Eppur si muove?: Activation policies in Austria and Germany

Current debates about activation policies focus only to a very limited degree on continental welfare regimes. The literature frequently deals with the Scandinavian countries, especially Denmark and Sweden, that have established labour market programmes maximising the employment opportunities of unemployed persons through individual action plans, training and jobs. Likewise, attention is often drawn to ‘liberal welfare regime’ countries, most notably the US, where activation has a quite different, and even opposite, meaning linked to the notion of the ‘workfare state’ (Peck 2001), i.e. a state that forces people to search for and accept jobs, of whatever kind, through reduction of benefits and similar ‘incentives’. In sum, activation policies seem to involve either the carrot (Scandinavia) or the stick (the US, and for some, such as Peck, the UK) (see Barbier 1998: 293), and little attention is paid to those countries that do not seem to fit in either category.

If continental welfare states are dealt with, they usually are rated quite unfavourably in terms of activation. Such judgments often are derived from single observations, not least from Germany's pervasive failure to reduce its high unemployment rate (Clasen 2000). This failure is said to stem not only from an increasingly globalising economy, from high wages and (especially in Germany) from low wage dispersion and the concomitant service gap (Fuchs and Schettkat 2000; Manow and Seils 2000: 296–301), but rather, and perhaps foremost, from the inadequacies of the continental social model with its ‘inbuilt tendency to eat the very hand that feeds it’ (Esping-Andersen 1996: 79–80) in a cycle of lack of employment → policies of labour supply reduction → rising pension and other costs → rising (non-wage) labour costs → lack of employment and so on (ibid.; see also Manow and Seils 2000; Ferrera and Hemerijck 2003: 108). Likewise, the tax wedge caused by these policies allegedly leaves little money in the pockets of those in employment, which reduces the gap between wages and unemployment benefits even more and thus enhances the ‘disincentive’ effect of unemployment benefits (Siebert 1997). To make a long story short, activation policies seem both urgently needed and impossible to achieve.

We do not maintain that these views about the difficulties of activation policies are completely wrong; rather, we would argue that the picture is more complex. This paper attempts to show this by analysing two countries with very similar features in terms both of labour market institutions and the organisation of social security, namely, Austria and Germany, which can be seen almost as prototypes of the ‘continental’ system of security. As the European Employment Strategy (EES) and the means to enhance it, such as the Open Method of Co-ordination, trigger more commitment on part of national governments towards broader activation policies, we may inquire whether and to what extent both countries develop activation policies in a more comprehensive sense, ‘as an increased and explicit dynamic linkage introduced in public policy between social, welfare, employment and labour market programmes’ (Barbier 2001: 2). Indeed, such ideas are more and more prominent in political discourse in both countries; yet, it remains to be seen to what extent they are translated into practice. The paper starts with an outline of the basic similarities and differences of Austria and Germany in terms of welfare state institutions, employment/unemployment and social expenditures. We will then go on to analyse, separately for each country, the development of active labour market policies during the past two decades, not least with respect to notions of social citizenship. The final section summarises our discussion.

Both Austria's and Germany's welfare states are commonly described as belonging to the ‘conservative’ (Esping-Andersen 1990), ‘continental’ (Bonoli 1997) or ‘Christian democratic’ (Huber and Stephens 2001a,b) welfare regime type. Social security is based on a state-regulated, contribution-based insurance system and entitlements to benefits such as unemployment compensation or old-age pensions are largely derived from paid work (and not from need or from citizenship; see Esping-Andersen 1990; Clasen and van Oorschot 2002). This also implies that normally only those who are granted unemployment benefits can take part in active labour market programs, e.g. training measures. A further consequence is the rather high degree of ‘stratification’ (Esping-Andersen 1990) or status protection provided. This is reflected in the equivalence principle (Äquivalenzprinzip) with the consequence that benefit levels are proportional to contributions. Another facet is the principle of occupational protection (Berufsschutz), based on the institutionalised system of vocational training. This principle is part of the definition of a ‘suitable job’ (the jobs to be offered to job-seekers by the Public Employment Service [PES]), thus securing the social status a person has achieved (on the changes to these principles, see below).

In line with the ‘conservative’ regime nature of both countries the state, while providing some services (like schools and universities), does not consider it its task to create a large public sector in order to promote high employment rates (even though in Austria after WW II the degree of state-owned enterprises was considerable). On the contrary, in accordance with what Germans like to call the ‘social partners’ policies of labour supply reduction like early retirement were a common reaction to the employment crises since 1975 (on Austria, see Hemerijck et al. 2000: 199; 205; on Germany: Manow and Seils 2000). But otherwise state involvement in many features of life is considerable. Of special importance is the fact that in both countries, both large parties, the social democrats and the conservative parties (CDU/CSU in Germany and ÖVP in Austria), traditionally subscribe to a strong welfare state. In the light of this, it has been considered remarkable that in Germany during the 1980s and 1990s steps towards cost-containment were taken (Alber 2000: 235; Pierson 2001: 450–1); Austria was much less successful (if we consider cost-containment a success!) in this regard (Alber 2000: 253).

One reason for this could be the fact the Austria was hit less hard by the two oil price shocks and the ensuing turbulences. Until the end of the 1970s, ‘Austro-Keynesianism’ with a strong emphasis on full employment was the dominant strategy. After the second crisis around 1980 full employment was replaced by budget consolidation as the central goal and was managed quite successfully (Hemerijck et al. 2000: 200). Despite this focus on keeping the public budget under control, public sector employment, while not reaching Scandinavian levels, increased considerably during the 1980s and 1990s, thus mitigating the loss of jobs that resulted from privatisation of formerly state-owned enterprises.

The current situation on the Austrian and German labour market looks quite similar concerning employment (see Table 1).² The employment rate (for men and women) is above EU-average, but still considerably below the Scandinavian countries. During the 1990s the employment rate of men declined, whereas the participation of women in gainful employment increased. Yet, Austria has very low unemployment, with long-term unemployment as a comparatively minor problem (1 per cent), although women are more affected than men. In Germany, however, the unemployment rate reached 7.9 per cent in 2001 which is slightly above the EU average, and long-term unemployment (2001: 3.9 per cent) is four times higher than in Austria. That Austria from the beginning was quite successful in keeping its unemployment rate at low levels, in spite of a system of social security that is very similar to that of its neighbour, casts doubts on the notion that it is the very fabric of the social security system that is at the core of the unemployment problem.

This situation is also reflected in social expenditures. While both countries spend similar percentages of the GDP on social security (Austria: 28.7; Germany: 29.5 in 2001, Mayrhuber 2003: 43), expenses for (un-)employment account for almost 10 per cent of social expenditures in Germany and roughly half of this in Austria (Mayrhuber 2003: 35 for Austria; BMSG 2001: Table I-3. for Germany).³ Interestingly enough, expenses for old age pensions (including those for surviving dependents) are proportionally higher in Austria than Germany even though the old age dependency ratio is markedly higher in the latter country (Abramovici 2003: 7).

How is this situation reflected in current debates about social policy? How do the demands for more ‘activation’ of the welfare system make themselves felt? In both countries, there is now indeed a rather broad consensus that non-wage labour costs are to be blamed for much of the current malaise. For the Schröder government in Germany cutting down these costs was one of the most prominent targets from the outset and attempts were made to shift the burden more on the shoulders of tax-payers or of employees. In the old-age pension system, two new measures have been introduced that are supposed to keep old-age pension contributions at acceptable levels: a tax on energy (with the additional advantage to be in line with the Green party's emphasis on penalizing energy consumption), and state subsidies to private old-age pensions [‘Riester-Rente’]). Past policies of labour supply reduction, which were one of the decisive means to keep unemployment levels low in East-Germany, have been cut down and early retirement is now penalized by reductions in pension payments. Reform of the health system has been on the agenda for more than a decade, and the most recent measures have tried to cut down contributions by putting higher burdens on those who actually use the health system (such as ‘practice entrance fees’ (Praxisgebühr) or increases in patient co-payment of medicines).⁴

In Austria there have also been several reforms to tackle the financial problems of the social security system, often quite similar to those in Germany. In the health sector, ‘practice entrance fees’ similar to Germany (Krankenscheingebühr) have been introduced and an increase in employees’ contribution rates is still under discussion. On the other hand structural changes in the health system are planned, as for instance a merger of health insurance institutions or a reduction in the number of beds in hospitals. The goal of the pension reforms is to increase the effective retirement age (to the statutory age of 60 for women and 65 for men) by abolition of early retirement schemes. However, the changes in the pension system led to increasing unemployment of the elderly.

Similarities also can be observed concerning family policies, and in line with the reliance on the family that is characteristic for conservative welfare states the activation tendencies are less pronounced in this field. It is true that in both countries during the past decade child care facilities were expanded; but the development was hesitant in Germany – there are still very few facilities for children under 3 and schools are just beginning to extend child care in the afternoon – and in Austria it was even brought to a halt during the past few years by the conservative government. In addition, both countries pay rather generous child allowances, especially during the first few years of children's lives (such as the German Erziehungsgeld, a means-tested benefit available only for persons working part-time or not at all, and the Kinderbetreuungsgeld introduced in Austria in 2002 which is designed as a family entitlement). This works as an incentive against female labour-market participation and in Austria even has reversed the earlier trend towards rising female employment rates.⁵

Thus, both countries indeed exhibit a very mixed picture concerning activation in several fields of social policy. Politicians subscribe to cut down pension and health costs, but they fail to achieve major cutbacks, not to mention a decisive shift of the system. In family policies, both countries give child allowances priority over child-care provision and thus work against encompassing female labour market participation.

Labour market policies have emerged as a field where attempts at thorough reform make themselves felt much more strongly. These policies are an interesting target for analysis of activation policies also for a second reason, often ignored in comparative analyses, namely, a strong heritage of active labour market policies. First, unemployment benefits, the alleged hallmark of continental ‘passive’ (or even ‘passivating’) labour market policies, have strong ties to activation in Austria and Germany (see also Kvist 2001: 200): A record of employment is a pre-requisite for obtaining benefits, and willingness and ability to take up jobs – and more recently to be able to demonstrate being actively engaged in a job search – are required during the period of benefit receipt.⁶ Second, and more importantly, activation has been by no means alien to both countries’ labour market policies from the outset. In Austria, during the times of full employment active labour market policy (ALMP), with the key objective to match supply and demand, was considered the basic measure to prevent a labour shortage – such that some observers have classified Austria together with the nordic, social-democratic welfare regimes in this respect (Huber and Stephens 2001a: 114–5). Similarly in Germany, ALMP – with emphasis on training the workforce to be upwardly mobile – was seen as a hallmark of the Arbeitsförderungsgesetz [AFG; ‘Labour Promotion Act’)] introduced in 1969. This may be explained to some extent by the fact that both countries, in spite of their belonging to the ‘Christian democratic’ cluster of welfare states (Huber and Stephens 2001a: 108; 2001b: 87), have exhibited considerable social democratic influence during these times (with Austria as ‘the only Christian democratic welfare state in which social democracy is more influential than Christian democracy’, Huber and Stephens 2001a: 111).

Current statistics about activation seem to indicate that both countries do not fare so badly after all in this respect. For instance, with roughly 2,000,000 unemployed persons that participated in training measures in Germany during the year 2001 and €14 billion spent it can scarcely be said that the country has remained ‘inactive’. The same holds for Austria, which even though starting from a smaller activation rate has taken considerable steps towards increasing the number of people affected by training and other measures. In 2001, for instance, ‘active’ expenses in Austria amounted to 0.53 per cent of the GDP. While this is clearly below the OECD average of 0.73 per cent, the percentage spent looks much better when unemployment rates are held constant,⁷ and also in comparison to the past as spending on ALMP increased by 138 per cent during the 1990s. The allocation of these means made it possible to bring into effect about 330,000 participations in the year 2000 (80 per cent of these in training), which is an overall increase of 68 per cent compared to 1990 (196,800 cases).

As mentioned above, ALMP has a long tradition in Austria. ALMP includes all measures intended to successfully match the supply of jobs with the existing demand, i.e. counselling, placement, and training activities of the PES (Public Employment Service; in Austria: Arbeitsmarktservice) as well as subsidised employment.

Until the 1980s the basic aim of ALMP was to upgrade the labour force, keeping its supply in line with demand through promotion of occupational mobility. However, by the end of the era of full employment various problem groups (e.g. a sizable number of long-term unemployed) had started to develop. In 1983 an amendment to the law on labour market promotion gave rise to a new ‘experimental labour market policy’ intended to develop new and innovative tools in the fight against unemployment. During this period, experimental labour market policy indeed seemed to be some sort of ‘playground’, an area to which resources were allocated to try out new measures. One of the better-known approaches in this context was the ‘Aktion 8000’, aimed to create 8000 new jobs that would also be ‘beneficial to society’. Several other new types of measures were developed, as for instance ‘Stiftungen’,⁸ employment projects, or counselling services for specific problem groups. Besides, measures were targeted for the needs of specific groups such as orientation and integration measures for young people, female returners, and long-term unemployed (e.g. Pinter et al. 1988). ‘Socio-economic enterprises’ and non-profit employment projects⁹ were set up in order to encourage the re-integration of problem groups (especially long-term unemployed people with psychosocial problems) via a ‘third labour market’. In the beginning most of these projects were small and had a rather innovative character. But during the 1990s those which proved to be successful were expanded and became institutionalised.

The 1990s were characterised by a substantial increase of the total number of people affected by unemployment, as well as by a more pronounced focus on training measures. However, ALMP covers a large variety of measures with regard to duration, intensity, and target groups. Considering the number of participants, training, orientation and job search assistance are the most important measures and represent the main part of the extention of ALMP. ‘Stiftungen’ and subsidised employment play a minor role but are more intensive (lasting from 6 months to 3 years) and therefore more expensive.¹⁰

The development of active labour market policy in the 1990s was mainly shaped by two factors: the PES gaining independence in 1994, and Austria joining the EU in 1995. These two events and the effects thereof have changed both the meaning and the appearance of active labour market policy.

The PES is the key force in Austrian labour market policy. In 1994, labour market administration became independent and the new PES was established as a public service company with its own legal title. The main focus was on essential labour market goals, the decentralisation of decision-making, and improving cooperation with the social partners at a national, provincial and regional level. The PES at this point also adopted a new management system – management by objectives.¹¹ This system, first implemented throughout the country in 1995, provides quantified targets and indicators to regulate the organisation's performance on the labour market. The controlling and monitoring system was extended as well, now placing more emphasis on the efficiency and effectiveness of active labour market policy.

After Austria had joined the European Union and especially after the EES had first been implemented in this country the EU became an important force in the shaping of labour market policy. Thus the European objectives now have to be implemented in addition to the PES objectives, whereby the former need to be integrated into existing regulations and institutional set-ups. The PES, for instance, adopted the NAP objective to implement gender mainstreaming and to get 20 per cent of all unemployed people to participate in training programs. But – in terms of quantity – active labour market policy also became more important, both with regard to the number of participants and the size of the budget allotted to such measures. In 2002 about 35,400 people have been on training (average per year) which is three times higher then in 1990 (12,250 people). Also the budget for ALMP increased from €258 million (1990) to €666 million in 2002.

In total, it is mainly women who have profited from the expansion of active labour market policy, especially since the mid 1990s. At the beginning of the 1990s the training rate of women had still been below that of men, but as from 1993 the female share of people in training was higher than the share of women who were out of a job and in 1996, when the PES initiative for female returners was started, more women than men participated in training measures. The goal to encourage equal opportunities for women and men through ALMP was established in the National Action Plan in 1998 and in turn led to an increase of ‘women-specific’ measures as well as to the development of new opportunities especially suited to the needs of individual target groups.¹²

Even though training measures were extended, some groups of unemployed people still do not benefit from this development as much as others: on the one hand, those with no entitlement to unemployment benefits or emergency assistance encounter more difficulties when it comes to participating in different measures. This primarily applies to women with insufficient skills who have been out of work for an extended period of time. They are clearly underrepresented as compared to women with better skills and shorter unemployment spells. But the same disadvantages also apply to unskilled and older unemployed persons and migrants, equally underrepresented in training measures.

The years 2000 and 2001 – i.e. immediately after the change in government – were characterised by the attempt to balance the national budget. Thus, the budget of the PES was cut at a time when unemployment increased unexpectedly, and the PES was forced to reduce and/or discontinue a number of measures. More expensive, albeit innovative and intensive, measures receded into the background, in favour of cheaper measures like the ‘job coaching’ program, a far-reaching but not very intensive measure.¹³

When unemployment rates skyrocketed during the second half of 2001, ALMP increasingly failed to ease the precarious labour market situation. The opposition parties and the employees’ representatives therefore demanded an increase of training measures, whereas the government – especially the Federal Minister for Economic Affairs and Labour – opted for tightening the requirements concerning ‘suitable work’ for persons unemployed for more than 3 months. Thus far the principle of occupational protection largely had been valid also for the unemployed, with mobility restrictions and child-care duties as additional legitimate reasons to decline a specific job. Plans were made which implied that from November 2001, unemployed people would be obliged to take any other job during the second half of the maximum period of unemployment benefit on penalty of discontinuation of payments for 6 weeks. A job would now still be considered suitable as long as it takes no more than 1.5 hours to get to work. Supporters claimed that these measures were necessary to make the labour market and the unemployed more flexible, to combat the misuse of unemployment benefits and emergency assistance, and to ‘activate’ the unemployed.

This situation gave rise to an emotionally charged public debate, lasting several months, on the rights and obligations of unemployed people (and especially on the reasonability of certain jobs or sanctions), which ended in a popular referendum initiative ‘Welfare State Austria’. One of the co-initiators, Prof. Talos, evaluated the reform plans as follows: ‘it would be quite contemptuous of humanity to tighten the suitability requirements and reduce the means for active labour market policy at the same time’.¹⁴

As a compromise, the social partners agreed on a support plan for each unemployed person, which should outline not only obligations, but – for the first time – also the rights of the unemployed towards activation. This was only partly realised. The newly established individual action plans take account of a jobless person's prior profession, his/her future career plans, and the resulting need for assistance or qualification. But they do not cover the right to participate in active labour market policy.

In sum, the meaning of the term activation in connection with labour market policy has changed during the last few years, and discussions on this topic were sometimes extremely controversial. The dispute, however, focused on subsidised employment – especially in the non-profit sector – rather than on PES training measures. One example for that would be the initiative ‘Integra 2000’, which aimed to re-integrate 1,000 long-term unemployed people into work in the year 2000. This was to be achieved by placing jobless people with non-profit organisations, a strategy often criticised for providing only low-paid ‘community-service work’ (a 20 per cent supplement to the regular unemployment benefit). Due to the huge media response and the accompanying ideological debates this measure was progressing rather slowly, i.e. only 500 people participated in it. Although with an estimated budget of about 100 million ATS (€7.3 million) it played a comparably insignificant role in labour market policy, it was extensively debated for several months and was the subject of a parliamentary inquiry.

In this context, discussions on the coercive nature of active labour market policy intensified, i.e. the discussion about sanctions for people who refuse to take a job or participate in a given measure.¹⁵ This debate was by no means new (‘social free-riding’, i.e. the misuse of benefits, was already debated in the 1980s), but now received an additional meaning, as the focus has shifted to the ‘activation of problem groups’, groups who are unable to find their way back to the labour market on their own and therefore need to be activated. However, this specific definition of the term also implies that the people in question are somewhat patronised.

These changes occurred during a period where labour market policy in general was on the agenda but did not dominate the political discussion to such an extent as in Germany. The last two election campaigns (1999, 2002) have been dominated by family policy and the pension reform in the second place. Although both measures had a direct impact on the labour market and worsened the labour market situation of certain groups, this context was hardly mentioned. The introduction of the new child benefit scheme (Kinderbetreuungsgeld)¹⁶ lead to a declining female employment rate because of longer career breaks (cf. Lutz 2003) and will in the long run increase the need for special assistance for female returners through ALMP, which is not yet seen as a problem. Similarly the abolition of early retirement schemes (in case of unemployment or illness) was related to increasing unemployment of the elderly. These two examples show prototypically the lack of coordination between different policy areas (labour market, economic, tax, family and social policy) to be able to implement a broader, more comprehensive employment policy.¹⁷

In sum, active labour market policy in Austria is mainly addressed to the unemployed and tries to re-integrate them by diminishing their lack of skills or qualifications. There are only a few attempts to change the labour market structure or to withdraw structural barriers concerning specific groups of unemployed. One of these measures was the introduction of a ‘bonus-malus-system’ in the mid 1990s, which aimed to support the reintegration of the elderly by offering a reduction in employer's contributions to social security (Sozialversicherungsbeiträge). In this case, companies which hire older employees (50 years or older) only have to pay reduced social security fees, but in case of a dismissal they have to pay a one-time contribution.¹⁸

As pointed out in the introduction, ALMP was at the core of the German Arbeitsförderungsgesetz from 1969, and unemployment benefits were not seen as opposing the aim of bringing the unemployed back to the labour market. When unemployment figures soared in 1975 the government responded with raising unemployment benefit payments and persons participating in training measures as from 1974 received generous allowances even for 6 months following the training measure in the case of persistent unemployment (Anschlussunterhaltsgeld – ‘follow-up allowance’). But already in 1976, the Federal government began to consider public deficits a problem and training measures were the foremost target for cuts. The allowance for people in training measures was reduced and the ‘follow-up allowance’ was abandoned completely. Support for training by the Federal Employment Service (Bundesanstalt für Arbeit, BA – now renamed ‘Bundesagentur für Arbeit’, literally ‘Federal Labour Agency’ instead of the earlier ‘Office’) was made contingent on contributions to unemployment insurance in the prior years; thus, new entrants or re-entrants into the labour market were barred from these measures (Clasen 1994: 128).¹⁹

The emphasis on cost-containment became stronger in the following years, resulting also in cuts of unemployment benefits (both with respect to eligibility for benefits and to replacement levels). Another cause for the partial abandonment of ALMP was a shift of the conservative–liberal (CDU–F.D.P.) government towards ‘de-regulating the labour market’ by extending the possibilities, very circumscribed up to 1985, to sign fixed-term labour contracts. However, as has been noted by several observers (Wood 2001: 388; Ludwig-Mayerhofer 2002: 86) the take-up of the new possibilities by employers was very limited, as they are not in line with the firm–employee relationship typical for a coordinated market economy. Thus, in addition to further benefit reductions (especially concerning eligibility), the principle of occupational protection was increasingly undermined with respect to the criterion of Zumutbarkeit, i.e. the ‘suitable job’ the unemployed are required to accept on penalty of losing entitlement to benefits temporarily or permanently. Since 1998, a job is defined as ‘suitable’ after 6 months of unemployment if the net income from this job is not lower than current benefit payments, which may amount to a very substantial loss of income and status. Also, tendencies towards ‘welfare to work’ increasingly made themselves felt during the 1990s. ‘Able-bodied’ recipients of general social assistance (Sozialhilfe) (usually long-term unemployed with insufficient or no insurance-based benefits) can be forced to work in public services, receiving only small additional payments, and without any application of the principle of occupational protection. Such measures, which were estimated to have affected about 400,000 persons per year in the late 1990s, frequently were started by administrations explicitly as a ‘work test’ (Voges et al. 2000).

Yet, this is not meant to imply that ALMP generally and more specifically training measures did not play a role at all. With the beginning of the 1980s they rocketed from a low of 150,000 participants (new entrants) per year in 1976 to a high of 600,000 in 1987. Likewise, job-creation measures (Arbeitsbeschaffungsmaßnahmen), even though to a lesser degree, were available, and the expenditure for ‘active’ measures (as percentage of total labour market expenditures) was markedly above EU average already in the second half of the 1980s (Wood 2001: 385, Table 12.2). Unification even brought a decisive increase, with several millions of East Germans undergoing training measures after 1990 and several hundred thousands of persons per year in job-creation measures, the conditions for the latter having been loosened considerably (Ludwig-Mayerhofer 2002).

Why, then, the persistent failure of Germany to cope with its unemployment problem? One problem is the shift of ALMP towards disadvantaged groups such as the long-term unemployed (Wood 2001: 387). But the central problem with ALMP is that their administrative and institutional design frequently worked against their alleged aim of ‘activating’ the unemployed. Job-creation schemes always were under attack as subsidies that undermined the appropriate functioning of market and thus were not considered a serious way of coping with unemployment;²⁰ other wage subsidies for firms offering jobs to ‘the hard-to-place unemployed’ likewise were too few to alleviate the situation. With training measures one major problem lies in the corporatist nature of their implementation at the local and regional level. Many of the enterprises offering training measures are run either by employers’ associations or by the trade unions, and unemployed persons were allocated to training measures not only in accordance with their own needs and perhaps the requirements of the labour market, but also in ways that supported the established system of training institutions. Finally, the PES (run by the local employment offices, i.e. local or regional branches of the BA) turned out to be quite ineffective as well, spending most of its time with administrative routine and unable to deliver efficient services to job-seekers and employers. Needless to say, ‘welfare to work’ measures on the municipal level did not much to relieve these problems. To sum up, in contrast to appearances, ALMP in Germany are not designed for effectiveness, let alone for efficiency. Indeed, there is much to sympathize with in the views of those who see the main functioning of German ALMP in ‘masking unemployment’ (Wood 2001) which certainly is true concerning East Germany (Ludwig-Mayerhofer 2002).

A central tenet of Gerhard Schröder's 1998 election campaign had been the promise to reduce unemployment figures to 3.5 million by 2002. Yet, measures announced as steps towards this aim such as the (corporatist) Alliance for Jobs, Training and Competitiveness and a youth labour market programme targeted at 200,000 young people without a job or an apprenticeship did not help much to deliver on this promise. After 2000, Schröder continued the former government's politics of ‘quieta non movere’, in line with predictions from social scientists (see also Kitschelt 2001: 295).

Tendencies towards change gained sudden momentum in spring 2002, however. First, a report by the Federal Accounting Agency (Bundesrechnungshof), stating that the PES embellished figures about job placement success, triggered a media outcry in February 2002 that led to overnight introduction of new measures such as wage subsidies for low-income jobs (Kombilohn) and entitlements for unemployed persons to vouchers for private job placement agencies. Further activities followed, notably the establishment of the ‘Hartz Commission’ – created to develop plans for a complete re-design of labour market policies –, when in May of the same year media contrasted rising unemployment figures with Schröder's earlier promise.

The Hartz Commission report, issued in August 2002, conveys the image of radical shift towards a new kind of activation (new, that is, for Germany). It was based on the notion, similar to that of the British Third Way, that getting people back to work should be the paramount aim of labour market policies. It suggested a variety of new measures that highlighted the PES's responsibility towards employers and job seekers alike and should increase the efficiency of the PES and activation in general. The local branches of the PES were proposed to be transformed into ‘JobCenters’ (sic), supposed to integrate social welfare agencies usually related to municipal welfare provision. This was seen as a response to complaints that long-term unemployed persons had been continually shifted between the PES (to claim benefits, to be counselled concerning job placement or simply to be checked as to their ‘willingness to work’) and the local social administration (concerning additional income support from social assistance and other forms of help, but also ‘welfare to work measures’). The Job Centers are to be assisted by Personal Service Agencies (PSA), agencies for temporary work (usually private firms contracted by the PES) supposed to help the unemployed back to work by contracting them as casual workers.²¹

A number of further proposals – several of which have been put into practice in the meantime – were aimed at the unemployed and (to a lesser degree) at employers. New subsidies are one element. Self-employment in the service economy is supported by the introduction of the Ich-AG (‘Me Inc.’) or Familien-AG (‘Family Inc.’), a scheme of financial support (for 3 years) for those unemployed who become self-employed on a small scale. Low income jobs between €400 and €800 (Midi Jobs) are supported by reducing employees’ social security contributions on a sliding scale; jobs below €400 (Mini-Jobs or petty jobs) are entirely exempt from social contributions and taxes on part of the employee. Even more innovative were a number of further measures. The PSAs receive subsidies up to €1,000 per month for each contract with an unemployed person; employers who hire an unemployed person can receive a loan of up to €100,000 with comparatively low interest rates from the Kreditanstalt für Wiederaufbau (a state-owned bank giving out loans on favourable terms, usually within specific state-sponsored programs). Applicants for training measures are not allocated anymore to a specific training measure by the PES, but are entitled to vouchers they can turn in at a training agency of their choice, provided the agency is approved by the BA.²²

Many of these measures signal a departure from the continental model of social security and a move towards more ‘self-reliance’, expanding jobs in the private services and creating a low-wage sector. What seems significant, however, is a peculiar state-market mix to be found in a number of these measures: More market and thus more choice is introduced, which is often backed up by considerable state subsidies, and also frequently an element of state supervision.²³ This perhaps testifies to a more general trend, as similar tendencies can be found in other fields, such as the state-subsidized private old-age pensions available since 2002 (‘Riester-Rente’), where subsidies can be received only for contracts the terms of which have been checked and approved by the state.

If such tendencies were to prevail, they might entail some decisive new elements in activation policies – retaining a substantial degree of state influence through regulation and supervision but avoiding both the ‘benevolent compulsion’ inherent in Scandinavian activation and the more or less unconditional surrender to market forces of workfare. For two reasons, however, it seems advisable to be cautious concerning the sustainability of these developments.

First, the outcome of the current processes of re-structuring is less then certain. The large variety of measures proposed is not the result of a clear design; the impression is rather that of a desperado feeling under attack and firing his guns in all directions in the hope of hitting some enemy somewhere, or perhaps simply to impress spectators. Not least, the implementation of these proposals testifies to the difficulties of re-arranging a tightly interwoven system of local and regional corporatism, to overcome the bureaucratic nature of the BA organisation and more generally of breathing the spirit of ‘market’ into a rather ossified institutional and bureaucratic structure. ‘JobCenters’ (or ‘Agencies for Work’) thus far have not been created with the exception of a small number of model centres, as the merging of the local PES offices with municipal administrations involves tremendous legal, financial and administrative effort and negotiation. The PSAs, introduced in January 2003 with the purported aim of taking up 50,000 unemployed person, by October of the same year had contracted only about 30,000 people of which about 40 per cent were (temporarily) placed with a company on an average day (even though only 14 per cent had been long-term unemployed). Institutions offering (re-)training services have difficulties of collecting sufficient numbers of participants, as the unemployed often are left alone with their vouchers and have little guidance as to which measure to participate in. Finally, the take-up rate of the loans from the Kreditanstalt is generally considered a failure.

Second, the variegated ‘carrots’ offered to the unemployed (but presently largely on paper) are backed up by a considerable degree of ‘stick’ (which will most likely materialise as from 2005). While the merging of PES with the municipal agencies will take years, the Federal government was very quick to create a new type of grossly reduced unemployment compensation, Arbeitslosengeld II (‘unemployment benefit II’), that combines the former unemployment assistance (Arbeitslosenhilfe) and social assistance (Sozialhilfe) and will be available for those whose unemployment benefits have been exhausted. In contrast to the former Arbeitslosenhilfe which, even though means-tested, was based on the ‘equivalence principle’ and thus related to previous earnings, the new Arbeitslosengeld II will be fixed at a level barely higher than social assistance and thus in most cases considerably lower than the previous Arbeitslosenhilfe. Also, the means-test associated with Arbeitslosengeld II is much more strict than that of Arbeitslosenhilfe.²⁴ As at the same time the previous comparatively generous duration of unemployment benefits for persons aged 45 and older has been reduced significantly, the number of persons entitled to no more than Arbeitslosengeld II will be considerable. At the moment it remains to be seen whether these measures will be put into practice (due to the tremendous legal and above all administrative difficulties in merging the unemployment benefit and the social assistance system), but the seeming generosity of the state in terms of subsidizing a number of new roads into (self-)employment looks much more ambiguous in the light of this generosity being financed to a large extent from the pockets of those who live in considerable financial hardship (a feature that is continuing previous politics, see Clasen et al. 2001: 218).

What are the likely consequences for the social citizenship of the unemployed? One possible outcome is an increase in discretion, already paramount in the current system, concerning the rights of the unemployed. Already in the past, the only precise regulations concerned the benefit level and related features (such as the details of the means-test), whereas ALMP measures were available at the discretion of the PES staff only, which in its turn to a large extent was governed by the availability of means; this tendency will most likely be increased in the future. On the other hand, obligations of the unemployed may become more demanding, with the newly introduced requirement to register immediately at the PES upon notice of being fired (on the penalty of benefit reduction) being one clear example. Likewise, it seems that currently discretion on part of the PES staff concerning penalties (as for ‘insufficient co-operation’ with the PES) is being reduced on behalf of the BA president, resulting in a wave of ‘tough hands’ at the unemployed and a steep rise of sanctions (i.e. temporary benefit withdrawals) (Lütge 2003).

Several authors have pointed out that the basic features of continental welfare states make it likely that cost containment or budget control is the key aim of reform in the face of current challenges (Pierson 2001; Ferrera and Hemerijck 2003: 109), and the developments in Austria and Germany during the past decades seem to confirm this. Indeed, both countries have not changed the fundamental fabric of their welfare system and rather have engaged in piecemeal reform aimed at keeping costs of social insurance contributions at acceptable levels. Judged in terms of employment levels, activation seems to be largely absent, as especially female labour market participation is growing slowly and mostly in spite, and not because, of family policies which some signs to the contrary notwithstanding do not deviate significantly from their ‘conservative’ or ‘Christian democratic’ framing.

However, as we have shown, in both countries activation has been on the agenda particularly of labour market policies for a long time (which, as we maintain, is also a feature of continental welfare states, but referring basically to the male working age population) and more recent figures exhibit considerable sums devoted to activation measures and sizable numbers of participants in such measures. But the problem is that active labour market policies cannot deliver on their promises for a number of reasons. Among these we may emphasize first that activation policies increasingly focussed on ‘disadvantaged’ or ‘problem groups’, to the neglect of more preventive measures in terms of adjusting the labour force to the ongoing changes of demand for labour; and second that in both countries special types of ‘active’ measures were used less and less with the aim of re-integration into the labour market, but rather as a ‘secondary labour market’ for those with few chances to find a regular job (with the additional effect of keeping unemployment figures at ‘acceptable’ levels). Currently, there are attempts at restraining such developments; in Germany now an unemployed person cannot participate anymore in two immediately consecutive measures and in Austria the pressure on the unemployed is increased to participate in shorter measures with low intensity and few prospects. But it is less clear to which extent such changes simply are a continuation of recent policies of reducing benefits and cutting down measures or whether they will be backed up by more constructive measures offering more help on the way towards integration into the labour market.

Even though it seems that our arguments are in line with standard notions about the limited capabilities for reform in conservative welfare states, we do not agree with most of the explanations for what seems to be a considerable degree of ‘path dependence’. Some of these arguments see the continental welfare states in a trap from which they cannot escape (see the first section of this paper); but such arguments seem to be based on too rigid notions about the ‘right’ share of the social budget and the ‘appropriate’ level of non-wage costs (which are not so high in the final analysis when productivity is taken into account). In addition, they seem to underestimate the capabilities of cutting down costs (see Pierson 2001 on Germany). Finally, the more recent developments towards reducing benefits and tightening suitability requirements are not an immediate outcome of the public deficit per se, but rather a matter of political priorities. This is most clear in the case of Austria where the centre/right-wing government defined a ‘balanced budget’ as the main political target and used budget consolidation as a pretext to shorten the PES budget.

Another line of arguments refers to the multi-tiered political system with many veto points that prevents radical transformations. Undoubtedly this argument contains more than just a grain of truth, but again it underestimates the amount of reform that actually has taken place and the even larger amount of reform that has been tried (especially during the past few years). At this point, it might finally be argued that – as was pointed out in ⁰⁰⁰² of this paper – the big political parties (and the voters!) favour a strong welfare state, which might constrain the scope of reform. But this probably underestimates the degree to which in the recent years political programs have become much more fluid and opposition to cuts in welfare services and benefits has increasingly been marginalised in Germany (both in the SPD and the CDU/CSU) and in the Austrian ÖVP.

Overall, we wish to emphasise the extent to which changes (or their absence) have become more dependent on political decisions, which in their turn increasingly are less predictable.²⁵ Debates about fundamentally reforming the health system in both countries are a good example; they show that now ideas gain wide acceptance that could signify a decisive shift away from the continental system of financing social security basically via wage-related contributions. But currently the attempts at reform only seem to render Austrias's and Germany's political economies more vulnerable, for two reasons. First, policies in different fields exhibit considerable inconsistencies; reforms in the fields of health, old-age pensions, education, family and the labour market do not follow a coherent plan and sometimes work along different lines (perhaps most notoriously in the case of family versus labour market policy). Second, in their zeal to ‘modernise’ politicians seem to be much more susceptible to popular but immature and misguided ideas. The case of PES reform provides a telling example. In both countries, PES rightly has been accused as too bureaucratic and therefore has been subject to reforms aiming at more efficiency. But to claim the imperfect functioning of the PES as one of the main causes of unemployment, as it currently happens in Germany, and to create the impression that an efficient PES together with cuts in benefits will cause significant relieve on the labour market is more than the public seems ready to believe. In contrast in Austria, where reform of the PES was brought into effect much earlier and with much more success, it is now curtailed for political reasons.

In sum, currently both governments intend to create the impression of energetic action to cope with the current challenges, but at the same time lack the skill to do so. In Austria the ‘speed-kills-strategy’ has led to several law revocations because of technical problems. Similarly the German political agenda exerts pressures towards reform of labour market policies, but after 18 months it is doubtful whether both the political and the legal-administrative obstacles will be overcome. One thing is to introduce new measures, but another is to introduce good measures; some of the current measures seem to be more a result of attempts at reform for reform's sake and not the outcome of a carefully crafted plan. It is unlikely that those pieces of the current reform packages that have been put into practice will make themselves felt on short notice; and if the current economic slack continues for a couple of years, everything may go up in smoke. Ironically, thus, there is some likelihood that the final outcome in Germany will be similar to the developments in Austria, i.e. ‘bashing the unemployed’ – prepared by repeated statements from the Federal Minister for Economics and Labour Wolfgang Clement – and the concomitant reductions of the entitlements of the unemployed, both with respect to benefits and to promotion of employment opportunities. While it is unlikely that we are facing the development of a mere ‘workfare state’ in Germany or Austria, it seems that politics have less trouble with getting tough at those with the least power to resist than with a courageous and energetic re-design of labour market policies.

Wolfgang Ludwig-Mayerhofer is Professor of Sociology – Empirical Research Methods at the University of Siegen, Germany. His main areas of research are inequality, poverty and unemployment, sociology of the family and sociology of (criminal) law. Among his recent books are Die Armut der Gesellschaft (edited with Eva Barlösius, 2001), Soziale Ungleichheit, Kriminalität und Kriminalisierung (ed., 2000) and Soziologie des Sozialstaats (edited with Jutta Allmendinger, 2000).

Angela Wroblewski is researcher at the Institute for Advanced Studies and lecturer at the University for Economics and Business Administration in Vienna, Austria. Her main areas of research are evaluation of labour market policy, female labour market participation and studies in the field of higher education.