Feb 24, 1999 - It is cheap labour for the employers. The young people ... project. During the integration period, they should be able to learn a real trade and ... school3 (unemployed or not) in the EU have had no more than primary education.
oecd
NICAISE I., Labour market programmes for disadvantaged young people in Europe, in: OECD, Preparing youth for the 21st Century: the transition from education to the labour market, Proceedings of the Washington DC Conference, 23-24 February 1999, Paris, OECD, 1999
LABOUR MARKET POLICIES FOR DISADVANTAGED YOUNG PEOPLE IN EUROPE Paper presented at the OECD conference ‘Preparing youth for the 21st Century’, Washington, 23-24 February, 1999
Dr. Ides NICAISE 1 HIVA (Catholic University of Leuven)
1. INTRODUCTION “In my municipality, social services place young people in integration projects for a period of 18 months. They have to work for up to 8 or 10 hours a day in voluntary organisations for $30 per day. It is cheap labour for the employers. The young people are forced to accept it because they have no alternative and they want to be independent. Once the “integration” is over, they are back on the street and are not entitled to anything because the law does not give them the same rights as other employees: they have no pay slip, no social security contributions, no sickness benefit fund, no minimum wage. When they leave they cannot go on the dole because their work is not recognised as real work, it is just an ‘occupation’. What we are demanding for young people is that the work in these integration projects should give them genuine rights as employees, both during and after the project. During the integration period, they should be able to learn a real trade and obtain a recognised certificate. Afterwards, they must be entitled to unemployment benefits. If they are recognised as unemployed, it is easier for them to be recruited by employers, who are subsidised for re-employing unemployed people.” This is how a mother from an underprivileged family summarises her son’s situation and her vision of the integration policy (quoted in ATD Quart Monde 1998). 2.
WHO IS INVOLVED?
The image of “disadvantaged young people” is somewhat fuzzy. We are dealing not only with unemployed young people who have to be “integrated”, but also with young people who started working when they were too young, for example, with no form of social protection. In some southern European countries, young people aged between
1
My special thanks to Nick Matheus for bibliographical assistance, and to Eurostat for providing unpublished data.
12 and 16 from needy families drop out of compulsory schooling in order to earn an additional income.2 In other cases, young adults try to earn a living or simply make themselves useful doing “informal work” (which is not the same as illegal activities), e.g. by repairing televisions or mopeds, doing voluntary work in their neighbourhood, etc.. It is therefore best not to measure “disadvantage” one-dimensionally (by unemployment), but as an accumulation of deprivations, even if we confine ourselves to the labour dimension. Each of these elements in itself delineates a fairly broad group, while the intensity and duration of the deprivation, together with the accumulation of factors, leads to genuine disadvantage. Unfortunately, the statistics (mostly derived here from Eurostat’s Labour Force Surveys, 1996 and 1997) are too general to give an accurate picture of our target group: - level of education: approximately 12.2% of the young people who have left school3 (unemployed or not) in the EU have had no more than primary education. Even though more than half of them are employed, these youngsters are faced with extremely high risks of unemployment and poverty at later ages.
Figure 1. Proportion of young people who have left school, by level of education (EU14, 1997)4 100% 80% More than low er sec.
60%
Low er secondary 40%
Primary or less
20% 0%
2
3
4
In Portugal the government introduced minimum labour legislation in 1993 for young people of compulsory school age because they could not stop them dropping out. Occasionally, serious work accidents involving young people under the age of 15 crop up in the European media. Young people aged between 15 and 24 who have left school (note that this definition implies an underestimation of graduation rates at upper secondary and tertiary level). The percentage is less than 5% in Austria, Denmark, Sweden and the UK and over 15% in Spain, France, Greece and Luxembourg. It exceeds 50% in Portugal. Figures for The Netherlands are not included because they are unreliable
2
Figure 2. Employment status of young people who have left school, by level of education (EU15, 1997)
100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%
inactive unemployed
-
TO TA L
te rti ar y
ec on up da pe ry rs ec on da ry
er s lo w
pr im ar y
or le ss
employed
social protection: only 23.3% of unemployed people under the age of 25 in the EU were receiving unemployment insurance or assistance benefits in 1994 (CEC, 1995, p. 88; Eurostat, 1996).5 It is not known whether these are better-placed or more underprivileged unemployed people: unemployment insurance favours those who have already worked long enough, while unemployment assistance is selectively targeted at families on the lowest incomes. Note, however, that in many countries young people also have no access to means-tested (public assistance) benefits;
Figure 3. Percentage of young unemployed covered by unemployment insurance or assistance
80,0% 70,0% 60,0% 50,0% 40,0%
5
I
P
EL
E
N L
F EU 15
S FI N D K
D
U K
A
IR L
B
30,0% 20,0% 10,0% 0,0%
CEC (1997) gives higher figures based on the EC Household Panel. However, these figures seem to us less reliable because of the limited size of the sample and possible selection biases in the calculations.
3
- duration of unemployment: 37.8% of young unemployed people in 1996 had already been without work for longer than one year, 20.5% for longer than two years; - regional disadvantage: both youth unemployment in general and long-term youth unemployment are highly geographically concentrated in southern Europe.6 While Finland too may well have high youth unemployment, the proportion of long-term unemployment in this country remains limited; - gender seems not to be a highly discriminating factor among young people. In some EU countries, the level of unemployment among young women is actually lower than among men, although this is compensated by a higher level of inactivity. Young women are more likely to work part-time than young men, but the differences between the two sexes are much smaller among young people than among those aged over 25; - qualitative aspects of employment: part-time and temporary work are not typical youth concerns (EC, 1998): they are actually more common among adults who go back to work after a period of unemployment. On the other hand, the labour market for young people is characterised primarily by low wages: the gap between them and higher age groups is increasing (OECD, Employment Outlook, 1996). Youth employment is also concentrated in sectors with higher flexibility requirements (hotel and catering, retail trade, building and personal services). Presumably, young people with the lowest qualifications are most likely to work in temporary, part-time, flexible and low-paid jobs, although no specific data are available to support this. - Other factors of disadvantage include ethnic origin, health problems, handicaps, previous institutional history, a criminal record, etc..
3. OBSTACLES TO PARTICIPATION IN LABOUR MARKET PROGRAMMES Underprivileged groups in general (and to a less extent underprivileged young people) are paradoxically heavily under-represented in programmes for re-integration into the labour market: although they would benefit the most from them, they are the least likely to take part. Statistical material about this issue is widely available, although harmonised data can only be found for training measures (see Figure 4). In this section, we look for possible explanations for this phenomenon.
6
Both youth unemployment in general and long-term youth unemployment are expressed here as a % of the total population aged between 15 and 24 (therefore including those still studying. Cf. EC, 1998 for a justification of this definition: suppose only 10% of those in a certain age group are available on the labour market and half of them are unemployed, this is less of a problem than if 50% are available and 1 in 4 of them is unemployed. In the former case, we are dealing with 5% youth unemployment, in the latter with 12.5%).
4
Figure 4. Participation of 15-24 year old individuals who have left school in vocational training programmes (EU14, 1997)7 7,5 8 6 % 4
4,3 2,5
3,3
2 0 primary
lower upper secondary secondary
tertiary
The role of cultural thresholds (the attitudes of young people towards work) in the decision to take part in active labour market policies (ALMPs) must be put into perspective. In contrast to what is often expected, young people in general and unemployed young people in particular attach at least as much value to work as other age groups (OECD 1996, pp. 113-117). So what are the genuine obstacles restricting access to active labour market policies (ALMPs)? An initial group of stumbling blocks is legal or administrative in nature. Given that many young people cannot receive benefits, many of them do not register as jobseekers (although they are looking for work) or withdraw from the labour market. In addition to the 3.5 million young people in the EU who were registered as jobseekers in 1996, the Labour Force Survey counted another 1.3 million unemployed but not registered and 2.9 million young people (1/3 of whom are men) who are “inactive” for reasons other than receiving education. Out of every four young persons declaring themselves unemployed in the Labour Force Survey, one is not registered (see Figure 5).
7
Figures for The Netherlands are not included because they are unreliable
5
Figure 5. Proportion of unemployed young people who are registered at the public employment service (inactives not included) 100% 80% 60%
not registered
40%
registered
20%
EL EU 15
L D K
L
P
N
S FI N
L
I
K U
IR
E
A
F
D
B
0%
Job-seekers who are not registered cannot of course be reached by the bodies implementing labour market programmes. However, even among those who are registered, attention is focused in the first instance on those entitled to benefit because the cost of their unemployment is higher for the government. This tendency is being reinforced by the new trend towards the activation of unemployment benefits. If no benefit is available, it cannot of course be activated. Furthermore, within the group of those entitled to benefit, discrimination often takes place between those covered by “unemployment insurance” on the one hand and “unemployment assistance” or the minimum income on the other hand. For example, in Belgium in 1997 these categories were legally brought into line with one another in terms of access to ALMPs. However, subtle administrative rules still exist under which those receiving public assistance are removed from the registers of employment agencies after a certain spell of time. Other legal or administrative reasons restricting access to ALMPs include: the inadequate definition of access conditions (e.g. a certain uninterrupted period of unemployment is required which means that those occasionally unemployed are excluded), the complexity of the legislation and the lack of publicity (towards both target group and employers). A second cause of non-participation is the mismatch between the programmes on offer and the needs and aspirations of the disadvantaged young people (Nicaise et al. 1995). Some (training) programmes impose excessively high requirements in terns of prior knowledge and skills, others follow too strict a pace; yet others are too schoolish and do not take into account young people’s aversion to school as a result of a career of failures. Those taking part in work experience programmes usually complain about their poor status or lack of prospects after the programme has ended. Most provisions suffer from a glaring lack of information and social guidance: they confine themselves to strictly work-related problems and lose sight of the link with other aspects of disadvantage. We believe that financial disincentives do influence the participation of young people in labour market programmes, but not specifically where one would usually expect to find them in the literature. One view sometimes put forward is that the replacement ratios of social security benefits are too generous and have a demotivating effect. In
6
other words, young people would turn up their noses at the low (minimum) wages in the segments to which they can gain access, compared to their unemployment benefits. We believe that this is too static an analysis. Generally speaking, the low wages paid to young people are still more attractive than the even lower or nonexistent social security payments meted out to young people. The replacement ratios of the social security payments seem to work out very low, particularly for schoolleavers (CEC, 1995, pp. 100-101). Moreover, we must not forget that participation in labour market programmes is itself a condition for receiving benefits in some countries (Nordic countries and the UK). However, more complex financial thresholds may be at work, that have been subject to little empirical research thus far. The considerations below are indeed based more on theoretical considerations and casual observation by field workers than on systematic econometric research. In the first instance, interaction may possibly take place with the benefits received by other family members in means-tested systems of social protection. Indeed, in some cases, young people can be restrained from accepting a job offer if, as a result, the benefits paid to their parents or other family members would be cut. The financial balance must also be examined dynamically, specifically with respect to low-income groups: sometimes acute short-term financial need takes precedence over the uncertain long-term benefit to be gained from investing in ALMPs. Even when the direct costs of participation (travel, child care etc.) are refunded, they often have to be pre-financed by the participants. If the refunding procedure takes too long, this in itself may be sufficient reason for those most in need to drop out. However, in addition to direct costs, indirect costs must also be considered, i.e. the fact that no other work can be accepted for the duration of the integration programme – albeit low-paid or temporary work. For anyone weighed down by long-term deprivation or debts, the short-term prospect of occasional work is often more attractive. In contrast to the static analysis of financial (dis)incentives, this approach advocates an increase rather than a reduction in the replacement ratios for underprivileged job-seekers. The increase can best be allocated in the form of a bonus for participation in active measures. Finally, we must mention the age-old tendency to cream off the target group in most labour market programmes. The cause may be found primarily in the desire of employment agencies to supply employers with the best possible candidates (Nicaise et al., 1995). But this brings us to another apparent paradox, one we shall examine more closely in the next section. Employment agencies appear to be convinced that labour market programmes are more effective for the stronger job-seekers (higher skilled, men, natives, short-term unemployed, ...), because ultimately the latter have better (gross) outflow opportunities. This is clearly a mistake: stronger groups may flow out faster, but the added value of labour market measures on their outflow rates is smaller. 4. THE EFFECTIVENESS OF ACTIVE LABOUR MARKET PROGRAMMES It is not our ambition to deliver a complete review of effect studies of ALMPs. Based on existing reviews (Fay et al., 1996; Friedlander et al., 1997; Meager & Evans, 1998; Nicaise et al., 1995; Nicaise and Bollens, 1998; OECD, 1994), we draw lessons
7
regarding disadvantaged target groups, young people, and the ‘intersection’ of these two.
4.1 The big misconception: ALMPs are more effective for disadvantaged groups Let us first consider the effects for the job-seekers in question at micro level. These effects are usually measured in terms of outflow rates from unemployment, or of wage rates or incomes. "Gross outflow rates" (as observed in bivariate analyses, i.e., without correction for the profile of the participants) argue against the disadvantaged groups: after participation, the long-term unemployed, the low-skilled, women and ethnic minorities still have lower outflow rates than stronger participants (the short-term unemployed, the better educated, etc.). However, the "net outflow effect", i.e., the increase of outflow rates attributable to participation is greater for the disadvantaged participant than for the average job-seeker (see, for European countries Axelsson, 1989; Bollens & Hooge 1996; de Koning, 1996; Gravesteijn-Ligthelm et al., 1995; Hofbauer & Dadzio, 1987; Raaum & Torp, 1996; Van der Burgh & Bavinck, 1995). This general tendency has also been observed outside Europe in the study by Friedlander et al. (1997). Where a difference is made according to age groups, the findings are rather mixed: young people derive now more, now less profit from ALMPs. Few studies investigate the cross-effects of age and other group characteristics: they report quite a strong net effect among disadvantaged youth (Hofbauer & Dadzio, 1987; Korpi, 1992).
4.2 Insignificant macro-effects? There is heated debate as to the effect of ALMPs at macrolevel. Attention has tended to focus on the unwanted side-effects, such as deadweight, substitution and displacement effects (see, inter alios, Fay, 1996). The increased outflow opportunities of one group are in fact partially offset by the lower employment opportunities of other groups. The net macro-effect for recruitment subsidies in particular is estimated at "only" 10 to 20% (or even less) of the gross effect. Two important comments must be made in this connection. *
First, there are various reasons to assume that the deadweight effects are smaller for programmes that selectively target the weakest groups. We are, after all, dealing with groups that would hardly stand a chance without government assistance (Bassi & Ashenfelter, 1986; OECD, 1993). As for substitution effects, the picture, in our opinion, is less clear: low and average qualified job-seekers are relatively more numerous and thus enter more readily into competition with each other (Nicaise & Bollens, 1998). Substitution can be avoided to the extent that ALMPs are focused on the bottleneck vacancies in the labour market.
*
Second, deadweight and substitution may indeed be important parameters, but they also reflect no more than a part (that is, the negative side) of the macro-effects. Other, positive side-effects are generally overlooked, e.g., the increased
8
effectiveness of the job-seeking behaviour of the unemployed persons concerned can be expected to result in wage moderation effects (which have actually been measured by the OECD, 1993). Furthermore, the returns for the government (lower expenditure on social benefits and higher tax revenues after the intervention) should create more room for manoeuvre and thus also indirect employment. In the light of the strict European convergence policy (the Maastricht criteria), this is a not inconsiderable advantage. In short, the macro-effects of ALMPs are broader and more complex than hitherto generally assumed. Such measures could be worth the trouble even with limited net employment effects at macro-level. The ultimate criterion of judgement is, in any case, not the ratio of net to gross employment effects depicted above, but the net social costbenefit balance of the conducted policy per created job (Friedlander, 1997). This argument can be illustrated by the example of a Belgian programme for disadvantaged young people and adults. It is a (rather expensive) combination of work experience with training and counselling for welfare recipients ("Training and Employment for the Disadvantaged"). The gross transition ratio to further employment after the programme was "only" 40%, the net employment effect in comparison with the control group "hardly" 15%. However, it may reasonably be supposed that this programme sooner or later becomes profitable for the community because of the increased employability and productivity of the participants and the recovery effects for the government8. Instead of throwing the programme overboard for yielding too low employment effects, we do better to increase investment given its positive cost-benefit balance (Nicaise, 1996). The above observations thus refute the prejudices held by many - including policy makers and public employment services - that ALMP has dubious effects on the hard core of the unemployed. If ALMPs have one advantage compared with other types of employment policy (wage restraint, redistribution of employment, industrial policy, etc.), then it is in this area: they can drive back structural unemployment, the so-called NAIRU - at least, if they are selectively targeted at the most disadvantaged target groups. The theoretical foundation of this last condition is illustrated in Figures 6 and 7, where the trade-off between unemployment and inflation is depicted by a Phillips curve. It is currently assumed that this trade-off is principally determined by the numbers of "effective" job-seekers. A reduction of the number of effective job-seekers implies an upward shift along the curve to the left; an increase is reflected in a shift along the curve, down to the right. Conversely, the number of hard-to-employ (or ineffective) jobseekers determines the lie of the entire curve: a fall in the numbers shifts the whole curve over to the left, an increase to the right. Point A is the point of departure for both figures.
8
The estimates are incomplete because displacement is left out of the equation but, conversely, so are the wage restraint effects. The time-horizon of the return period appears to be crucial: the balance is certainly positive if the profits on the investment last at least 2 to 3 years, which is plausible.
9
Figure 6. Effects of non-targeted labour market policies on unemployment and wage inflation
w B
C
A
U Figure 7. Effects of targeted active labour market policies on unemployment and wage inflation
w
B
A
C
U
If the active labour market policy is now brought to bear upon the average (effective) unemployed, the stock of the unemployed will decrease in the short term, but most likely at the cost of more inflation (shift from point A towards point B in figure 6). On outflow, some of the participants find work in the regular employment market but, given the limited "added value" of the measures for the employability of the persons concerned, a substantial percentage flows back into unemployment (shift from B to C). Figure 7 shows the effect of selective measures for those caught in structural unemployment. In that this group has no influence on the wage formation, unemployment falls in the short term without exerting upward pay pressure (from A to B) and nudges the Phillips curve to the left. When some of the participants end up in unemployment again after outflow, they are transformed into effective unemployed and exert a moderating influence on pay (shift from B to C along the new curve). The above reasoning further explains why the empirical literature on the aggregate impact of ALMPs is not unambiguous (Layard et al. 1991; Fay, 1996; Meager and Evans, 1998; Anxo et al., 1998; Van der Linden & Dor, 1998). Our hypothesis is that
10
the macro-effects regarding unemployment and pay pressure are more favourable as the policy is better focused on the structurally unemployed. An incorrect targeting brings wage-push inflation and a limited decrease in unemployment; whereas selective measures for disadvantaged groups give a greater outflow from unemployment plus wage restraint. Thus far, macro-studies have tended to ignore the aspect of targeting.
4.3 What sort of measures, which modalities? Can differences in effectiveness be observed between training, job search assistance, recruitment subsidies, relief work programmes, self-employment schemes or "intermediate employment (community business) initiatives"? Here again, the research findings are ambiguous: at present, it would seem rather premature to make firm statements regarding the superiority of this or that approach. For disadvantaged youth in particular, only certain tentative conclusions can be put forward: *
recruitment subsidies appear more prone to deadweight and substitution effects than direct employment schemes (relief work), and certainly more than job training (Meager and Evans, 1998). Temporary wage subsidies on the other hand allow employers to "screen" candidates from risk groups. Candidates can thus prove their employability and disabuse employers of their preconceptions. In other words, the main function of selective wage subsidies is that of redistribution of opportunities.
*
Direct employment programmes have very mixed results (Meager and Evans, 1998). Some studies detect a substantial short-term effect on the outflow from unemployment, macro-level included (see also Anxo et al., 1998). There is however a problem with the throughflow from relief jobs to regular employment. Some cases display a "lock-in" effect, i.e. the beneficiaries reduce their job-seeking because of the fact of already having work. On the other hand, good throughflow results are booked in some countries (e.g., Flanders, the Netherlands, Germany, Austria), presumably because it is clear from the outset that these work experience programmes are of a temporary nature. Here, work experience acts as a springboard, especially for the long-term unemployed.
*
Job training clearly has a greater impact on later employment chances when coupled with practical training periods with regular employers. More generally, combined measures (job search assistance, training, work experience, social guidance, etc.) appear to increase the chances of success. This, according to Meager and Evans (1998), argue the case for community businesses, which often adopt a more holistic approach than do other types of programmes. However, very little is yet known regarding the effectiveness of community businesses.
*
There is strong evidence of the inefficiency of compulsory programmes such as workfare or learnfare programmes in which young people (sometimes also the longterm unemployed or welfare recipients) must participate in order to qualify for unemployment benefits. It can hardly be an accident that all these programmes have had negative effects on the outflow chances from unemployment and on the
11
wages (Sweden: Andersson, 1993; UK: White & Mcrae, 1989, Dolton, 1993; Andrews et al., 1996; Denmark: Jensen et al., 1992, Rosholm 1994, 1995; Finland: Hämäläinen, 1998). These disappointing results are ascribed to different factors (Aakrog et al., 1991): a.
"mismatched" programmes, too homogeneous to accommodate the diversity of the needs and interests of the job-seekers. Compulsory participation relieves the providers of the programmes from any creativity in handling those needs;
b.
the demotivating effect of the obligation in itself (the training is seen more as a way of keeping the right side of the employment regulations than as a genuine springboard to employment);
c.
the stigma of the obligation, its aura of blameworthiness making the job-seekers even less attractive in the eyes of the employers.
Denmark has drawn lessons from these evaluations and, since 1994, its labour market policy has become appreciably more flexible: the "obligation to co-operate in a re-integration plan" still remains, but more alternative options are now on offer, and initiatives are more made-to-measure and more preventive. In the same spirit, the UK abandoned the erstwhile Youth Training Scheme in favour of the national apprenticeship programme and the "Entry Level Provision" having better-fitting service packages for job-seekers. *
The pedagogic aspects of the measures have received scant critical attention. Brun et al. (1991) and Godinot et al. (1995) list the following quality criteria for the success of projects for the most deprived people: a holistic approach (see above), a sufficiently long training period (two or three years is not considered a luxury), alternating formulas combining training and work, preferably starting with work experience and with a gradually increasing training component, and sufficient remuneration in order to prevent dropout.
*
Be this as it well may, we must not lose sight of the macroeconomic aspects of (youth) unemployment, nor of the problem of its regional concentration. Several studies (EC, 1998; OECD, 1996) show that the aggregate volume of economic activity is the No. 1 determinant factor in youth unemployment. Active measures designed to increase the employability of youth may help, but are drops in the ocean in the context of areas such as Southern Italy, where more than half the young population are jobless. It goes without saying that an industrial and regional policy is also required there, directed towards the demand side of the labour market and towards the creation of new enterprises (Pugliese, 1995).
5. RECENT TRENDS AND PERSPECTIVES It is impossible to give a quantitative summary of the different strategies currently in progress throughout the EU designed to combat youth unemployment. Many programmes are, in any case, not aimed at specific target groups. Low-skilled
12
unemployed youth are targeted very actively in the countries of Northern Europe, more often than not by linking social allowances to work experience and job training programmes. Such a "watertight" approach is hardly feasible in Southern Europe. Here, we find greater reliance on the apprenticeship system and on the "intermediate labour market" (social co-operatives, workshop-schools, etc.). France and Germany make use of large-scale direct employment programmes. The Structural Funds (the European Social Fund in particular) streamline policy at EU level. Here, the major accent rests on job training, though recent years have witnessed an interest in a diversification of measures. "Integrated pathways to employment" are the new fashion, with (a) a provision of individualised services; (b) partnership between diverse actors (government, the social partners, the third sector, ...); (c) a package of counselling, training, work experience, job clubs and labour cost subsidies; and, (d) a ‘trajectory’ that must lead on to permanent integration. This approach is better suited to the needs of disadvantaged groups, though still too often limited to services in the field of employment integration and hard to implement. Among the priority target groups of the ESF, a special place is accorded to (lowskilled) young persons as well as "persons threatened with exclusion from the labour market". Unfortunately, the category of "young persons" casts it net too wide for practical purposes, with the result that many countries complain of a "creaming-off" effect (European Commission, 1998b). The European Social Fund is gradually being fitted into a broader joint EU employment policy9. This policy finds concrete form in the EU Employment Guidelines which are being applied in national action plans. Guideline no. 1 merits particular comment here: "Within five years (...) each Member State will ensure that every young person is offered a new start before reaching six months of unemployment, in the form of training, retraining, work practice, a job or other employability measure ...". A similar approach applies to persons over the age of 25 before they hit the limit of one year of unemployment. These objectives go beyond the very ambitious. They also signal a new approach in social protection determined to go beyond the mere replacement of income from paid employment. We could, for want of a better name, call it "activating social protection". This is anchored in a growing awareness that not only the right to a (minimum) income but, also, the right (and duty) to work is essential for social cohesion. What is more, it is common knowledge that both traditional Bismarckian and Beveridgian social protection systems can no longer cope with the present-day kind and scale of social exclusion. With Bismarckian systems, we note that more and more people are unable to consolidate their rights to social security through a track record of employment. Beveridgian systems are littered with the poverty trap. This being the case, a new sort of social contract is taking shape in which social protection is assured in exchange for a
9
The four pillars of wisdom of this policy being: improving employability, developing entrepreneurship, encouraging adaptability in businesses and their employees, and strengthening the policies for equal opportunities.
13
personal commitment to integration, via community service, training or participation in work experience programmes. The new approach is very much inspired by the experience of the Nordic countries. All these countries have rubbed shoulders with one or another form of "learnfare" and "workfare", where young people receive benefits only if they attend a training scheme or sign up for a work experience programme10. The government then goes on to guarantee a sufficient supply of job training and work experience places, so as to strike a certain balance between the rights and obligations of individual and community. The New Deal in the UK and the Jobseekers’ Integration Act (WIW) in the Netherlands travel the same road. In many European countries, the access to public assistance is also linked to the conclusion of "integration contracts", where the receiving end and the municipal welfare agency join forces to put an end to the situation of dependence. This trend can, in principle, be called positive. Whereas in the traditional Bismarckian systems millions of young people were unable to acquire rights to social security through earlier employment (retro-active insurance), they should now have meaningful chances to prove their readiness to work and to integrate if the new, pro-active formula is true to its promise. The system also meets two or three basic needs at a stroke: income, work and/or training - where previously there was naught but income replacement. In the third place, the creaming-off effects in ALMPs are countered, meaning that the most underprivileged will also be invited to the table when the government seriously undertakes to make an offer to anyone who meets certain criteria. Finally, these European guidelines are redolent of the will to prevent social exclusion by long-term unemployment rather than to cure it. Denmark, Sweden and Finland are, at first sight, "runaway" success stories in this connection. In one twelvemonth, Finland saw its youth unemployment nosedive by more than a third while the percentage of school attendance took the upward gradient (Aho & Vehviläinen, 1997, cited in MISEP 1998); the number of low-skilled unemployed young persons in Denmark fell by a full 70% just one year after the introduction of the new compulsory training scheme (MISEP 1997). There are, however, certain fundamental dangers. 1. To begin with, there is some fuzziness about the description of the target groups and their mutual relations. According the young such high priority - surely this is to eat off the plate of other, possibly even more disadvantaged target groups? By no means all young persons necessarily have integration problems. A sudden new tack on behalf of short-term unemployed youth may de facto, paradoxically, give way to even more stubborn exclusion of (younger and older) long-term unemployed ("reverse substitution effect"). 2. Will there be a priority for those who already draw benefits ? The latest lever in the ALMPs is the activation of benefits and allowances. We have seen that only a small minority of young people draw any benefit at present. A watertight system of
10
The difference with hard-line workfare is that remuneration usually meets normal standards.
14
social protection, as described above, should provide cover for all school-leavers who appear unable to find employment past a certain cut-off point. Bearing in mind that many young people currently live at the margins of the labour market, there is a risk that many a national action plan, unless re-thought, will miss those most in need of assistance. 3. Targeting also has implications regarding the effectiveness of the measures put up. Measures aimed at the earliest possible intervention, irrespective of the job seekers’ level of educational attainment, can by all means be expected to suffer considerable deadweight and substitution effects. In our view, the urgent call for prevention has not altogether been founded in reason. A remedial approach can sometimes be more efficient than a "full frontal" preventive one. de Koning and van Nes (1998) have demonstrated that the a priori "expected duration of unemployment" (forecast based on econometric analyses of individual characteristics upon entry into unemployment) involves a high uncertainty margin. Allocation of services on this basis may to a large extent serve groups who do not need help.11 Or, to put it another way, European Guideline no. 1 might well turn out to be an expensive affair. 4. In some countries, it has become clear that government was not up to the job of matching supply quantity with quality within the set time-frame. In Denmark and the UK, the "iffy" quality of the early 90s resulted in disappointing outflow rates (see section 4.3.). The Finnish authorities have recently been forced to loosen up as regards obligations on the part of youth: instead of actual participation in a training scheme, they now accept three ‘applications’ for job training sufficient to qualify for unemployment benefit. 5. A further danger with "activating social security" is that of growing authoritarianism. There is already evidence of a systematic stiffening of unemployment regulations in Europe. The obligation to accept offers has always existed, but this was shaded by a set of criteria as to the nature of a "suitable offer": it must satisfy minimum or maximum requirements regarding mobility, working hours, match between the offer and the needs or capacities of the individual, pay, etc. These standards become increasingly unfavourable (EC, 1997, pp. 102-105). So, we end up in the paradox that the public authorities themselves are developing a "secondary labour market", in which compulsion, uncertain statutes and, sometimes, near exploitation rule. There is the added danger of increased exclusion from social protection: jobseekers who refuse the offer or drop out lose their benefits in greater numbers. It is puzzling, for instance, that the proportion of unemployed young people who are not registered at the public employment service is so high, not only in countries with a weak social protection system, but also in the Nordic countries where social
11
de Koning and van Nes advocate continual re-assessment of the outflow rates. Given that non-observed handicaps of individuals can only be ‘revealed’ by the duration of unemployment, appropriate action should be taken as and when the outflow chance dips below a certain level.
15
protection is well developed (see figure 5). Does this signal non-compliance of youngsters with the coercive rules of the system, or sanctions, or is it a matter of definition ? Could there be a trade-off between social protection and ALMPs ? Different studies appear to point to a limited positive impact of sanctions on the outflow rates, but it should also be mentioned that the outflow often leads to inactivity instead of work (Abbring et al., 1995; Aho & Vehviläinen, 1997; van den Berg et al., 1998; Van der Linden & Dor, 1998).12 There are practically no studies regarding the social side-effects of exclusion from unemployment (Wets 1998, ATD Quart Monde 1998). In the UK, the alarming increase in poverty and homelessness among young persons was brought into connection with the shortage of places in the Youth Training Scheme and the resultant exclusion of those young persons from social protection (Nicaise et al., 1995, p. 59). In short, the new model of social protection has opportunities and pitfalls. What is important is that the final objective (combating social exclusion) continues to have the upper hand over the pressure to score "good figures" as regards pushing back registered unemployment. These two objectives may not be mutually exclusive, but that does not make them synonymous. If we hope to prevent "activating social security" from degenerating into a bureaucratic mill or an exclusion machine, high standards will have to be imposed regarding quality, customised programmes and freedom of choice for job-seekers.
6. CONCLUSION Disadvantaged youth in Europe will probably be the touchstone of the new social model. Given the combination of their vulnerable position in the labour market and their marginal position with regard to social protection, the national governments are obliged to search for new ways of integration. Guideline no. 1 of the joint EU employment policy is, perhaps, the catalyst in this process of exploration. We are evolving - ideally towards the linking of income transfers with packages of ALMPs. In contrast to "common knowledge", research to date shows that ALMPs, both at individual and at macro-level, are more effective the more they are selectively targeted on the most disadvantaged groups. It is then all a question of: a) reaching out effectively to these groups (many disadvantaged young persons fall outside any system of social protection and have no more connection with the competent authorities); b) putting together a suitable offer, i.e., a holistic approach in the form of "pathways to integration", sufficiently long and intensive, with a genuine prospect of paid employment at the outcome, a good balance between rights and obligations,
12
It may further be assumed that outflow to work resulting from sanctions is also accompanied by deadweight and substitution. The aforementioned microstudies do not take this into account. Van der Linden & Dor (1998) report a macro effect near zero.
16
positive support rather than a coercive approach, a solid status and sufficient freedom of choice. The will in the Member States of Europe to arrive at a watertight offer within a period of five years must be viewed as ambitious. The most disadvantaged groups do not in fact really have a chance to benefit from policy measures until these take the form of guaranteed services. The undertaking, however, is not without risks, e.g., wrong definition of the target group, a pressure to ‘statistical performance’, and authoritarianism.
BIBLIOGRAPHICAL REFERENCES Aakrog et al., V. (1991) UTB, Uddannelsestilbud for langtidsledige, København: Danmarks Lærerhøjskole & Udviklingscenter for folkeoplysning og voksenundervisning Abbring, J., van den Berg, G.J., van Ours, J. (1995), The effect of unemployment insurance sanctions on the transition rate from unemployment to employment, Working Paper, Dep. of Economics, Univ. of Amsterdam Aho, S., Vehviläinen J. (1997), Activating the young unemployed into education ? Studies on the effect of a recent policy in finland and on the hidden rationalities among uneducated young people, Univ. of Tampere, Research Inst. for Social Sciences, n° 53 Andersson H. (1993), Choosing among alternative non experimental methods for estimating the impact of training : New Swedish Evidence, mimeo, Swedish Institute for Social research, Stockholm University. Anxo, D., Carcillo, S., Erhel, C. (1998), Aggregate impact analysis of ALMPs in France and Sweden: a regional approach, paper presented at the 10th Annual EALE Conference, Blankenberge, Sept. 1998 Axelsson R. (1989), Svensk Arbetsmarknadsutbildning - en kvantitativ analys av dess effekter, in Umea Economic Studies, Umea University. Andrews, M.J., Bradley, S., Upward, R., Estimating the costs and benefits of Youth Training in th the UK, paper presented at the 8 Annual EALE Conference, Chania, September 1996 ATD Quart Monde, Sortir de l’inactivité forcée, Dossiers et Documents de la Revue Quart Monde, n° 8, 1998 Bassi L.J., Ashenfelter O. (1986), The effect of direct job creation and training programs on low-skilled workers, in : Danziger, S., Weinberg, D. (eds), Fighting poverty : what works and what doesn’t, Harvard Univ. Press, 1986, p.133-151 Bollens J., Hooge J. (1996), Bereik, kwaliteit en effectiviteit van de VDAB-beroepsopleiding voor werkzoekenden, Luik 1, deelrapport 1, HIVA, Leuven. Brun, P., Guillot, F., Viard, T., Le Crédit Formation Individualisé au regard des jeunes issus des milieux très défavorisés, Pierrelaye, IRFRH, ATD-Quart Monde, 1991 Commission of the European Communities (1995), Social protection in Europe, Dir.-Gen. Employment, Industrial Relations and Social Affairs, Luxembourg, Office for Official Publications of the European Communities Commission of the European Communities (1997), Social protection in Europe, Dir.-Gen. Employment, Industrial Relations and Social Affairs, Luxembourg, Office for Official Publications of the European Communities
17
de Koning, J. (1996), Strategies and measures to fight long-term unemployment : the Dutch case, NEI-paper, Nederlands Economisch Instituut, Rotterdam de Koning, J., van Nes, P. (1998), Prevention in active labour market policy: is it possible and is it desirable ?, paper presented at the 10th Annual EALE Conference, Blankenberge, Sept. 1998 Dolton, P.J. (1993), The economics of youth training in Britain, Economic Journal, 103(420), September 1993, p.1261-78 European Commission, Dir.-Gen. Employment, Industrial Relations and Social Affairs (1998), Comparative analysis of national performance on integrating young people into the labour market, Brussels, mimeo. European Commission, Dir.-Gen. Employment, Industrial Relations and Social Affairs (1998b),Conclusions of the ESF mid-term evaluations, Brussels Eurostat (1996), Labour Force Survey, Luxembourg, Office for Official Publications of the European Communities Fay R. (1996), Enhancing the effectiveness of active labour market policies : evidence from programme evaluations in OECD Countries, Labour Market and social Policy occasional Papers, No. 18, OECD. Friedlander, D., Greenberg, D.H., Robins, Ph.K. (1997), Evaluating government training programs for the economically disadvantaged, Journal of Economic Literature, XXXV (Dec. 1997), p.1809-1855 Godinot X. et al. (1995), On voudrait connaître le secret du travail, Paris, Ed. L’atelier / Quart Monde, 350p. Gravesteijn-Ligthelm J., De Koning J., Olieman R. en Van Der Weijde I., Ex post evaluatie van de ESF hulpprogramma’s voor doelstellingen 3 en 4 in Nederland, NEI, Rotterdam, 1995. Hämäläinen, K., The impact of active programmes on employment in the eras of high and low th unemployment, Paper presented at the 10 annual EALE-Conference, Blankenberge, September 1998 Hartog J. (1997), On Returns to Education : Wandering along the Hills of ORU Land, Keynote speech for the Conference of the applied Econometrics Association, Maastricht, May 1997. Hofbauer H., Dadzio W. (1987), Mittelfristige Wirkungen beruflicher Weiterbildung. Die berufliche Situation von Teilnehmern zwei Jahre nach Beendigung der Massnahme, in Mitteilungen AB, 2/87, p. 129-141. Jensen, P., Pedersen, P.J., Smith, N., Westergård-Nielsen, N. (1992), Measuring the effects of labour market training programmes, Fourth Annual Conference of the European Association of Labour Economists, Warwick, 14p. Korpi T. (1992), Employment stability following unemployment and manpower programs, Stockholm Research Reports in Demography, n° 72, Stockholm University LAYARD, R., NICKELL, S., JACKMAN, R., Unemployment: macro-economic performance and the labour market, Oxford University Press, 1991, 618p. Meager, N., Evans, C. (1998), The evaluation of active labour market measures for the longterm unemployed, ILO Employment and Training Papers n° 16, Geneva MISEP, Policies (quarterly journal of the EC’s Employment Observatory), several issues Nicaise I. (ed.), Bollens J., Dawes L., Laghaei S., Thaulow I., Verdié M., Wagner A. (1995), Labour market Programmes for the Poor in Europe : Pitfalls, Dilemma’s and how to avoid them, Avebury, 146 pp.
18
Nicaise I., Bollens, J. (1998), Training and employment opportunities for disadvantaged groups, in : M. Tessaring (ed.), Vocational education and training – the European research field. Background report, Luxembourg, Office for official publications of the European Communities, vol. II, p.121-153 Nicaise I. (1996), Donner du poisson ou apprendre à pêcher ? Analyse des coûts et bénéfices sociaux des projets d’insertion EFD des CPAS, Revue Belge de Sécurité Sociale, 38(4), déc. p.909-923 OECD (1993), Employment outlook, Paris OECD (1994), The Jobs Study, Paris OECD (1996), Employment Outlook, Paris OECD (1998), Employment Outlook, Paris Pugliese, E. (1995), Special measures to improve youth unemployment in Italy, in McFate, K., Lawson, R., Wilson, W.J. (1995), Poverty, inequality and the future of social policy, New York, Russel Sage Foundation, p.439-460 Raaum, O., Torp H. (1996), Impact of training on earnings. Testing alternative models, Paper presented at the EALE Conference, Crete, sept. 1996. Rosholm, M. (1994), Effektmaling af ATB m.v., Centre for Labour Market and Social Research, Univ. of Aarhus Rosholm M, (1995), The effect of public job creation on the exit rate from employment and unemployment for long-term unemployed, paper presented at the 7th Annual EALE Conference, Lyon, Sept. 1995 Speckesser, S., Hilbert, C. (1998), Active labour market policy and unemployment outflows: an aggregate analysis for Germany, paper presented at the 10th Annual EALE Conference, Blankenberge, Sept. 1998 van den Berg, G.J., van der Klaauw, B., van Ours, J.C. (1998), Sancties in de bijstand vergroten de kans op werk, Economisch-Statistische Berichten, 83(4161), 17 July 1998 Van der Burgh Y, Bavinck S. (1995), Scholing en werkervaring voor de onderkant van de arbeidsmarkt, in: Huygen et al., Naar volwaardige werkgelegenheid?, SISWO Van der Linden, B., Dor, E. (1998), The effect of unemployment benefits, sanctions and training on unemployment outflows, paper presented at the 10th Annual EALE Conference, Blankenberge, Sept. 1998 White, M., Mcrae. S. (1989), Young Adults and Long-Term Unemployment, Policy Studies Institute London, 289p Wets, J. (1998), Communicerende vaten ? Een onderzoek naar de relatie tussen werkloosheid en sociale bijstand, Leuven, HIVA
19
