Employment strategies for Europe: lessons from Denmark and the Netherlands.
Over the last twenty years, labour market reform has been seen as an important policy tool aimed at improving economic performance in the United Kingdom. It is widely thought that it has had a major effect on the way the British economy has worked, leading to increased flexibility of the whole economy. The last two decades have also witnessed a rise in income and wage inequality, a growth in social exclusion and long-term unemployment, with specific groups of the population, and especially those with fewer skills and experience, being particularly adversely affected. With the implementation of the New Deal programme, the United Kingdom is embarking on a new wave of structural labour market reform designed to deal with some of these problems and tackle wider issues by activating and motivating further underprivileged groups (young people, long-term unemployed).
The UK government has begun a series of reforms that are designed to increase employment opportunities and increase both incentives and abilities to take on paid work. The Welfare to Work scheme is designed to give the unemployed help in finding work, and, if need be, it provides subsidised jobs for the young unemployed and for those who have been unemployed for six months or more. There is also job search assistance for the disabled and for lone parents. This scheme is based on in-depth interviews that are designed to galvanise the unemployed, and assist them in job search. They are also presented with alternatives that are exhaustive, and do not include continued unemployment. The scheme involves both assistance to find a job and disincentives to stay registered unemployed. This can easily lead to a fall in unemployment matched by a rise in inactivity, much as happened in 1996 with the introduction of the Job Seekers Allowance interviews. The government has recognised this problem and it has introduced tax and social security reforms that are designed to make work more attractive for those with low earnings potential. In particular the Working Families Tax Credit and the Disabled Person's Tax Credit improve the attractiveness of employment for those covered by these schemes. The intention of these reforms is to increase pressure on individual unemployed persons to leave the register, and to increase the returns from taking paid employment when they do.
Other economies, such as Denmark and the Netherlands, have been following a similar path to try to improve their economic performance and we can learn from both their successes and failures. Both have had New Deal style programmes in operation for some years, and since 1994 the Danes have been experimenting with something almost identical to that in the UK. Both have carried out in-depth economic reforms, including market liberalisation, social security system revamping and strategic changes in their labour market policies. This process deeply altered the nature of the Welfare State and the relation of the state to employment and income guarantees. All three are now displaying unemployment rates well below those in the rest of Europe, as can be seen from Chart 1.
Our objective in this article is to discuss the effects of these reforms on labour market performance. We discuss recent labour market performance in Denmark and the Netherlands, and draw up a catalogue of the major reforms these two countries have carried out. We wish to analyse two sorts of reform, those which are designed directly to create jobs, and those designed to make the labour market more flexible, and hence raise the efficiency of its operation. These two goals can sometimes be in conflict, as a more flexible labour market may involve job shedding or a loss of job security, and job creation may involve reductions in flexibility either through reducing competition or increasing rigidities. We will also stress the distinction between increasing incentives to work and changing incentives to participate. Unemployment performance can be improved either by persuading the unemployed and the nonparticipants to take up work, or persuading them not to register and participate in the labour force. It is important to discuss what Danish and Dutch reforms have done in this respect.
We will look at the effects of labour market transitions on employment and unemployment, stressing the need for reforms to increase flow rates into employment, and hence raise activity in the long run, rather than raise inflows into inactivity which may cosmetically reduce unemployment in the short run. Labour market flow rates for these three countries are tabulated in the Appendix. These flow rates are revealing, as it is common for New Deal style reforms to raise the flow into inactivity, as they did in the UK in 1996 and Denmark in 1994. Labour market reforms have to induce the inactive to flow into employment, and the unemployed to return to employment. Reforms in the two countries we study have had mixed effects, with the Dutch raising activity rates and reducing unemployment, whilst the Danes have raised inactivity. The Dutch have been particularly successful at introducing family friendly reforms that have raised (albeit low) female participation, partly through an increase in the ability to create and maintain part-time employment, and hence reduce average (but not overall) hours worked.
1. A significant improvement of classic labour market performance indicators
As reform of active labour market policies and education has been proceeding at a rapid pace since the beginning of the 1980s, labour market performance in the Netherlands and Denmark has distinctly improved and has been well above the European average. Unemployment rates have trended down. Chart 1 gives the quarterly evolution of the OECD standardised unemployment rate in Denmark and the Netherlands, and compares them with the UK and with the European Union average.(1) The Netherlands experienced a trough in GDP growth in 1993, and unemployment rose slightly in 1994 and 1995 but, overall, the trend is clearly decreasing, passing under the 4 per cent mark early in 1998. Denmark experienced a slowdown in early 1990, which partly explains the rising trend of unemployment prior to mid-1993. Since 1994, when labour market reforms began to be implemented, unemployment has fallen steadily to 4.5 per cent of the civilian labour force in the second quarter of 1998.
This fall in unemployment is mostly due to a rise in employment growth. Chart 2 gives an employment growth index (1980 = 100) for both the Netherlands and Denmark and compares it with employment growth in the UK and the European Union. The number of employees in employment in the Netherlands grew at an average rate of about 1.2 per cent a year between 1980 and 1998. The corresponding figure for the European Union (15 current countries) was only 0.2 per cent. In terms of employment growth, the Netherlands has clearly been outperforming the rest of Europe by a considerable margin during the last twenty years. Matters are less clear for Denmark but, unlike the Dutch, who started to implement wage moderation in 1982, the Danes started their labour market reform in 1994. If we focus our analysis on 1994-8, employment growth in Denmark has picked up to an average rate of 1.4 per cent a year, compared with 0.5 per cent a year for the rest of Europe. However, as can be seen from the chart, this may just be a repeat of the cyclical pattern seen in the 1980s.
Unemployment is unevenly distributed among age groups and generally those under 25 years of age are much more affected than other groups. Chart 3 presents the incidence of youth unemployment in Denmark and the Netherlands and compares them with the UK and the European average. Both countries clearly experienced a significant fall in youth unemployment in the last two decades, although the rate of decline was not constant in Denmark. The UK has also seen falling youth unemployment. The fall in the unemployment rate for the younger group of the population can be seen as one of the main consequences of improvements in labour market policies. Indeed, there seems to be quite a clear improvement of the labour market performance in Denmark and the Netherlands, especially when judged by the improvement in unemployment without a continuing fall in participation over the last decade.
2. A gradual implementation of structural reforms
At the beginning of the 1980s, both Denmark and the Netherlands faced serious problems in the aftermath of the two oil shocks: large current account deficits, strong inflationary pressures, widening public sector deficits and rising unemployment. These mounting macroeconomic imbalances urgently needed to be addressed and both countries dramatically re-oriented their economic policies in the course of the 1980s, and especially in the 1990s.
The early stages of Dutch economic policy reform relied on voluntary wage moderation, agreed between social partners. The first agreement, at Wassenaar, in 1982, traded working time reductions against lower taxes and social charges for employers and a significant slowdown in wage growth. Minimum wages were also reduced, initially sharply, and they have been frozen for much of the period, and hence have continued to fall relative to the modal wage ever since. Many social benefits are linked to the minimum wage, and hence they have also fallen in relative terms. The agreement was designed to create jobs by reducing real wage growth, and this has been uniquely successful. Slower real wage growth produced significant improvements in international competitiveness, with Dutch labour costs increasing significantly less than in partner countries. Between 1983 and 1998, trend unit labour costs declined at an annual rate of 1.2 per cent, whilst Danish labour costs growth was 2 per cent per annum, and British labour costs grew at 2.56 per cent per annum (Chart 4 - all figures in a common currency). OECD labour costs grew at almost 3 per cent on average between 1983 and 1998.
However, real wage growth per person hour has been maintained in the Netherlands, albeit at only 0.7 per cent per annum over the period 1982-97 as hours have fallen but the number of employed persons has risen more rapidly. This real wage growth can be compared to 1.3 per cent in Denmark and 2.1 per cent in the UK, as can be seen from Chart 5. The reform process was not without its problems, and there were significant, but unsuccessful, public sector strikes in 1983 in protest against the application of 'wage moderation' to civil service salaries.(2) Strife of this nature is almost inevitable when policies are designed to reduce the share of labour in national income and to raise the share of profit in order to increase productive investment.
The revamping of social security systems
The deterioration of public finances was an important trigger for reforms in Denmark and in the Netherlands. By the early 1980s, about two thirds of Dutch GDP was spent or redistributed by the government and for every ten persons in employment, there were about eight on social benefits. The situation deteriorated through the 1970s up until 1984, when the number of social security beneficiaries rose to 3.7 million, as compared to 2 million in 1970. The level of employment dropped from 4.6 million to 4.2 million over the same period. Social security beneficiaries continued to increase in number to 4.2 million in 1994, albeit more slowly than total employment which rose to 5 million. After 1994 the dependency rate began to fall from around 83 per cent to below 75 per cent at the end of 1998.(3) This fall after 1994 was mainly the result of the effects of measures to make unemployment and disability benefits less generous or more stringently applied. Qualifying conditions for unemployment benefit were tightened in 1991 and 1995 and voluntary unemployed individuals were effectively excluded after 1995.
The disability insurance regime was the most likely to be reformed because the numbers of benefit recipients involved had risen by more than 30 per cent between 1980 and 1990. In 1990, disability benefits claimants represented 7.6 per cent of the workforce. There was a latent defect in the system in that employees and employers both had an interest in using this insurance to avoid redundancies. Individuals could be declared disabled by a firm, with the support of a doctor, if they were no longer deemed to be able to work in their current occupation, and not because they were unable to work. Employers could thus avoid the financial fallout from redundancies while employees were granted compensation higher than unemployment benefits. In the early 1990s, measures were taken to redefine and tighten the eligibility criteria and to fine employers over using the scheme. An employer hiring or retaining a 'disabled' person became entitled to a bonus representing six months of salary. Conversely, in the case of an employee being declared disabled, the employer was obliged to pay a penalty to the disability scheme, which was initially also set to six months of the salary of the disabled person. This enabled the government to break the trend rise in disability scheme claimants: the January 1994 peak of 925,000 claimants dropped down to 861,000 one year later, thanks to both a decrease in inflows into the scheme and a pick-up in outflows. However, they subsequently turned positive again, and further reforms were needed. In 1998 a new form of disability insurance was introduced in order to induce employers not to declare workers disabled. Employers now have to pay premiums, either to the government or to a private insurer, relating to the five-year claimant history of workers whose employment they terminate in this way. The objective is to give firms an incentive to rehabilitate those it has declared disabled and also to give them an incentive to combat disability in the workplace. This initiative contributed to the fall in the dependency ratio in 1998.(4)
However, as can be seen from Table A2, 98 per cent of those who are over 55 who are inactive are likely to remain so the next year, and there has been little change in this since 1994. Inactivity also remained quite an 'absorbing' state for the prime age group with more than 80 per cent of those who were inactive remaining so between 1994 and 1997. This rise, from levels of 67 per cent in the early 1990s, if sustained, will eventually lead to lower overall participation rates and has to be addressed in the ongoing reform process.
Denmark faced similar problems, as its benefit systems were very generous. In 1982, unemployment and sickness benefits were frozen for two years. Later, the generosity of unemployment compensation was reduced, as the maximum duration dropped from nine to five years in 1993. Denmark, like the Netherlands, still remains a country with one of the highest public transfer levels among OECD countries, with average income replacement ratios rising to 90 per cent for an unemployed low-wage worker.(5) In Denmark, pension transfers have increased following the demographic trend, but transfers to the working-age population have been rising continuously. The number of social benefits recipients rose from less than 20 per cent of the working-age population in the early 1980s to almost 30 per cent in 1996. Several steps have been taken: the administration of transfers has been decentralised and municipalities are now required to investigate sickness spells over eight weeks (instead of thirteen prior to 1992). Unemployment benefits are now subject to the claimant's participation in one of the active labour market policies. Increased stringency has also been extended to social assistance benefits, where since July 1998 all individuals have to have a locally administered action plan within a year of entering the scheme (thirteen weeks for those under thirty), and all recipients who can be registered as unemployed now have to register, and be available for work.
Measures to relax institutional rigidities
In both countries, relaxing institutional rigidities made up an important part of overall structural reform, and progressive but rigorous market liberalisation and deregulation policies were implemented. Both countries had had a long tradition of restarting competition in many goods and services markets as well as in the public sector. Deregulation, or even privatisation, has been proceeding at a fairly rapid pace in most utilities sectors since the beginning of the 1990s. The Kok cabinet implemented a number of reforms in the Netherlands after 1994, and competition legislation has been gradually relaxed and rationalised in order to allow more flexible operation. Important changes include changes to shop opening hours in 1996 in the Netherlands for instance, and regulatory obstacles have been removed in some markets. As a complement, Danish and Dutch authorities introduced a number of measures to promote labour market flexibility.
In the late 1980s, the Dutch government streamlined the administrative clearance required for laying off employees, relaxed hiring procedures and extending the maximising period from fixed-term contracts to one full year. Rules governing working-time became more flexible, in order to encourage part-time working as the most common model of work sharing. Legislation regarding overtime was also relaxed. A step further in the reform of job protection legislation was taken in 1996 and early in 1998 with a new law on fixed-term contracts, probation periods and dismissal procedures. It has notably eased restrictions on the renewal of fixed-term contracts with a maximum of two renewals in three years.
Comparatively, the Danish labour market has always been characterised by a lower level of legislative regulation. There has never been any legal limitation to working hours. Working time has always been regulated by collective agreements and, as in the Netherlands, it has been a major issue in collective bargaining during the 1980s. In 1990, the weekly working time was set to 37 hours in most agreements, which also provided arrangements for a flexible organisation of working time. The use of temporary and part-time work is also mostly unregulated. Employment protection and dismissal rules were tightened in 1989 to prevent employers laying off workers with little or no compensation. There are few limitations to the renewal of fixed-term contracts. The Danish authorities chose to increase the flexibility of the labour market with a move to the decentralisation of wage bargaining to fourteen regional Labour Market Councils, the increased involvement of counties and municipalities in the formulation of labour market policies and the introduction of management focused on objectives.(6)
More active labour market policies
Most importantly in the wave of structural reforms, the focus of Dutch and Danish labour market policy shifted somewhat from income support measures in the 1970s and early 1980s to more 'active' policies and stronger incentives to work. In the 1970s, 'active' labour measures were mainly focused on job creation with public support in labour intensive sectors such as construction, and on job preservation through subsidies to firms with financial difficulties. These types of measures were largely abandoned from the mid-1980s, being judged far too costly. In the early 1990s, 'active' programmes gained momentum in both countries. Active labour market policies can be divided into measures and actions directed towards persons without employment and those directed towards employed people. In the first case, the objective is to ensure better integration into the labour market through job subsidies, direct job creation or assistance for becoming self-employed. In the second case, they aimed at reducing the normal working time or the life working time, through various leave schemes, which have been used extensively in Denmark. In both cases, education and training remained central features of the programmes.
In the Netherlands, direct job creation and improved integration into the labour market are coordinated by the Public Employment Service, which was completely restructured in 1990 to make it both more decentralised and more needs orientated. A series of special employment programmes has been introduced. Among the most innovative and widespread are the 'Labour Pools' and the 'Youth Work Guarantee Plan'. 'Labour Pools' were introduced in 1990 and targeted specific groups with high unemployment rates and limited chances of finding a job. They were especially focussed on the long-term unemployed and involved the creation of jobs so that individuals could perform beneficial work for society. The 'Youth Work Guarantee Plan', which was begun in 1992, involved the creation(7) of 40,000 new jobs in the public sector and 20,000 in the private sector for those young unemployed who were threatened with long-term unemployment. Employers receive a GLD18,000 subsidy (a bit more than [pounds]3 per working hour) if they provide training and create permanent jobs. These jobs are for 32 hours per week and wages are lower than 120% of the statutory minimum wage.
The Youth Work Guarantee Law was primarily designed to aid the transition to a permanent job. However, a first assessment of the scheme, in 1993, underscored the low proportion of young people actually getting a job when leaving the scheme. Therefore, in 1994, the programme was complemented by a training scheme for young people who were inadequately prepared for a move directly into a normal job. In 1996 it was decided that reintegration should be further promoted, and the government began to create new subsidised jobs for the long-term unemployed. These 'Melkert jobs' (named after the Minister of Social Affairs and Employment) include:
* the creation of 40,000 permanent new public jobs in the fields of security, health and child care, maintenance of public areas and monuments and education;
* the creation of temporary jobs, mainly in the private sector, thanks to agreements between the employment services and temporary work agencies;
* the compulsory activation of long-term benefit claimants who have a very limited prospect of finding paid work; and
* the experimental exploration of the market for domestic services. Private cleaning services companies can receive a subsidy when hiring a long-term unemployed person with social security benefits and conclude cleaning contracts with private households.
These programmes are mostly financed through the related savings in social security benefits and partly through a contribution from the central government (around 9 billion guilders - 1.3 per cent of GDP - have been earmarked for these purposes for the 1995-9 period).(8)
The Danes began to introduce new policies in 1994 based around the concept of 'activation'. These reforms resemble recent policy initiatives in the UK in that a new emphasis was placed on activating the unemployed, particularly youths with an incomplete education. The activation programmes place a 'right and duty' on an individual to undergo activation, either through a private or public job offer, education or training. All Danish unemployment benefits or social assistance recipients are now obliged to take part in compulsory activation, starting after thirteen weeks unemployed for everyone up to thirty years of age, and after two years of unemployment for the rest of the population. Activation is administered by municipalities and has to be underpinned by an individual action plan for all over 25 years old. This involves a 'New Deal' style personal interview. This strategy has been described as an "effort-commitment strategy" (Thustrup and Tranaes, 1997), designed to change the choice facing the individual.
The activation programme places a right and duty on individuals to undergo activation, and the incentive structure is designed to persuade participants who can that they should search for a job that allows them to leave benefits. Individuals receive the hourly rate from their last regular job, but only up to a specified maximum, which is normally the level of benefit they would receive. Hence many activated persons will be no better off than when on benefit but will have to work part or all of the week. Hours would generally be less than in normal employment. Individuals cannot exit from these programmes into unemployment and receive benefits, and hence they have a greater incentive to actively find or keep a job.
Part of the incentive to work involves a job placement interview, much as in the New Deal. This in depth discussion appears to have a positive effect on the attitudes and motivation of the interviewee, especially among the long-term unemployed. Early evidence (see Langager, 1997, and correspondence with Larsen of the Social Forsknings Instituttet) suggests that these interviews have had an effect on participation decisions and are one of the more positive effects of the programme.
However, activation can lead to low incomes, especially for the unskilled. A 1997 survey among Danish employed members of unemployment insurance funds indicates that 10 per cent have a lower income from work than they would receive if unemployed, a share that had increased from 7 per cent in a similar survey in 1994.(9) This disincentive to work is being tackled in the recent tax reform (1998-2002) which targets taxes on labour.
Efforts have also been made to tackle long-term unemployment and results have been encouraging since the implementation of the 1994 reform and the compulsory activation programmes, and long-term unemployment has fallen by more than 2 percentage points. In the Netherlands, in contrast, long-term unemployment has been tackled mainly with passive labour market measures: public job creation and wage subsidies for hiring a long-term unemployed person (more than two years). So far, there are few improvements and long-term unemployment has not followed the decreasing path of unemployment in general. This tends to suggest active measures to persuade the long-term unemployed to take work can in some sense be seen as more 'desirable' than the creation of public sector jobs in order to soak up the long-term unemployed.
Extensive use of leave schemes
A first revamping of the labour market policy took place in 1993 in Denmark and it greatly facilitated outflows from employment or unemployment to inactivity, as can be seen from Table Al. First, early retirement was encouraged, resulting in a 16 per cent increase in the stock of pensioners between 1986 and 1995. There was an even larger outflow from the labour force in 1995 when 47,000 workers (1.7 per cent of the Danish labour force) chose to retire just before the scheme was abandoned. Three popular leave schemes were also created in 1993: the Childcare Scheme, the Sabbatical leave scheme and the Educational leave scheme. Overall, these paid-leave schemes led to a large initial take-up (3 per cent of the labour force) and a tightening of the labour market, with an increasing shortage in the skilled labour force.
Fearing increasing labour market imbalances, the Danish government swiftly changed course in 1994. Paid-leave schemes were made less generous; access to early labour market withdrawal benefits for the long-term unemployed in the age range 50-59 were abolished; the maximum duration of unemployment benefits was reduced from nine to five years; and the eligibility criteria were tightened. Subsequently a tax reform programme phased in over the 1994-8 period gradually lowered the taxes on labour and hence increased incentives to work for the low-paid.
Education and training
Despite a high level of public spending on education and training (9 per cent of GDP), the Danish economy had to face shortages of skilled workers in the 1980s. Moreover, results of tests of academic proficiency in Denmark appeared to be on the middle to low side in international comparisons. One of the major problems was the relatively low degree of skill acquisition, due to a lack of motivation and incentive to complete upper secondary level education and to pursue higher studies. As a result, the share of population with uncompleted secondary or primary education reached 20 per cent, ranking Denmark somewhere just above middle position in a comparison across OECD countries.(10)
In response to this weakness, the 1990s have been a period of extensive educational reform in Denmark. Measures have been taken to improve the education system and lower the number of dropouts from formal education. The Danish authorities have broadened access to existing institutions and stimulated individuals, notably young people, to continue their formal education. For example, the membership of an insurance fund (and hence entitlement to benefits), for a young person who has never been employed, is granted if this person has undergone post-school education or training of at least 18 months. There is also extensive public involvement in the provision of adult education and training. In 1995, 1 per cent of GDP was granted to these programmes and 13.7 per cent of the labour force were participating, the largest figures in all European Union countries.
Denmark has had a long tradition of apprenticeships, and hence the employment probability for new school leavers has always been among the highest in the European Union (more than 75 per cent on average). But, as employment prospects for younger workers worsened at the beginning of the 1990s, Denmark introduced wide-ranging changes in its active labour market policies to target youths. After six months of unemployment, an individual action plan is formulated for persons under 25 years of age. They are offered either 18 months of education, on-the-job training or a two-year grant to start their own firm. Since 1996, a young person under 25 has been unable to receive unemployment benefits for more than six months and then at only one half of the normal level. In addition, since 1996 unskilled youths have received even lower benefits in order to induce them to take up and successfully complete normal education or undertake a skills course.
Similar concern has also emerged in the Netherlands, as recurring labour shortages in certain technical fields appeared whilst there was a concentration of unemployment among university graduates in other fields. This suggested further attention needed to be given to vocational programmes. In 1995, the Dutch government implemented a new law on education and vocational training. This included the creation of regional training centres for basic education, the strengthening of the practical aspect of vocational training and general adult education. In 1996, a tax advantage was granted to firms employing apprentices. This scheme was designed to create 18,000 new apprentices by the end of 1999.(11)
Balance between increased labour market flexibility and social cohesion
Denmark and the Netherlands have embarked upon a fairly extensive programme of reforms, and the UK is in many ways following them. Labour market policies lay emphasis on the principle of individual responsibility with reinforced controls, tightened access to unemployment or disability compensations and more active measures to stimulate job creation and 'activate' inactives. Overall, reforms in both countries increased the flexibility of the labour market with the encouragement of part-time work and fixed-term contracting, wage moderation and working time reduction.
So far, the reforms have not thrown the foundations of the welfare state into question. Indeed, both countries have maintained a comprehensive social security system, with high-income replacement ratio benefits, and governments acting as provider of last resort to secure a relatively generous minimum standard of living, despite the tightening of the last two decades. Although unemployment has fallen in the Netherlands, for instance, there has been a continuing increase in those on unemployment benefit not seeking work, as can be seen from Chart 6.
There is a strong safety net provided by the Dutch and the Danish states. Despite all recent efforts, there still is very high welfare state dependency and a potential benefit trap, which is still to be reduced. Despite recent efforts in the Netherlands to lower the proportion of people falling into inactivity and benefit dependency, inactivity seems to have become a more and more absorbing state when one analyses the flows from one labour market status into another (Appendix A). The probability of staying inactive from one year to another in the Netherlands has increased by 19 percentage points since the beginning of the decade for the 25-54 age group. This progression has been slowed down for the younger age group whose probability of staying in inactivity has only increased by 8 percentage points since 1990, mostly as a result of policies specifically targeting young people.
3. A preliminary assessment
We should assess these reforms not only in terms of their effects on job creation but also in terms of overall labour market performance (labour market participation rates, discouragement and quality of work found). It is too early to evaluate these programmes thoroughly, because most of them have only been recently implemented. A few preliminary conclusions may however be drawn.
We have seen that unemployment has decreased dramatically in both countries, especially as compared with the European average. In Denmark, attempts have been made to give a preliminary examination of the activation programmes. Andersen and Bach (1998) suggest that the effects have so far been small, and a large proportion of those affected are only influenced in the short term. The question still remains about the 'after' activation programme status. The effects of job training in the private sector appear to be more positive than those in the public sector, however. Retrained persons in the private sector have a tendency to stay on after periods of training when their subsidy stops. Private sector training appears to be more effective at re-skilling individuals and hence reintegrating them into the workforce. However, this could just reflect the fact that private sector schemes are likely to screen applicants with long-term attachment in mind. This is particularly probable as private sector schemes have absorbed only a very few individuals, and it is hard to pick out any overall effects.
Active labour market policies often have disadvantageous aggregate effects, in that they reduce the pool of available labour whilst leaving effective output unchanged. However, Hansen and Tranaes (1997) suggest that the effort requirements in the Danish schemes protect against this, as participants do not reduce their job search intensity whilst on the schemes. There has been no significant re-emergence of skill shortages in the economy in the 1990s, despite falling unemployment, and this may in part reflect the incentive effects of the new activation schemes. However, there is to date little evidence that these schemes have actually improved the performance of the economy and increased output.
Young people in Denmark seem to have responded to changes in economic incentives, and completion rates in formal and vocational education have risen, and search intensity for non-subsidised jobs has also risen. Changes in benefits for those without completed education seem to be effective and have reduced unemployment among the young under-educated without reducing inactivity, as can be seen from Chart 7, and have also begun to improve educational levels in the workforce.
The overall effects of these schemes can be seen from the labour market flows in the first table of the Appendix. By 1997, flow rates out of youth unemployment and inactivity had risen to around a half compared to a third in 1993. Retention rates in employment have been much higher since 1994 than before the reforms started. Schemes like this raise the attachment of youths to the workforce and raise their long-term chance of employment.
Studying labour market flows suggests there has also been improvement for the rest of the working-age population. Inactivity has become a much less absorbing state since the beginning of the Danish labour market reforms in 1994. Inactive people from the prime age group (25-54 years old) have had an increasing probability of flowing into employment (from 15 per cent in 1990 to 24 per cent in 1997).
In the Netherlands, public employment programmes and education have also been provided on a large scale, although employment probabilities of new school leavers are lower than in Denmark (52.5 per cent in 1996). However, youth unemployment has declined over the last two decades, as can be seen from Chart 3. Once again, wage moderation is a useful tool of the Dutch policy as young people under 22 years of age can be paid as little as 85 per cent of the legal minimum wage. This is confirmed when studying the labour market flows (Table A2). Clearly, unemployment has become a much less absorbing state for the whole working-age population, and this is particularly true for the younger age group.
However, we should not judge performance in terms of unemployment and employment changes alone. An important part of the working-age population does not actually participate in the labour force for various reasons, and the evolution of non-participation can be symptomatic of changes in labour market policies.
Labour force participation and inactivity
Danish non-participation in the labour force has traditionally been much lower than anywhere else in Europe, and it has the lowest non-participation rate in the European Union, averaging 20 per cent of the labour force, with only Sweden experiencing similar levels. Women have a particularly high participation rate in Denmark, with around 75 per cent of the female population taking part in the labour force. The early 1990s saw a large increase in non-participation in Denmark, as can be seen from Chart 8, which plots activity rates for men and women. This rise in Danish non-participation is not only relative to a low level, but also linked to a specific change in policy measures, notably with the introduction of generous leave schemes between 1993 and 1994, which, as we discussed earlier on, attracted an significant share of the labour force.
However, at the same time, 'activation' measures have been taken to put inactive people back to work. This can be seen from an analysis of the flows of workers between employment, unemployment and inactivity since the beginning of the labour market reform in Denmark in 1993. Chart 9 plots the transition probability from being inactive in the year (t-l) to being employed in the year (t), for two age groups.
Inactivity has become a much less absorbing state in Denmark, as the probability of staying inactive from one year to another has fallen from 0.8 to 0.69 between 1990 and 1997. At the same time, people are actually going back to work, with the probability of finding employment when inactive increasing by 10 percentage points on average since 1990.
This statement has to be qualified by observing what has been happening to unemployed people. Looking at the prime age group before the reforms, an equal share of 40 per cent of the unemployed had a chance either to become employed or to remain unemployed. It appears that reforms, although lowering the probability of staying unemployed to 0.3, actually facilitated flows towards inactivity. In 1997, an unemployed person still had a 40 per cent chance of finding employment within a year, but now had also a 30 per cent chance of becoming inactive. This evolution is even more obvious for the older age group, as shown in Chart 10: the transition probability from unemployment to inactivity increased from 0.16 in 1990 to 0.58 seven years later.
This would explain the overall increase in non-participation in Denmark since the beginning of the 1990s. Although policies have successfully helped inactive people to find employment, they also appear to have reduced unemployment and the costs of the benefit system by raising inactivity. In October 1998 further reforms were introduced that aimed to reduce early withdrawals, amongst other things. In particular, 50-54 year olds had their right to up-to-ten years unemployment benefits removed, and the total length of unemployment benefit receivable by any group was reduced from five to four years. Compulsory activation was extended to all groups after twelve months. These measures are designed particularly to reduce early withdrawal from the labour market, and are consistent with reforms to the disability benefit system made earlier.
Conversely, the Netherlands has always been characterised by a low participation rate, notably among women. Only 55 per cent of women actually participated in the labour force in 1998 and, although participation has been on the increase over the last fifteen to twenty years, it is still far below other European countries which average 70 per cent. Overall, the Dutch participation rate has increased steadily for the last twenty years, to reach 63 per cent of the working age population in 1998, as shown in Chart 11.
Looking at the transition probabilities from one labour market status to another, however, it appears that inactivity has been an increasingly absorbing state for the entire working-age population, despite a renewed effort to curb the number of social benefits recipients in the Netherlands. This is particularly clear for the prime age group whose probability of staying inactive rose from 0.68 in 1990 to 0.87 in 1997 (see Table A2). In addition, flows from unemployment to inactivity have also increased quite dramatically in all age categories. These flows more than doubled for people under 55 between 1990 and 1997, reaching a probability level of 0.5 in 1997.
Labour market flexibility and its effects
We have seen that another important policy change was to encourage flexibility of the labour market. This has been enhanced by the continuous increase in part-time and temporary employment, especially in the Netherlands. Charts 12a and 12b plot the evolution of temporary and part-time work over time in both countries.
* In Denmark, the share of temporary workers in total employment is slightly above the European Union average but has remained rather stable, even on a slight decreasing trend, over the last twenty years, averaging around 12 per cent since the early 1990s.
* In the Netherlands, however, temporary work has been following a different path. Without being exceptionally high, the incidence of temporary employment in the Netherlands doubled between 1980 and 1994, whereas it remained fairly stable in most other European countries (European Commission, 1994). Dutch employers seemed to make more extensive use of temporary work agencies than any other European country and this has increased substantially since the beginning of the 1990s due to the relaxation of legislation regulating the role of temporary work agencies (Bovenberg, 1997).
Part-time employment is another atypical and more flexible form of employment, and Charts 12a and 12b describe its evolution in both countries. It has remained fairly stable, and even fallen in Denmark, although this might change in the near future: part-time voluntary work is now encouraged for persons who satisfy the conditions for joining the voluntary early retirement pay scheme, in order to ensure a gradual withdrawal from the labour market.(12) In the Netherlands, however, part-time working is much more common than elsewhere, and it is twice as high as the European Union average. Since the beginning of the 1980s, the Dutch government has supported the use of part-time employment. Legislation has been amended to remove legal differences in status and conditions between part-time and full-time contracts. The growth of part-time employment in the Netherlands is mainly supply-side driven and strongly related to the rise of the female labour force since part-time employment meets the needs of women with young children who want to enter the labour market (Bovenberg, 1997). However, since female participation in the Netherlands remains relatively low, there may be further increases in such forms of employment to come, even though subsidies for childcare are currently being improved.
Another aspect of labour market flexibility, although closely linked to the organisation of work, lies in the use and duration of working time in the economy. In the Netherlands, reducing the standard number of hours has clearly been seen as an instrument to fight unemployment through redistributing work across people. During the 1980s, most Dutch collective agreements cut working time, mainly by increasing the number of holidays and encouraging part-time work. A renewed call for work sharing in the 1990s led to a further legal reduction of working hours. In the public sector, some 1997 agreements decreased standard hours by about 3.5 per cent. Also, in some parts of the private sector, unions and employers have agreed to further reduce the number of standard hours. As a result, the average annual number of hours worked per employee fell by 5.2 per cent between 1980 and 1995. At the same time, the total number of hours worked in the Dutch economy increased by almost 14 per cent, as shown in Chart 13b. This chart suggests that the reduction in the number of standard hours worked per employee successfully led to a rise in the number of workers, since the overall effort of the Dutch economy, in terms of hours, has constantly increased since the mid-1980s.
In Denmark there have also been political incentives to reduce working hours per employee. This can be seen from Chart 13a, where the average annual number of hours worked per employee dropped by 8 per cent between 1980 and 1996. This is a similar decline to that experienced in the United Kingdom (7 per cent) but much less than in the Netherlands (12 per cent). However, the total effort of the economy in terms of hours only regained its 1980 level in 1996. Denmark experienced neither such a strong downward trend in hours per head as the Netherlands, not did it see such a large increase in actual hours worked in the economy.
A 1998 OECD survey(13) shows that the average year-on-year decline in the Netherlands is entirely due to increases in the proportion of part-time workers, whereas virtually all the Danish working time decline is attributable to a fall in the hours of full-timers.
Part-time and temporary work enhance working time flexibility in regulating fluctuations in the workloads, but they can also indicate the low quality of the new jobs underlying improved employment performance, especially in the Netherlands. For example, there might be a growing number of involuntary part-time workers. These individuals would usually work full time but have been required to work part time for economic reasons, because they have had their working hours cut or simply because they could not find any full-time work. OECD studies(14) have shown that involuntary part-time workers are overwhelmingly female (typically between 70 and 85 per cent of involuntary part-timers) and in most cases could not find full-time work. In the Netherlands, the share of involuntary part-time workers in the labour force remained relatively stable, falling from 5.7 per cent in the mid-1980s to 5.6 per cent in 1991. This remains a higher proportion than for most other European countries where involuntary part-time work is around 3 per cent of the labour force on average. However, the high Dutch level of involuntary part-timers is partly due to the high proportion of part-time work in general. Also, this overall stability of the figures hides the rise in involuntary part-time working among men and a decrease among women. Given the significant rise in Dutch part-time work we can conclude that this increased flexibility has been largely voluntary.
The picture is not so good when looking at the number of discouraged workers in the Netherlands. Discouraged workers are those individuals who leave the labour force in the face of poor job prospects and others who decide not to enter because of poor prospects. These individuals would like to work but are not seeking work, because they feel no suitable job is available. Over the last twenty years, the number of Dutch discouraged workers grew from 0.16 per cent of the labour force in 1983 to 1.18 per cent in 1995. Once again, women are more affected than men. In Denmark, on the contrary, the number of discouraged workers remained fairly stable (under 0.4 per cent of the labour force on average) until 1991. Then, there was a sudden upsurge to 1.6 per cent as the economy was experiencing a slowdown of activity. Since 1994 and the implementation of labour market reform, the number of discouraged workers has been reduced by 35 per cent in less than two years. Overall, it seems therefore that increased flexibility and labour market reforms in Denmark and the Netherlands have not been significantly translated into an increase in the number of discouraged or involuntary part-time workers.
We have to be cautious about the lessons we can learn from Denmark and the Netherlands as models of labour market strategies for the rest of Europe. First, they have benefited from country-specific positive characteristics, which have helped them to attain falling structural unemployment. In the Netherlands, for example, the particularly low rate of female participation in the labour force is unusual in Europe. Because of Dutch neutrality during the first World War, women did not replace men at the front in factories, as women did in other European countries. In Denmark, the long tradition of apprenticeship helped to maintain a lower rate of youth unemployment than in the rest of Europe.
There are also reasons to believe these two countries have a high proportion of hidden unemployment. Benefits claimants are numerous and in the last decade the Dutch authorities have been trying to reduce their number without much success. The Netherlands still relies heavily on public job creation, which does not easily lead to integration into the labour force. Similarly, in Denmark, activation programmes may be leading to an increasing proportion of hidden unemployment, where a proportion of activated people flow from one programme to another and do not enter the labour force.
In both countries, the last few years of economic expansion have been characterised by relatively strong employment growth and falling structural unemployment, although improvements in Denmark are very recent and may be transitory. Activation programmes in both countries of the sort implemented in the New Deal in the UK are already bringing encouraging results, especially when implemented in the private sector. Persons activated in the private sector have a tendency to stay on after periods of training, when their subsidy stops (Jensen et al., 1996). It seems that the approach combining stricter eligibility rules for benefit claims and effective controls with appropriate training and counselling on the labour market help the return of some persons to the active labour market.
1. European wide data prior to 1993 are annual and from a Eurostat source.
2. See Visser (1998) for a discussion of these problems, and also those in 1990 and 1995.
3. See Don and Besseling, 'Social Security Reforms; How and why' CPB Report 1996-4 and CPB Report 1998-4 Appendix A.
4. See Besseling, Bovenburg and De Mooij (1998) for further details.
5. OECD (1998), Benefit Systems and Work Incentives, Paris.
6. PLS Consult (1996), Labour Market Studies: Denmark, Luxembourg, European Commission, pp. 41 and 96-103. 7. Number of creations up to 1996. The programme is still active.
8. See van Opstal, Roodenburg and Welters (1998) for a description and evaluation of these schemes.
9. OECD Economic Survey, 'Denmark', 1999.
10. OECD Economic Survey, 'Special feature: Education, Training and Labour Market Reform' in 'Denmark', 1997, table 22, p. 91.
11. OECD (1996), 'Netherlands', OECD Economic Survey, Paris.
12. Introduced in 1995, this new early retirement scheme covers workers aged between 60 and 66. They must reduce their working time by at least a quarter and be replaced by a younger worker and they are thus entitled to a partial early retirement benefit of 58 Danish Krones (around [pounds]5) per hour by which working time is reduced.
13. OECD (1998), 'Working hours: latest trends and policies', in Employment Outlook, June, Paris, table 5.2, p. 156.
14. OECD Employment Outlook, 1995 and 1993.
Andersen, D. and Bach, H.B. (1998), 'Virksomhedernes brug af lontilskud - og langtidlediges senere jobsituation', Copenhagen, Social Research Institute 98:12.
Besseling, P., Bovenburg, L. and de Mooij, R. (1998), 'Premium differentiation in social insurance', CPB Report 1998-1.
Bovenberg, L. (1997), 'Dutch employment growth: an analysis', CPB Report, 1997/2, pp.16-24.
Don, H. and Besseling, P. (1996), 'Social security reforms: why and how?', CPB Report, 1996/4, pp.13-17.
European Commission (1994), 'Employment in Europe', Luxembourg, Office for Official Publications of the European Communities.
----- (1997), 'Employment in Europe', Luxembourg, Office for Official Publications of the European Communities.
Jensen, P. and PLS Consult (1996), 'Denmark', Labour Market Studies, Luxembourg, Office for Official Publications of the European Communities.
Langager, K. (1997), 'Indsaten over for de forsikrede legide. Evaluering af arbejdsmakedsreformen I' Kobenhavn, Socialforskninginstituttet.
Opstal, R. van., Roodenburg, H. and Welters, R. (1998), 'Low skilled jobs through job creation and wage subsidies', CPB Report 1998-2.
OECD (1998), 'Netherlands', OECD Economic Surveys, Paris. ----- (1999), 'Denmark', OECD Economic Surveys, Paris.
----- (various years) Employment Outlooks.
Thustrup, C. and Tranaes, T. (1997), 'Effort commitment in active labour market policy', in Andersen, Jensen & Risager (eds) (forthcoming), Macroeconomic Perspectives on the Danish Economy.
Vidal, D. (1997), 'Miracle ou mirage aux Pays-Bas? L'Etat-Providence neerlandais en question', Le Monde Diplomatique, July.
Visser, J. (1998), 'Two cheers for corporatism, one for the market: industrial relations, wage moderation and job growth in the Netherlands', British Journal of Industrial Relations, 36, 2, pp. 269-92.
[TABULAR DATA FOR TABLE A1 OMITTED]
[TABULAR DATA FOR TABLE A2 OMITTED]
[TABULAR DATA FOR TABLE A3 OMITTED]
We should like to thank our colleagues Hilary Metcalf, Rebecca Riley and Dirk Willem te Velde for their comments and assistance in preparing this article. We have gained from discussions with a number of people in the Netherlands and Denmark, and we have greatly appreciated the help and written communications of Barthold Kuipers of the CPB-Netherlands Bureau of Economic Research and Mona Larsen of the Danish Social Forsknings Instituttet. Comments should be addressed to the authors at the National Institute of Economic and Social Research, 2 Dean Trench Street, Smith square, London, SW1P 3HE. Tel. no. 0171 654 1925, e-mail: firstname.lastname@example.org.
|Printer friendly Cite/link Email Feedback|
|Author:||Barrell, Ray; Genre, Veronique|
|Publication:||National Institute Economic Review|
|Date:||Apr 1, 1999|
|Previous Article:||Flexibility, quality and competitiveness.|
|Next Article:||Family-friendly management: testing the various perspectives.|