labour offices’ boards in 1991, based on the idea that this would commit them more strongly to a just and effective matching of demand and supply. The results of this were disappointing, leading to forcing labour offices in 1996 to pay more attention to problematic categories of unemployed. From then on they are expected as well to co-operate with other players in the rather complex field. The main other players are municipal social services, responsible for the re-integration and activation of social assistance clients (most of whom are long-term unemployed). And social insurance administrations, which used only to collect social insurance contributions and pay benefits, but were assigned a responsibility for the activation and re-integration of their clients in the early 1990s. These three actors are expected to co-operate in the newly developed Centres for Work and Income (CWI’s).

From the beginning of the 1980s it was increasingly felt that the traditional active labour market policies and institutions could not cope with the initial mass unemployment, and the structural high proportion of long-term unemployment that was left after the deepest crisis was over. A large number of extra measures was taken, of which we can present here only the main lines. We distinguish between activation measures proper and work-related benefit criteria. The first we define as measures that are directly and explicitly aimed at the (paid or unpaid) labour participation of unemployed people and/or at preventing employed people to become unemployed. Introduction and tightening of work-related criteria for benefit eligibility and entitlements have an indirect (but mostly not unintended) activation effect, mainly because they make unemployment less attractive for workers and result in higher pressures on unemployed to find work. We will discuss measures aimed at the (re-)integration of disabled workers separately. Disability claimants have no formal obligation to find work, but their labour market participation has become a major concern.

An inventory of measures taken since the mid 1970s is presented in the appendix schema A. A close look leads to a number of general conclusions.

First, there is an ‘explosion’ of measures in the second half of the 1980s . Before that the main scheme was Loonsuppletie, offering a temporary wage supplement for all categories of unemployed when they accepted a job with a wage below what they use to have in their last job. Newly added measures specifically aimed at young and (very) long-term unemployed, which by then formed large parts of total unemployment.

A second conclusion is that three groups with notoriously bad labour market chances, older unemployed, women and ethnic minorities, are absent as explicit target groups for activation measures. There are specific schooling and training programmes for these groups (e.g. (part of) CV for ethnic minorities, Vrouwenvakscholen for women), and within the measures for long-term unemployed they are sometimes seen as groups deserving extra attention, but apparently government has hesitated to design any explicit measures for them. We could not find public statements about its reasons for this. We speculate that is has to do with a fear of (further) stigmatising unemployed women and people from ethnic minorities in the eyes of employers, which could be the result of making them the subject of explicit measures. It might also be related to governments sensitivity for public opinion, which is not in favour at all of positive discrimination of ethnic minorities and women (see table 7). Regarding older unemployed the situation is different. It was among the first reactions to the economic and budgetary crisis of the early 1980s to exempt unemployed of 57.5 years of age or older from the work obligation as a means of creating more opportunities for younger cohorts. Only very recently, in 1999, by the time when unemployment has dropped substantially, this exemption has been abolished.

Thirdly, only few activation measures are solely aimed at the unemployed persons themselves. Wage supplement for those accepting a job with a lower wage level, re-orientation interviews aimed at designing individual re-insertion plans (HOG) and ‘social activation’ (Melkert III) are the only measures that do not require the immediate co-operation of employers. All other schemes do, in the sense that they try to persuade employers to employ long-term unemployed by means mainly of temporary or permanent wage subsidies and reduction of taxes and social security contributions levied on wages. Apparently, the perceptions and attitudes of employers, and their related selection behaviour, is seen as more of a concern than the motivations and qualifications of the unemployed. Studies have indeed shown that by far the largest part of all unemployed is very eager to find a job (e.g. Jehoel-Gijsbers 1989, Kroft et al. 1989). In the beginning of the 1990s it is even tried to stimulate private employment agencies to concentrate their services more strongly on long-term unemployed (under VU and KRU) by means of a subsidy for every successful case. Soon, however, this was seen as an unjust way of merely subsidising the agencies with little lasting effects for the unemployed and the schemes were abolished.

Fourthly, there is a mix of measures aimed at employment in additional jobs and in regular jobs. Both types of job can be realised in either profit or non-profit organisations. Additional jobs are mainly created for very long-term unemployed (Banenpool, Melkert I and II) and young unemployed (JWG). In case of the very long-term unemployed additional jobs are seen as the only way of avoiding the strong barriers for this group on the regular labour market. Barriers mainly consist of stigmatisation of this category among employers, their older than average age, their lower than average skill level and the higher proportion of ethnic minorities. In case of the young unemployed additional jobs have replaced the right to social assistance benefit.

Fifthly, schemes vary in the number of people participating in them. Generally the participation rate, as the percentage of participants relative to the total target group, is very small. For instance, in 1988 about 7000 young unemployed participated in the TV-GWJ, while nationally about 45.000 met the criteria. The Wet Vermeend-Moor covered 170.000 people in 1987, but only 6000 participated (SCP 1992). Even the ‘larger’ schemes, like Banenpool (23.000 participants and Melkert I (40.000) and II (20.000) only cover small parts of their target groups of a few hundred thousands of long-term unemployed. A great success in terms of participation seems to be only the WVA-SPAK, which offers a tax and social security contribution reduction for employers on the wages of those of their employees with a salary below 115% of the minimum wage. This measure aims to stimulate employment at lower job levels for which low-skilled unemployed could qualify. First evaluations of the scheme shown that it is also a success in terms of enhancing the labour participation of unemployed people (SZW 1999).

Finally, most of the various measures have recently been integrated in two framework laws. One, the WIW, administered by municipal social services, aims at the participation of long-term unemployed in additional jobs. It incorporates e.g. the Youth Work Guarantee scheme and Banenpool The other, WVA, integrates schemes aimed at other types of unemployed or employed people, mostly schemes offering reductions and subsidies to employers. The reasons for integration are various. There was overlap between some of the measures, competition even, clients and administrations had difficulties in distinguishing between the various conditions and target groups, and there was lack of overall co-ordination.

Evaluating the effect the whole of activation measures thus far has had on unemployment is problematic. It is e.g. difficult to separate it from the effects of other factors, like employment growth due to economic revival or the effect of restructuring social security. Given the fact that at present the number of unemployed still is higher than the level that was regarded as unacceptable in the 1970s one could conclude that policies have not been very successful (SCP 1998). On the other hand, since the mid 1980s unemployment has dropped, and the Dutch unemployment rate of about 6% is among the lowest now in the OECD countries. Long-term unemployment, however, still remains as a structural problem, despite all the targeted measures just discussed. Whether the situation would be even worse if the measures would not have been taken is a question that is difficult, if not impossible, to answer. Maybe one could conclude that policies aimed at employment growth (like national wage moderation) have been more successful than those aimed at reducing long-term unemployment.

Evaluations of individual activation measures are plentiful. In a recently published report in which various types of Dutch and foreign measures are compared the conclusion is drawn that those measures aimed at reducing wage costs for employers are most successful in terms of re-inserting unemployed on the regular market of paid labour (NEI 1999). This conclusion is in line with that of an international comparative report issued by the OECD (OECD 1996). The problem of measures creating additional labour is that ultimately they result in a few participants streaming into regular jobs. Once in an additional job there seem to be lack of possibility and motivation to move on. For, another conclusion is that the stream from subsidised to regular jobs is smaller when the subsidy period lasts longer, and when the subsidised jobs are in the public sector. Finally, the report shows that re-insertion of older people, lowly educated people and people from ethnic minorities is least successful. Administrations that want to show impressing success rates tend to cream their clientele, i.e. concentrate their efforts on those with the highest labour market chances. Creaming in the activation business is a recurrent finding of many evaluation studies (SCP 1992). Other typical problems of activation policies are the relatively high administrative costs and, in case of additional labour, the issue whether the jobs created are in effect additional or just pushing away other people’s regular jobs.

From the mid 1980s work-related benefit criteria and conditions have been introduced and tightened in Dutch benefit schemes for unemployed people. Initially they mainly served the purpose of cutting back on social expenditure, later they are explicitly motivated on the basis of their activating effects. Work-related criteria and conditions make unemployment less attractive for workers and result in higher pressures on unemployed to find work. We will discuss unemployment insurance WW and social assistance separately.

Dutch workers are entitled to unemployment insurance if they become involuntarily unemployed for at least 5 hours a week or lose at least half their working hours, their income drops as a result of that and if they hold themselves available for re-insertion on the labour market. The latter implies a number of specific conditions that have to be met: 1) registration as jobseeker at the labour office; 2) acceptation of ‘suitable’ work; 3) co-operation in assessments, training and scholing; 4) actively trying to find work. These criteria are traditionally attached to unemployment benefits, and have not changed in the last two decades. What did change was the rigidity they are applied with. Sanctioning those who did not comply used to be a ‘competence’ of the administrative bodies. However, with the Law on Penalties and Measures (Wet Boeten en Maatregelen) of 1996 it became an obligation. This law, from 1997 covering social assistance too, aims at intensifying the sanctioning policies of social security administrations in order to activate unwilling unemployed more strongly. Penalties are nationally prescribed per type of misbehaviour and administrations are policed on their implementation.

The most important change, however, is that in the 1980s and 1990s the work-relatedness of eligibility and entitlement criteria has increased significantly. Mainly through linking them (more strongly) to a person’s work history. Schema 2 offers an overview.

Regarding benefit eligibility the 1987 reform of the WW scheme implied a higher threshold: from 130 days work history in the year previous to unemployment, to 26 weeks. The prolonged benefit after the half year of standard benefit became dependent on a further requirement of having worked at least 3 years in the previous 5 years. After expiration of the prolonged benefit a follow-up benefit was paid for a period depending on age. In 1995 eligibility criteria for the standard benefit were tightened again. Now one has to have been working 26 weeks in the previous 39 in stead of 52 weeks, and 4 years in stead of 3 years in the last 5 year period. Most significant is, however, that both conditions are now combined, which implies that nowadays it is rather difficult to be entitled to even the standard benefit. It is estimated that only about 45% to 50% of the present workers would meet the combination of criteria. Those who only meet the criterion of 26 weeks out of 39 are entitled only to short-term benefit of 70% of the minimum wage. Those unemployed who do not meet any of the criteria have to rely on means-tested social assistance.

Regarding benefit level the schema shows that traditionally the WW insurance benefit is fully wage related. In 1987, however, benefits for long-term unemployed, who expired their rights to the standard and prolonged benefit, were converted into flat rate payments of 70% of the minimum wage. In 1995 the flat rate was introduced further for those who expired their standard benefit only. So, benefit levels are increasingly linked to work history through their relation with the various types of benefit within the WW scheme.

Regarding benefit duration we see that the maximum duration of WW-benefit has extended from 2,5 year (for those under 60 years of age), to maximally 7,5 years. However, the work history requirements for entitlement to prolonged or follow-up benefit, i.e. the benefits following after standard benefit has expired, have been tightened strongly. So, for those unemployed with extended work records duration has increased, for those with smaller work records it has become more difficult to get (wage related) benefit for more than half a year.

On average being unemployed now means lower benefits, for shorter periods. As a result keeping or finding paid work has become more important and compelling to workers.

Being the safety net of the Dutch social security system the means-tested social assistance scheme (Abw) has no work-related requirements regarding benefit eligibility. Every citizen lacking sufficient means of subsistence is eligible. Benefit level and duration are independent of work history criteria too, but in practice they may vary according to whether clients conform to work related obligations or not, which obligations are imposed upon them as soon as the benefit is paid. Not conforming to them may result in penalties (of 5% to 20% of the amount of benefit) or a withdraw of the full benefit. Penalties and withdraw may only last for a maximum of one month to prevent people from falling under subsistence level structurally. The work related obligations have not changed in the last decades, but as in case of unemployment insurance the previously mentioned Law on Penalties and Measures of 1997 has resulted in a more rigid and systematic implementation. The obligations are: 1) to try and find work actively; 2) to register as jobseeker at the labour office; 3) to accept suitable work; 4) to co-operate with assessments, training and schooling; 5) to refrain from activities that hinder labour market participation.

In the 1996 restructuring of the social assistance scheme a number of work related elements were introduced, aimed at activating the assistance clientele. First, for young people under 21 entitlement to benefit was replaced by an entitlement to a job in the framework of the Youth Work Guarantee scheme (JWG). Young people are expected to work for their subsistence. Secondly, single parents with children over 5 years in stead of 12 years of age became subjected to the work related conditions. Third, the standard of ‘suitable work’ has been broadened, implying that beneficiaries are expected to accept work well below their educational and former job level. Fourth, for each client with a reasonable labour market chance the administration has to design and implement an individual re-insertion plan, in close collaboration with the labour office. Fifth, municipalities, which administer social assistance, are obliged to design and implement a system of bonuses for clients that either participate in training and schooling, or accept an additional or regular job. And finally, it has been recognised that nearly half of the social assistance beneficiaries have very little real chances of finding a job. To prevent social isolation and apathy municipalities have the possibility of implementing ‘social activation’ policies: that is, to stimulate clients to do voluntary or community work in exchange of exempting them for the work related obligations.

So, the social assistance clients are ‘activated’ in various ways, but the possibilities of sanctioning those who do not comply are limited.

The (re-)insertion of incapacitated workers on the labour market traditionally has had little attention in Dutch social policy. Fully disabled workers have no formal work obligation. Partly disabled do, but only in case they are for the other part unemployed and receiving a partial unemployment or assistance benefit.

Traditionally there are sheltered workplaces for a (limited) number of handicapped people, and the WAO disability benefit scheme offers some possibilities for adjusting workplaces to handicaps, but structural and effective measures were not taken. This all changed, however, when the number of WAO-disability claimants kept growing and stayed at very high levels, even after the economic recovery at the end of the 1980s. A variety of measures were taken to try and stimulate labour market participation of (partly) disabled workers (see van Oorschot 1998c for a full review).

In 1986 the ‘act on work for disabled workers’ WAGW was introduced as the first major Dutch re-integration act for disabled workers. The WAGW did not contain many new concrete measures. It mainly offered a legal framework for existing measures and imposed on employers and unions the responsibility to promote equal chances for disabled workers.They could use already existing (but, according to the government, insufficiently used) possibilities under the WAO benefit scheme. New was that a quotum of 3% to 7% would be imposed on an industrial branche if three years after the introduction of WAGW results were lacking. The WAGW has never been seriously implemented, however, since the proportion of disabled workers in companies did not increase, and a quotum has not been imposed on any branche. A second new measure was that the administration of the disability benefit could claim payments from an employer equal to the wage a fired disabled employee would have earned if the employer had made employment possible. Few if any of such payments have ever been requested. A third measure, holding that employers were given the obligation to adjust (access to) the workplace of any disabled employee they have in their firm, has never been seriously controlled. It is generally felt that the act was used as a ‘sweetener’ for the cutbacks on WAO-benefits that were discussed in 1986 and implemented in 1987. Yet, the WAGW is not without significance. It is the first act in which a general concern for the labour participation of disabled workers is expressed.

In 1987 the WAO disability insurance benefit changed importantly. Before the revision partly disabled people without a job received a full wage-related disability benefit for as long as their disability lasted, on the basis that their chances on the labour market were nearly zero. With the revision, however, partially disabled workers without a job became entitled to a partial, instead of a full, wage-related disability benefit. For their unemployment part they became entitled to an unemployment benefit WW or social assistance, and thus subjected to work obligations.

In the early 1990s two other major laws were introduced. One in 1992, the ‘act on reducing the disability volume’ (TAV), which aimed at creating greater incentives for employers to prevent claims for disability benefits. It introduced a ‘bonus-malus-system’ under which employers receive a subsidy if they employ a disabled worker for at least a year. In addition to this once-off subsidy, a 20 per cent wage subsidy is also provided. On the other hand, employers have to pay a fine or ‘malus’ if one of their employees gets disabled at work and has to be fired as a result. In 1993 a second ‘act on reducing disability claims’, the TBA, was implemented. It broadened the reference standard for the assessment of the degree of disability from ‘suitable work’, which was defined as work suiting one’s educational level and former type and level of job, into ‘generally accepted work’, a standard not connected to educational and former job level. As a result more jobs are regarded as in principle being available for the disabled. Furthermore, every existing beneficiary of the WAO benefit younger than fifty years of age had to be re-assessed according to the new standard. In the first two years after its implementation this rule resulted in a withdrawal of the full WAO benefit in 50 per cent of all reassessed cases. People concerned were declared to be fully unemployed, instead of being (partly) disabled, and had to claim unemployment benefit, with its work related conditions.

AMBER of 1995 was the next step. It abolished the fine or ‘malus’, which had to be paid under TAV by employers if they fired a disabled worker. This fine met strong resistance among employers unions and administrative bodies had large difficulties with implementing it. Partly to compensate for the loss of this incentive for employers to employ disabled people, but also to carry further the ‘fight’ against the disability volume, new financial incentive measures were taken for stimulating participation of disabled workers. Among the measures are: an extension of the wage subsidy for employers (introduced earlier under TAV); a wage supplement for disabled workers in case they loose income when accepting work; a guarantee for disabled workers over 50 that they will get their old benefit level back if they have to stop working again; the possibility of working in a ‘try out’ job without loss of benefit rights.

Fourthly, in a further attempt to get control on the WAO inflow, the 1998 PEMBA law just recently took effect. It aims at influencing behaviour of employers in such a way that they feel an individual responsibility for the prevention of disability as well as for the (re-)insertion of disabled workers. The law introduced two measures, premium differentiation and opting out or privatisation. Before PEMBA contributions for the WAO scheme were not differentiated according to risk, i.e. to the number of disability claims coming from individual firms and sectors of industry. All paid a uniform percentage of wages. Under PEMBA such percentages, and thus the amount of contributions, is differentiated according to risk. As a result, firms and sectors of industry that generate more disability claims have higher costs. They can cut costs by preventing disability claims. This can be done either by an improvement of working conditions, or by adapting work places for disabled employees. PEMBA also offers individual firms the option to leave the collective system and assume responsibility for the disability and subsequent benefits that it generates. (Some large companies have already chosen to ‘opt out’, but the first signs are that only few will follow).

Finally, the 1998 act ‘on reintegration of handicapped persons on the labour market’ (REA) replaces WAGW as the comprehensive legal framework for re-insertion policies for disabled workers. To government’s opinion there is not so much a lack of measures, but of results. Government acknowledges that this might have to do with measures, taken e.g. with regard to short-term disability, that induce employers to avoid employing workers with health problems. Therefore, it introduces with REA that employers get a fixed budget in case a job is offered to a disabled person. From this budget the employer is expected to pay the necessary workplace adaptations and access improvements. If costs are less than the budget the surplus does not have to be re-imbursed. If costs exceed the budget an extra amount is possible. Furthermore, under REA the sickness pay for an employed disabled person will be paid from the national sickness fund, in stead of by the employer, so that the possible extra sickness risk will not burden the wage cost. The employer can get a reduction on his WAO-contributions if 5% or more of his wages is paid to disabled employees.

The main revision of the ZW started in 1994. Before that benefits for workers who were ill (for less than a year after which period the disability scheme comes into force) were paid from the collective sickness fund for the full period. The fund was financed by contributions from employers and employees. The relation between degree of sickness absenteeism and costs of insurance was not strong, since contributions were only differentiated between branches of industry. Thus, at the level of individual firms and workers, incentives to prevent sickness were lacking. This changed with the ‘act on reducing sickness absence’ TZ of 1994. Under this law, employers were obliged to pay sick employees at least 70 per cent of their wage for the first six weeks of absence (two weeks for companies with less than 15 employees). Thus, the first weeks of sickness were privatised and did not burden the national sickness fund anymore. Either employers now paid wages for sick employees directly, or like most of them did, reinsured the risk with private insurance companies. Reducing sickness absenteeism was further promoted by a second obligation, which held that every firm has to develop and implement a sickness absence prevention and control policy. In 1994, another revision took place as a result of the earlier mentioned TAV law. This law introduced a further differentiation of contributions for sickness benefit within industry sectors. Firms with a higher absenteeism than their sector’s average pay higher contributions.

The TZ, especially the part concerning privatisation of the first weeks of sickness benefit, had an immediate and large effect on the national sickness fund. In 1993, 345.000 sickness beneficiaries were paid compared to only 175.000 in 1994. In 1994 the overall percentage of reported absenteeism dropped from 7 per cent to 4.5 per cent of total labour time and has stayed at that lower level since then (Ctsv, 1995, p.65). Although there is a natural floor to this percentage, central government hoped that a further privatisation would result in a further decrease and extended the period in which employers had to pay wages to sick personnel to one year. This measure, known as the WULBZ law that came into effect in 1997, practically implies the abolition of the ZW. ZW still exists at present, since it still covers the sickness risk of specified categories (estimated at 15 per cent of the previously covered population), like pregnant women, (partly) disabled workers, people on temporary contracts and apprentices. But for the largest part of the Dutch workers it is replaced by the employer’s duty to keep on paying wages during sickness leave.

To conclude, the labour market participation of long-term and short-term disabled workers has been the subject of major policy concern. Access to the schemes is restricted, more work is seen as suitable for people with disabilities, partly disabled unemployed have stronger work obligations, and a number of financial incentives have been introduced for unemployed, but mainly for employers. It is mainly through the workings of these latter that a remarkable tension has grown between the intended activation intentions of these measures and their actual effects. For, the employer incentives are set such that they profit from having a workforce with a minimal disability and sickness risk. Evaluation studies (discussed in SCP 1996) have already shown that chronically ill people and (partially) disabled people nowadays even have more in stead of less difficulties in (re-)entering jobs, because employers screen new employees more stringently on their health status. It is also found that the chances to be fired have increased for workers with a worse health status. And that the number of temporary labour contracts, as a means of prolonging the period of screening employees on their ‘sickness leave behaviour’, nearly doubled from 1993 to 1995 from 11 per cent to 20 per cent of all labour contracts. Hiring workers via employment agencies, to reduce the risk of sickness pay, rose in the same period from 4 per cent to 9 per cent.

The public’s view on activation policies is relevant for the societal legitimacy of the policy trends just sketched. There are some data available (van Oorschot 1998b) that cast light on this issue.

Table 4 shows that the public has no restrictions in applying the obligation to (find) work strongly to young unemployed. A majority of nearly two thirds also favours strict application towards the long-term unemployed. However, when unemployed are active in voluntary or community work, when they clearly contribute to society in other ways than paid work, the public is more generous. In that case a majority would favour a less strict application of the obligation to find a paid job. The public is even more generous when unemployed have no real chance on the labour market what so ever, when they care for a sick household member, and when they are older than 55. In these cases a quarter to nearly half of the public would not even apply the obligation at all. With regard to single parents the Dutch public is rather considerate generally. Least in case of children older than 16, but clearly to a large extent when children are younger than that. In general the idea is that single parents should not be set free from the obligation completely, although 40% would prefer that if children are under 5. The public is more considerate towards single parents on benefit than the social assistance law, because from 1996 onwards single beneficiaries with children over 5 are obliged to look for a paid job. In the case of young widows a majority would like to be strict on the obligation to find work, while a majority would prefer a supple application towards older widows. In practice, the young widow actually has to find a job, because she would not be entitled to widows’ pension and would have to claim social assistance in stead. The older widow would be entitled to a pension, with no work obligation. If people are partially incapacitated to work the overwhelming majority of the Dutch population would like them to be treated supple, while it would prefer fully incapacitated people to be set free from the obligation completely.

Table 5 shows that the majority of the Dutch public is of the opinion that unemployed people should accept a job with a lower skill level and with a lower salary than they used to have. However, only about a quarter would be very strict on that, in the sense of requiring of an unemployed person to accept much lower levels of skill and salary. Opinions are clearly divided about whether or not unemployed should accept jobs with a salary below benefit level. Half of the population offers unemployed the right to say no to such jobs, while the other half would not accept such an attitude in case the salary is only a bit below the amount of benefit.

Despite the differences in opinion with regard to the last item, the general conclusion still can be that the Dutch population sees the acceptance of paid work as a duty of unemployed people. But, would one like to go as far as to oblige unemployed people to accept unpaid work, like in a system of workfare? That is, to accept work that is useful for society or the community, in return for which the unemployed does not get a salary, but just keeps his of her present benefit. And what about the unemployed’s partner? Should he or she be obliged to look for a job and earn an income for the household too? Table 6 shows that a relatively large part of the public has no clear opinion on these matters. About a third says not to know what to answer. Of the remaining part the majority is in favour of workfare, i.e. of obliging unemployed to contribute to society by way of carrying out unpaid work. With regard to the partner’s duties opinions are clearly divided.

With regard to measures aimed at redistributing available jobs to the advantage of specific categories the Dutch public has some outspoken opinions (table 7). Measures that are supported by the majority of the public are: extending the possibilities for early retirement, so that the older make place for the younger; give a financial bonus to employers if they hire long-term unemployed or incapacitated workers; and, extent the possibilities for educational leave.

Less popular measures, to which most people are in favour, but which would meet resistance of about a quarter to third of the public, include: oblige employers to hire long-term unemployed and incapacitated workers; extent parental leave; shorten the workweek for full time employed; and, limit the number of working hours for earning couples. Clearly unpopular measures would be to stimulate financially or to oblige employers to give priority to women or ethnic minorities for filling up vacancies. So, according to the public the young should have more access to available work, as well as long-term unemployed and incapacitated workers, but not women and ethnic minorities.

From this overview of available public opinion data we conclude that the shift towards the activation of social security beneficiaries has a sufficient degree of societal legitimacy. People want unemployed people to be obliged to look actively for jobs, and they want them to accept less favourable job offers. Clearly, however, opinions of the Dutch are shaded. They agree with activation, but it should not be implemented too rigidly in all cases. Older unemployed, unemployed who are active in socially valuable ways, single parents and incapacitated workers should be treated supplely. A second conclusion is that the Dutch public agrees with measures that aim to re-distribute available work more evenly. For this it accepts special treatment of long-term unemployed and incapacitated workers generally, but women and ethnic minorities are not seen as especially deserving in this respect.

Dutch policies aimed at the labour market participation of unemployed people started to develop as a reaction to growing unemployment in the 1970s and, especially, the early 1980s. Despite the economic revival and large employment growth in the second half of the 1990s, participation is still high on the social policy agenda, for long-term unemployment has not declined substantially among older unemployed, the low-skilled, single parents and ethnic minorities, and the number of partly disabled unemployed is still very high.

Over the years a permanent and substantial element of the Dutch approach to employment and participation has been the policy of wage moderation, aimed at keeping wage costs down. It was and still is seen as a main macro economic condition for stimulating labour demand. Initially this ‘macro-demand’-approach was accompanied by other such types of measure, like Keynesian investment or extending the public sector, and by ‘macro-supply’ types of policy trying to reduce the total amount of labour supply, e.g. through stimulating early retirement and ‘shoving ‘ older workers into disability benefit.

When the budget deficit kept growing, long-term unemployment remained a problem and the growth of the number of (partially) disabled unemployed could not be stopped, the emphasis shifted towards micro directed policies aimed at influencing qualifications of unemployed, and behaviour and motivations of employees and employers. An incentives approach, based on a rational conception of men, evolved into a central position. A variety of financial incentives were created to stimulate unemployed and disabled workers to find and accept jobs, but most measures are directed at employers, trying to stimulate them to employ long-term unemployed and disabled workers and to create jobs at lower wage and skill levels. Indirect incentives are created through an increase in the ‘work-relatedness’ of benefit eligibility and entitlement criteria. Although the effectiveness of incentives measures usually is hard to assess, there are clear indications that measures aimed at stimulating regular jobs through financial incentives for employers are more effective in the long run than subsidising additional jobs. With regard to the (re-)integration of disabled unemployed incentives measures seem to have been leading to contra-productive effects. An enforcing and disciplining approach to participation and activation of unemployed people has never been strong in the Netherlands, but was strengthened in the 1990s through new legislation on sanctioning unemployed who do not comply with work obligation rules. The guidance approach, consisting of training and schooling, job brokerage and helping unemployed in jobseeking has a long history in the Netherlands, and services have been extended in the last decades, but its institutions are being criticised for their ineffectiveness with regard to combating especially long-term unemployment.

The traditional guidance approach existed alongside the income protection system. Institutionally, because the administration of the related active labour market policies was separated from that of benefits, as well as structurally, because there was little or no interlinkage between benefit entitlements and participation in ALMPs. Social protection as a whole used to be ‘passive’, since it focused mainly on paying benefits. In the last two decades this changed towards a more ‘active’ approach in which income protection became more strongly linked to labour market participation. Not only through an increased work-relatedness of benefit entitlement, and strengthening of work obligations, but also through the various incentives measures. Most recently there is a trend towards the originally Swedish ‘arbetslinjen’ approach, in which participation policies even precede income support, in the sense that benefits are only paid after all suitable ALMP possibilities for (renewed) labour market participation have been tried. The newly developed Centres for Work and Income (CWI’s) are the main institutional outcome of this approach.

These CWI’s are a form of institutional integration trying to overcome the ineffectiveness of the administrative differentiation that exists not only between assessing entitlements and paying benefits on the one hand, and implementing participation policies on the other, but also between different types of benefits (unemployment insurance, disability insurance, social assistance). Structurally the integration of policies is promoted by the recently introduced encompassing framework laws for the re-insertion of social assistance clients (WIW) and other types of unemployed and employed people (WVA), as well as of disabled people (REA). Their aims are to reduce the complexity of the totality of existing measures and to co-ordinate and harmonise between them. Ultimately, clients should profit from the institutional and structural integration. It is too early yet for any evaluation.

Evaluating the total effect of Dutch participation policies one could conclude to success, as well as possible failure. Successful seems to have been the macro-demand approach of wage moderation, which is generally seen as the driving force behind the Dutch jobs-machine, and which has allegedly led to the relatively low present unemployment rate of around 6%. Possible failure perhaps with regard to the micro directed policies. They did not succeed in reducing substantially long-term unemployment, which is the main goal of most of the measures taken. The (re-)integration of disabled unemployed still is more of a noble wish than a reality, and policies here even seem to be contra-productive.

AAW Algemene Arbeidsongeschiktheidswet

General Disability Act

ABW Algemene Bijstandswet

General Social Assistance Act

AKW Algemene Kinderbijslagwet

General Child Benefit Act

AMBER Wet Afschafing Malus en Bevordering Reïntegratie

Act on Abolishment of Malus and Promoting Reintegration

ANW Algemene Nabestaanden Wet

General Survivors’ Benefit Act

AOW Algemene Ouderdomswet

General Old Age Benefit Act

AWW Algemene Weduwen en Wezen Wet

General Widows and Orphans Benefit Act

Act on Reducing Income Consequences of the New Disability Standard

C(A)VV Centra voor (administratieve) vakopleidingen voor volwassenen

Centres for (administrative) Vocational Training for Adults

CBB Centra voor Beroepsoriëntatie en Beroepsbeoefening

Centres for Professional Orientation and Practice

CV Centra voor Vakopleiding

Centres for Vocational Training

CWI Centrum voor Werk en Inkomen

Centre for Work and Income

EAA Experiment Additionele Arbeid

Experiment Additional Labour

EAU Experimenten Activering van Uitkeringsgelden

Experiments on Activation of Benefitexpenditure

EWLW Extra Werkgelegenheid voor Langdurig Werklozen

Extra Employment for Long-term Unemployed

HOG Heroriënteringsgesprekken

Individual Re-insertion Plan

JOB Jeugdontplooiingsbanen-maatregel

Youth Developing Jobs Enactment

JWG Jeugdwerkgarantiewet

Youth Work Guarantee Scheme

KRA-RAP Kaderregeling Arbeidsinpassing – Reguliere Arbeidsplaatsen

Framework Law on Labour Integration – Standard Jobs

KRA-WEP Kaderregeling Arbeidsinpassing – Werkervaringsplaatsen

Framework Law on Labour Integration – Workexperience Jobs

KRU Kaderregeling uitzendwerk

Framework Law on Temporary Work

MOA Maatregel Ondersteuning Arbeidsinpassing

Enactment on Supporting Labour Integration

PALL Premieregeling Aanvaarding Lager Loon

Regulation on Bonusses for Acceptance of Lower Wage

Act on Premium Differentiation and Market Competition in the Disability

RWW Rijksgroepsregeling Werkloze Werknemers

Social Assistance for Unemployed Workers

TAV Wet Terugdringing Arbeidsongeschiktheids Volume

Act on Reducing the Disability Volume

TBA Wet Terugdringing Beroep op de Arbeidsongeschiktheidsverzekering

Act on Reducing Disability Claims

TZ Wet Terugdringing Ziekteverzuim

Act on Reducing Sickness Absence

VU Vergoedingsregeling Uitzendarbeid

Compensation Scheme for Temporary Work

VLW Vermindering Langdurige Werkloosheid

Reduction of Long-term Unemployment

WAGW Wet Arbeid Gehandicapte Werknemers

Act on Labour for Disabled Employees

WAO Wet op de Arbeidsongeschiktheidsverzekering

Act on Disability Insurance

WIW Wet Inschakeling Werkzoekenden

Act on the Mobilization of Job-seekers

WULBZ Wet Uitbreiding Loondoorbetalingsverplichting bij Ziekte

Act on Extension of Obligation to Pay Wages in Case of Sickness

WVA Wet Vermindering Afdracht

Act on Reduction of Tax Contribution

WVM Werkverruimende Maatregel

Work Increase Enactment

WW Werkloosheidswet

Unemployment Act

ZW Ziektewet

Sickness Act