NORMALIZING ACTIVE EMPLOYMENT POLICIES IN THE EUROPEAN UNION: The Danish case

The paper examines some of the ways in which the problematic of activation is put into play in the European Employment Strategy (EES) and Danish employment policies. The limitations of causal analysis and convergence approaches are discussed and a different analytical framework based on the concept of normalization is proposed. It is argued that while various forms of activation were present in Denmark well before the launching of the EES in 1997/1998, the latter has contributed to the normalization of the Danish employment policies by sustaining activation as the key element of such policies and by disregarding political interventions based on alternative problematizations of labour and well-being.

The notion of activation has become a key element in a wide array of political strategies and programmes over the last 10 years or so. It is used both inside and outside European polities and is increasingly part and parcel of the policy documents pouring out of Brussels. Thus, activation is not only an issue of national employment policies, but also a fundamental feature of the European Employment Strategy (EES) launched in 1997. The EES has been the object of quite extensive analyses by political scientists and sociologists. The bulk of this research tends to revolve around one or two rather narrow questions: Are national employment policies being Europeanized and are they converging (or not)? Hence, while this research has produced many insights, the EES does seem to be in need of an analysis based on other conceptual and methodological approaches.

Based on the notions of normalization and problematic, this paper examines the attempt of the EES to promote activation in the Danish employment policies. More precisely, I pursue two broad questions. First, in what ways is the problematic of activation both informing and engendered by the EES and the Danish employment policies? Answering this question implies scrutinizing how the various programs, schemes, procedures and techniques claiming in one way or another to promote activation are put into play through the EES and the Danish employment policies. Second, what normative assumptions about working life is the problematic of activation supporting? Answering this question includes analyzing the ways in which alternative suggestions about the governing of working life, that is, alternatives to increasing participation in the labour market, are regarded.

The choice of the Danish employment policies may seem a curious one considering that Denmark embraced activation well before the launching of the EES in 1998. Yet, by choosing the Danish employment policies, I am trying to show that the EES serves to promote and sustain the problematic of activation rather than claiming that the EES is trying to push a new policy approach (activation) onto the employment policies of the member states. Thus, I try to shed light on the ways in which the EES promotes and sustains the problematic of activation in the Danish employment policies at the expense of other problematics. My overall argument is that the significance of the EES rests not on its ability to make Danish employment policies converge to some abstract ideal, something that the EES really does not intend to do. Rather, the major significance of the EES springs from the ways in which it contributes to sustaining and updating the problematic of activation as a key element in Danish employment policies at the expense of alternative problematics of work and wellbeing.

There is currently little agreement on the methods to be used in studying the significance of the EES for national employment policies (Barbier 2005). This paper uses the concept of normalization to examine the significance of the EES. In order to circumscribe this specificity of this concept, I will first provide a brief account of two dominant methodological approaches (causal and convergence analysis) and then describe the concept of normalization and outline its analytical implications.

A large part of the literature on European politics and policies adopts a cause–effect approach. One such example is Börzel and Risse's ambition of trying to measure the domestic effects of ‘Europeanization’ (Börzel and Risse 2000; see also Caporaso et al. 2001). Here the EU as an organization or more broadly ‘Europeanization’ as a ‘process of institution-building at the European level’ (ibid. 3) is taken as the independent variable and domestic effects as the dependent one. In accordance with this approach, Goetschy postulates that ‘the EU employment guidelines will induce a gradual Europeanization of certain elements in national employment policies’ (Goetschy 1999: 134). In short, the essential question in the causal analysis becomes: How and why are the goals of certain politico-administrative organizations/actors able to produce policies that are implemented by politico-administrative units at a lower level in a legally stipulated hierarchy? However, as Börzel and Risse, and many other scholars are aware, it is extremely difficult to measure the Europeanization effect because the process of Europeanization is itself shaped by member states. Thus, member states use their seats in the Council and various access channels to the Commission in the attempt to shape EU policies according to their (national) interests. Consequently, it is very difficult if not outright impossible to uphold a clear distinction between Europeanization or EU policies as the independent variable and member state policies as the dependent variable.

Some have tried to add nuances to the Europeanization approach by talking of a ‘leverage effect’, which essentially involves analyzing how national actors use EU policies, such as the EES, for their ‘own’ purposes (Erhel et al. 2005: 229). While this is no doubt an important correction to the view of the policy process as a rather straightforward top–down process, it retains a traditional causal analysis in which various actors at different politico-administrative levels try to maximize the realization of their ‘own’ pre-given interests. I think this approach is not entirely suitable to account for the ways in which the EES is able to shape what is regarded as problematic, what requires intervention, and thereby what political actors show an interest in. If political interventions and the interests articulated through these take place in relation to one or more concrete political problems, or rather problematics (see below), rather than in a void, then we need to address the EES’ ability to promote and sustain these problematics.

A second dominant approach in the EU policy literature is to study the effects in terms of convergence–difference. Here the concern is less with the disentangling of causal relationships and more with the question of to what extent we actually see a convergence of national employment policies around the broad political line suggested by the EES. Principally, this question only allows two answers: those arguing that difference prevails due to nation-specific historically institutionalized values and policy-making styles, and those arguing that convergence has taken place because of either causal or isomorphic pressures. In the first group, which currently is the dominant one, we find, for example, Jean-Claude Barbier who argues that while the EES has influenced the member states’ employment policies in terms of procedures by creating a ‘common language in the area of employment and social policy’, it has done little if anything to change the ‘substantive aspects’ of these policies (Barbier and Ludwig-Mayerhofer 2004: 434; see also Erhel and Zajdela 2004). Henning Jørgensen, who shares this opinion, perceptively likens this to karaoke: the attempt to sing like others rarely succeeds (Jørgensen 2005: 36). Similarly, Claudio Radaelli has argued that, ‘it is reasonable to expect that even a “perfect” application of the OMC will result in a more diverse Europe’ (Radaelli 2003: 9). In the second group, which seems to constitute a minority, the elements of convergence are stressed (e.g., Lindsay and Mailand 2004). Whether the mechanism of convergence is a causal one (where actors with a higher level authority manage to persuade or coerce actors with a lower level of authority) or a form of isomorphic pressure enabled by shared moral and/or cognitive frameworks (DiMaggio and Powell 1983), the prediction is that member states will pursue increasingly identical policies.

In an attempt to make the answers provided within the difference/convergence approach more nuanced, Amparo Serrano Pasqual makes a distinction between the methods of activation, which according to her still display national differences, and the ethical and ideological legitimation of policies, which are converging (Pascual 2004). She goes on to argue that activation is associated with cross-national changes in the discourse on social protection, including both a reworking of what is seen as problematic and the type of subject invoked. I think Pasqual's approach is valuable in as much as it allows us to address the ways in which political procedures and language interrelate, and thereby enable certain types of political practices at the expense of others. However, I also think we have to fully transcend the convergence approach in order to open an analytical space that may make visible the ways in which the EES contributes to normalize member state employment policies.

Thus, instead of pursuing a strict causal analysis or an analysis of convergence, I propose to shed light on the ways in which the EES contribute to normalize a very specific and therefore excluding problematic of employment informing the employment policies of the member states. Normalization is the process and procedures through which a norm is informing the set of political practices that it seeks to regulate (Ewald 1990; Triantafillou 2004). It is a process where all have to refer to a particular norm or a particular problematic in order to develop, debate or contest existing political practices. Normalization could, for example, refer to a process through which one may contest particular understandings of activation and/or suggest various reforms to improve activation, but in which one can neither morally be against activation nor factually dismiss (the lack of) activation as a key problem that any employment policy has to deal with. It should be stressed that normalization does not imply, at least not necessarily, a consensus on a particular way of doing things. We may very well see intense disputes, as in the case of the EES and the Danish employment policies, over how activation should be defined and over how it should be promoted. Yet if the problematization of the labour market in terms of activation, and thereby the very need for activation (in whatever form) remains unquestioned, we can have a situation with both intense disputes and strong normalization.

A problematic (or problematization) is a name for the way in which a social phenomenon is rendered problematic, e.g., the kind of problems we identify in relation to the functioning of the labour market or labour in general (cf. Foucault 1987: 11–12). By enframing the problems of the labour market in a particular way, we also limit the possible answers suggested to deal with these problems. For example, if the major problem of the labour market is regarded to be a lack of economic incentives to work, the answer is obviously to invoke some kind of mechanism(s) that will augment economic incentives. I regard problematics and political activity as mutually constitutive. On the one hand, all political activities in the sense of actions trying to govern the action of others are informed by one or more problematics of governing. In other words, political actions are, by my definition, always informed by more or less well developed problematizations, i.e., by calculations, forms of knowledge and moral valuations that turn certain social phenomena into ‘problems’ calling for political intervention. On the other hand, the problematics of governing activities only exist to the extent that they are inscribed in social and political practices. The notion of problematics is somehow akin to ‘the argumentative turn’ in policy analysis (Fischer and Forester 1993). However, much in line with Bruno Jobert's and Pierre Muller's concept of ‘referential’ (Jobert and Muller 1987), a problematic can neither be reduced to a mentally constituted belief system (Sabatier and Jenkins-Smith 1999: 132–5) nor to a linguistically constituted discourse (Fairclough 1995). If we want to render intelligible the ways in which new understandings of a social issue is produced, disseminated and updated, we need to pay attention to the material or technical dimension of these processes as well. Thus, a problematic owes its existence not (only) to mental predispositions or linguistic representations, but to the concrete techniques of data production, policy-making procedures, and schemes for the exchange of knowledge. Consequently, the analysis of the normalizing significance of the EES implies paying attention to the mutually constituting interplay between the problematic of activation on the one hand, and the mundane techniques, procedures and schemes found in the EES and the Danish employment policies on the other.

Before proceeding to the analysis, I need to specify what I intend to examine under the heading of ‘activation’. Some authors suggest a broad understanding of activation. For instance, Halvorsen and Jensen characterize Danish and Norwegian labour market policies since the late 1950s as activist, because they persistently tried to expand the rate of labour market participation (Halvorsen and Jensen 2004). However, such a broad conception, which essentially equates activation with more or less extensive political interventions, not only makes it difficult to circumscribe the analytical object, but also risks overlooking the novelty of the employment policies emerging from around the late 1980s onwards. Others operate with a much more narrow definition that reduces the analysis of active employment policies to the ways in which the unemployed are being activated (e.g., Lødemel and Trickey 2001). However, such a narrow definition risks overlooking some of the ways in which the problematic of activation informs not only other parts of the employment policies, but also other policies such as social and tax policies (Barbier and Ludwig-Mayerhofer 2004: 423). I have chosen a middle of the road solution, whereby I will mainly focus on activation interventions found within employment policies, but I will also touch upon some of the ways in which the problematic of activation has informed changes (or proposals to change) social security and taxation policies.

Denmark has quite a long history of interventionist labour market policies. Since the early 1960s, Danish labour policies have relied on a high level of income compensation for the unemployed, education and skills upgrading, and mobility enhancing policies (Lind 1992). Despite the persistent economic recession starting in the early 1970s, which soon made unemployment rates rise to high levels, employment policies only saw few changes. The changes that were made essentially sought to ameliorate the economic conditions of the unemployed, they did not impose obligations of, for example, active job-seeking.

The first significant change of the Danish labour market policies in the direction of activation took place in 1989. That year the Liberal-Conservative government's proposal for a law on the activation of unemployed youth was adopted (Torfing 2004: 175). The law changed the voluntary job training offer into a precondition for young people's reception of social assistance. Despite several criticisms of the law from the opposition (including the Social Democrats) and the labour movement, the idea of activation gained further impetus. A key document published by the Ministry of Social Affairs entitled ‘There is a need for everyone’, states that the new social policy for both youth and adults assumes that ‘everybody who can, should contribute to society in return for support’ (Socialministeriet 1990: 1–2). The new policy tried to ‘turn the passive income support into an active something-for-something approach’ (ibid. 1–2). However, the real breakthrough for the activation approach in Danish employment policies came in January 1994 when the law on active labour market policy proposed by the new Social Democratic government came into effect (Torfing 2004: 204–6). The background of the reform was a record high unemployment rate of more than 12 percent and a new understanding of the labour market couched in terms of ‘bottleneck’ problems (the result of inadequate skills and mobility of the workforce).

The key change of the 1993/1994 reform was to divide the 7-year unemployment insurance period into two: in the first 4 years, the unemployed will receive voluntary job or training offers, in the last 3 years the unemployed must accept job or training offers to remain eligible for unemployment insurance payment. The law also introduced individual action plans, i.e., contracts between the public employment service and the unemployed person that identified job seeking opportunities and job training activities. Only through satisfactory participation in these activities would the unemployed be entitled to receive unemployment insurance payment. Finally, the law reduced the unemployment insurance payment rate for young people to a level equal to student allowances. Over the next few years, the Social Democratic government gradually shortened the period in which the unemployed were eligible for unemployment insurance (Torfing 2004: 34, 214). In 1998, the law on active social policy was adopted whereby compulsory activation was extended to all social benefit recipients (Torfing 2004: 238). While this shift in employment policy from a problematic of social security to one of activation was met with resistance by the political parties on the far left and some of the unemployed, it was supported by all the major political parties, the representatives of the employers and the employees, the economic experts and the major newspapers. In sum, the inscription of activation as the key problematic supporting Danish labour market policies emerged only from the late 1980s and was consolidated around 1994. By the same token, it is clear that the problematic of activation heavily informed Danish employment well before the advent of the EES in 1998.

The speeding up of European economic integration in the early 1990s (through the completion of the internal market), the establishment of the Economic and Monetary Union and the adoption of the Stability Pact rendered some of the traditional employment policy tools obsolete (Goetschy 1999). Attempts to improve international competitiveness through devaluation or adjustments of national interest rates together with attempts to stimulate national demand through public deficit policies and state subsidies were more or less de-legitimized. By formulating a new link between economic growth and employment, the 1993 Delors White Paper ‘Growth, Competition and Employment’ contributed importantly to paving the way for the EES (CEC 1993). The White Paper thus regarded unemployment less as a ‘cyclical’ (macro-economic) problem to be regulated through fiscal and monetary policies and more as a ‘structural’ (micro-institutional) problem to be dealt with through interventions spurring on technological development, enhancing flexibility in labour markets, and improving the skills of labour through education and training.

The new problematic supporting a strategy tackling unemployment through micro-institutional interventions within a framework of macroeconomic stability was further developed over the following years. At the Essen summit in December 1994, the member states agreed on five broad priorities and a multilateral procedure to monitor the member states’ progress on these (European Council 1994). Activation played a key role in three of the five priorities. The Amsterdam Treaty of 1996 subsequently devoted an entire chapter to employment making ‘a high level of employment’ an explicit priority to be dealt with through ‘a coordinated strategy for employment’. In order to convert this goal into practical policy-making in the member states, the Amsterdam Treaty introduced four broad mechanisms:

• 1.

  Yearly guidelines envisaged to be the dynamo of the employment scheme.

• 2.

  Annual assessment that could result in national specific policy recommendations.

• 3.

  The creation of an Employment Committee with the task of monitoring the employment situation in dialogue with the social partners.

• 4.

  Authorization to apply incentive measures, such as benchmarking and pilot projects.

In the following, I examine the guidelines and the benchmarks on activation (mechanisms 1 and 4). Next, I examine the recommendations issued to Denmark (mechanism 2). Finally, I analyze the making of the Danish National Action Plans (NAPs) (part of mechanism 4). I have decided not to analyze the regular peer review/mutual learning sessions organized by the Employment Committee (mechanism 3) because my preliminary study suggests that these sessions play a minor role in the normalization of the Danish employment policy.¹

The activation of the unemployed played an important role from the first EES adopted in 1997, notably in the first pillar dubbed ‘increasing employability’ and in around five of the 20 odd guidelines.

After an evaluation of the first 5 years experience with the EES (CEC 2002), the EES was reformed with effect from 2003. This entailed that the four ‘pillars’ and 24 ‘guidelines’ were replaced by 10 new guidelines and three ‘overarching and interrelated objectives’. The latter were: full employment, improved job quality and productivity, and social cohesion and inclusion. The 10 guidelines, which were to remain constant for 3 years, contained little new. Five of them were directly targeted at enhancing activation by activating the unemployed, developing human capital, making people remain longer on the labour market, and making work pay (guidelines 1, 4, 5 and 8).

In 2005, the employment guidelines were changed once more, though the focus on activation was retained. In order to consolidate the attempts to make member states promote activation, the EES guidelines were merged with the Broad Economic Policy Guidelines (BEPG) (Council European Union 2005). By the same token, the EES guidelines were reduced from 10 to eight. Four of these deal directly with activation of the unemployed by enhancing the employment rates (GL 17) promoting a life cycle approach to work (GL 18), ensuring inclusive labour markets (GL 19), human capital development (GL 23), and reforming education systems (GL 24). The Council had decided that the guidelines should remain constant until 2008 (Council European Union 2006).

It is this comparative data that forms the basis of benchmarking analyses found in the annual Joint Employment Reports (starting in 1998) by the Commission. The identification of ‘best practices’ or ‘good practices’ is essentially based on a comparison of two elements, namely the quantitative employment indicators and more qualitative evaluations of the member states policies’ attempts to fulfil the guidelines (or the pillars). The benchmarks were adopted by the Council in July 2003 (Council European Union 2003).

In Table 1, I have mapped the employment guidelines dealing somehow with activation adopted in 2005 and the corresponding benchmarks adopted in 2003, which are still valid. I have only included the guidelines that deal in some way with activation. In order to provide an idea of the room for national manoeuvre, I have indicated whether or not the benchmark has a deadline. Likewise, I have indicated whether or not a benchmark is formulated authoritatively (‘shall'or ‘will’) or simply as a declaration of intent.

The overview of the guidelines shown in the table suggests a rather wide room for the making of national employment policies. Even if six out of the eight benchmarks actually have a deadline, namely 2010, only three out of the eight benchmarks are formulated authoritatively. In sum, only two guidelines have both a fixed deadline and are formulated authoritatively, namely GL17 and part of GL18. I take this as an indication of a strong consensus among member states on the goal of increasing employment rates (in general and for particular groups) and making people stay longer the labour market. In most other areas member states apparently want to retain a wide room for manoeuvre with regard to the way in which they address activation.

While the member states have a wide scope for the way in which they address activation, it is clear that anything does not go. Rather paradoxically, guideline 19, which stands out by its lack of a direct benchmark, is perhaps the clearest sign of this limited room for the way in which activation may be turned into practice. In contrast to all the other employment guidelines, the GL 19 on promoting inclusive labour markets has two benchmarks that refer not to the substantive goal of an inclusive labour market, but to the procedural aspects of member state policies. Both these procedural benchmarks are targeted towards the activation of the unemployed, not with the reforming of the labour market structures with a view to making this more inclusive. Thus, the pursuit of activation is primarily restricted to inciting, training, and ‘preparing’ the unemployed to getting a ‘real’ paid regular job.

Perhaps the most significant effect produced by the guidelines and the benchmarking analyses may not be the relative positioning of member states policies according to common defined activation indicators, but that the member states are induced to set their own activation targets. For example, guideline 23 (formerly GL 6) not only urges Member States and the social partners to develop possibilities for lifelong learning, but also urges Member states to set their own targets for participants benefiting from such measures. Thus, after having judged the Danish measures for developing lifelong learning as ‘adequate’, the Joint Employment reports proceed to evaluate how the Denmark fared according to its own goals (CEC 2000: 42–3). Of course it could quite rightly be objected that by letting the member states set their own goals, these are bound to be lax in order to minimize changes of their employment policies. Yet, I would argue that it is exactly by letting member states define their own goals that activation becomes their own policy, i.e., member states subject themselves to the activation problematic. Thus, no matter how lax the member states’ own goals may be, they are nevertheless being formulated within the problematic defined by the guidelines on activation, not within some other problematic. But let us now turn to the way in which the EES has been played out in Denmark.

Since 2000, each and every member state has received annual recommendations. These recommendations, which are based on the guidelines and the benchmarking analyses, are proposed by the Commission and formally adopted by the Council (Table 2).

An X signifies that Denmark received a recommendation on the mentioned issue in the said year. ²

While the member states are not obliged to do anything to accommodate the recommendations, the Danish case suggests that even if the recommendations do not give rise to a new national policy, they do provoke very comprehensive responses from the government. The persistent response of changing Danish governments to the issues raised by the recommendations is that they are already dealt with by Danish employment policies.

In the following, I analyze the responses of the Danish governments to the three most persistent recommendations, namely the ones on integration, retirement and income taxation. First, the integration of immigrants refers to initiatives seeking to alleviate the significant difference in the employment rates between immigrants and native Danes. Both the Social Democratic and the Liberal governments had acknowledged well before 1998 that immigrants in Denmark have been and still are inadequately integrated into the labour market. As a result of their relatively low level of employment they are regarded as being too dependent on welfare benefits and by implication not active in the proper manner, i.e., by having a paid regular job. Thus, the Commission's recommendation that Denmark intensify its efforts to raise the employment rates for immigrants has been absolutely uncontroversial.

Second, the recommendations concerning taxation vary in some ways over the years, but essentially propose that Denmark reduce its high level of marginal income tax, particular for the low and middle income groups in order to increase the economic incentive to work. When the new Liberal-Conservative government came into office in November 2001, it decided to implement a tax freeze. The new government argued that the tax freeze would enhance employment not only by a general stimulation of purchasing power, but also by providing a higher economic incentive to work. Now, while it would be absurd to claim that this decision was the result of the Commission's recommendation, it is interesting to note how well the Danish government's argumentation resonated with the Commission's recommendation. Accordingly, in the NAPs issued since 2002, the Danish government has explained its tax reform partly as a response to the recommendation of lowering income taxes (e.g., Ministry of Employment 2002: 4, 12; Ministry of Employment 2003: 28, 40).

The third recommendation concerning retirement refers to initiatives seeking to make the workers stay longer (more years) on the labour market before they retire. In particular, the Danish early retirement benefit was criticized by the Commission for lowering the supply of labour. Despite the fact that so-called early retirement was the topic of a very controversial reform in 1999, the Commission's recommendation on this issue was accepted by the Danish government (Ministry of Labour 2000: 26). On the one hand, the left wing parties and the nationalist Danish People's Party were against a reform of the early retirement benefit scheme. On the other hand, the Liberal, the Social Liberal, the Conservative Party, the employers, and most leading economic experts were strongly favouring a reform. The Social Democratic Party was caught in the middle, felt compelled to make a ‘half-way’ reform and ended up loosing popular confidence and the 2001 election. Thus, even if it is correct that the Social Democratic led government, around 2000, made the Commission soften its public recommendation on the Danish retirement policy (anonymous oral source), this only testifies to the normative strength with which the problematic of active ageing/late retirement from the labour market influenced the actions of the Danish government. The problematic of active ageing simply required government action directed towards some way of inciting older people retire later from the labour market.

In brief, all three recommendations revolve around the understanding that in order to be active one has to participate in the labour market in the sense of having a paid regular job. The recommendations on better integration of immigrants and the postponing of retirement directly states that activation implies labour market participation. And the lowering of marginal income level tax is a device to make it profitable to take (more) paid labour, and thereby to be active – in the right way. This problematic seems by and large to be accepted by the changing Danish governments. At least, they are adamant in their insistence that their positive response to the Commission's recommendations on activation is due to the close fit between these and the Danish long-term national employment strategy, not due to normative, legal or other forms of pressures. Perhaps then this process of internalization or subjectification is the most powerful aspect of the EES: any recommendation on strengthening activation in the sense of a regular paid job issued by the Commission is not only accepted (in some form or another) by shifting Danish governments, they also claim and for all intents act as if the accepted recommendation is part of their ‘own’ policy!

If the recommendations have made successive Danish governments insist that the problematic of activation is their own, what about the Danish employers and employees (the so-called social partners)? A survey of the major Danish social partners’ view on the recommendation shows that they do not feel challenged by the recommendations (Ginger-Mortensen 2006: 86 ff).³ Even the recommendation to postpone retirement, which could be seen as a critique of the Danish early retirement scheme that the Danish labour unions are defending, has not produced any outcry. While the Danish National Trade Union admits that they are ‘irritated’ every time the Commission issues this recommendation, the Union – according to itself – does little to try to stop these recommendations from being issued (Ginger-Mortensen 2006: 86–7). The Danish National Trade Union actually shares the position that measures should be taken to create incentives for people to retire later from the labour market, only they want to decide how or by what mechanisms this should be accomplished. As long as the Commission's recommendations on activation remain at the level of overall goals, the Danish social partners seem inclined to preserve their efforts for the domestic scene where the struggle over of the choice of concrete policy instruments are played out.

It would be wrong to say that there are no critical voices in Denmark of current activation measures. In fact, the forced participation of unemployed persons in futile job training programs and the recent (2007) reduction of social benefits for long-term unemployed persons have been heavily criticized by the leftwing parties, social workers and academics. However, even the political opposition implicitly supports activation in the sense of getting as many people as possible into a regular paid job. The only difference is that they envisage these new jobs being created through public financial support. They agree that unemployment creates a situation of personal dependency which is erodes the personhood of the unemployed. Therefore, even if Danish society could afford paying social benefits to the unemployed, it is a moral obligation of society to activate the unemployed. In short, even those strongly critical of current forms of activation are actually supporting the norm informing these measures.

Since 1998, all member states have obliged themselves by signing the Maastricht Treaty to formulate annual National Action Plans (NAPs) for employment. They were succeeded in 2005 by the National Reform Programmes (NRPs) which includes not only national employment policies but also macro-economic policies and reflections on the ways in which they may be coordinated with the employment policies. The NAPs are issued by the national governments who structure the plans according to the joint employment guidelines. Hence, all member states structure their plans in the same way, though the plans vary significantly in length. Moreover, all national actions plans have to respond to the recommendations issued by the Commission in the Joint Employment report as a result of the benchmarking of the performance of each and every member state the previous year (see above).

The process of formulating the Danish NAPs and later the reform program has run more or less constantly since 1999 in the following way.⁴ A specific set of social partners are urged every year by the Ministry of Employment, which is responsible for formulating the NAP to send in joint written contributions. Parts of the contributions by the social partners are incorporated in the main text of the NAP/NRP, and the entire contributions are put in appendices. The social partners emphasize that their contributions be seen as reports on already adopted measures that in one way or another resonate with the joint guidelines. This perception is in line with the content of the final NAP/NRPs that above all report on the decisions already made by the Danish government. As explained by a senior civil servant in the Danish Ministry of Employment, the (few) future plans that are mentioned in the NAPs are invariably explicated as the result of the Danish government's own strategy (Binder 2005).

The participants in the formulation of the NAP include the ‘usual suspects’ from the formulation of national labour market agreements and policies. They are the representatives of the employers and employees in the private and public sectors (at both state and municipal levels). This also means that several groups such as the National Organisation of the Unemployed, who would have liked to participate in the process, are not included. Although the actors involved in the NAP are identical with those involved in usual domestic labour market policy-making, their roles differ significantly. Usually the making of nation wide labour market agreements is characterized by highly institutionalized negotiations between the representatives of the employers and the employees (at both private and public levels) with the state as a mostly silent observer. However, in the case of the NAP, the state (the Ministry of Employment) plays a much more active role by being responsible for formulating the NAP and coordinating the process. The social partners are invited and participate regularly and contribute written inputs to the NAPs, even if they (the partners) complain about being included rather late in the process each year (Jacobsson 2005: 118; Ginger-Mortensen 2006: 83).

While the Danish employment policies mentioned in the NAP/NRPs are never put forth as a response to the joint guidelines or specific recommendations, they are nevertheless invariably framed in line with the structure, vocabulary and lines of argumentation put forth in the joint guidelines. Thus, every year the Danish government reviews its employment policies and their performance in augmenting employment through various forms of activating the unemployed. Accordingly, all Danish NAPs on employment and the two recent NRPs invariably begin with an account of the main employment figures pertaining to GL17, namely the overall employment rate, the female employment rate, and the senior's employment rate (e.g., The Danish Government 2006: 40). And each year, the successive Danish governments have had a hard time hiding their happiness with the fact that not only do Danish employment rates exceed the 2010 goals stipulated by the EES, they are also well above the EU average employment levels. So far, the results of these annual benchmarking analyses have persistently been taken by the Danish government and the major Danish social partners alike as a clear sign of the moral ‘fact’ that activation – in the sense of inciting, training and preparing the unemployed to take a regular paid job – is working and should remain the cornerstone of Danish employment policies.

Danish policy-makers and top civil servants in the Ministry of Employment tend to depict the EES in general and the NAP/NRPs in particular as technical reporting mechanisms with limited if any impact on Danish employment policies (Binder 2005). Academics have similarly argued that the effects of the EES on Danish employments are negligible. If anything, they argue, it is the Danish policies that have influenced the EES (Madsen 2003: 154). While I agree that the EES is not a tool for the formulation of new laws and regulations, I have tried to show that the significance of the EES may lie elsewhere, namely in promoting and sustaining a very particular understanding of labour and employment, what constitutes the problems of labour and employment, and thereby what types of interventions may be launched to tackle these problems.

To be a bit more precise about the significance of the EES, let me try to answer the two questions posed at the beginning of this paper. First, in what ways is the problematic of activation at the same time informing and engendered by the EES and the Danish employment policies? In many ways the Danish – and Swedish – activation approach was a source of inspiration for the EES, notably its pillar 1 on enhancing employability (Lefresne 1999). Unsurprisingly then, the benchmarking analyses of the Danish activation policies for the unemployed have generally been very positive and resulted in discussions on how these policies could be transferred, in a more or less modified form, to other member states (e.g., CEC 2001). Therefore, the Commission's recommendations to Denmark since 2000 rather than pushing a new problematic have served to sustain and update the problematic of activation which has fundamentally informed the Danish employment policies since the early 1990s. Since then, successive Danish governments have consistently identified dependence and inadequate activation as the key problem of the labour market. Accordingly, the increased and more systematic activation of the unemployed in order to get them into paid regular jobs is regarded as the most important, albeit not the only, element in solving this problem.

Since the inception of the EES in 1998, the successive Danish governments have actively participated in the EES, a fact that should not be taken for granted. As I have showed the Commission issued numerous recommendations that at least in some areas contain a quite strong criticism of Danish employment, social security and taxation policies. Moreover, the activation approach pursued by the shifting Danish governments have at times been strongly criticized by parts of the oppositions (the left wing parties), parts of the unions and not least by the unemployed themselves for forcing people into what they see as poor/low paid jobs or meaningless job training schemes. On top of all this, even if the Danish government is bound by the employment objectives of the Amsterdam Treaty, the latter contains no means of sanction. Yet, instead of simply disregarding uncomfortable benchmarking analyses and recommendations, the successive Danish governments have provided detailed explanations to show how their policies already or in the future accommodate these recommendations.

Now, one may speculate as to whether the perseverance with which shifting Danish governments have pursued employment policies informed by the activation problematic would have been any different had the EES not been there. Of course this can be neither affirmed nor rejected. But what we can positively see is that on the one hand the EES has consistently since 1998 contributed to reaffirming the activation problematic through a number of soft governing mechanisms, such as the national benchmarking analyses that apparently provide hard evidence of the success of Danish activation model in creating and sustaining high levels of employment. On the other hand, we can also see that the recommendations given by the Commission often are at odds with the way in which shifting Danish governments have envisaged the concrete unfolding of activation. In short, while the problematic of activation is thoroughly informing both the EES and the Danish employment policies, the latter two are not identical. Accordingly, we cannot simply reduce the one to the causal effect of the other.

Let us then move to the second question: What normative assumptions about working life are the problematic of activation found in the EES and the Danish employment policies supporting? The problematic of activation found in the Danish employment policies seems to imply a shift from a problematization of employment in terms of the overall demand of the national economy, which could/should be regulated through fiscal and possibly monetary policies, to one in which employment is regarded as a problem of structural and institutional barriers within the economy. According to the latter understanding, the enhancement of the supply of labour (through employability and/or economic incentive structures) is seen as a necessary, but an insufficient measure. Thus, we are not simply moving from a Keynesian inspired demand-driven policy to a Monetarist inspired supply-driven policy. Demand management is still a fundamental part of employment policies, though now as an issue of micro-institutional settings not one of macro-economic conjunctures. As suggested elsewhere, the neo-liberal problematic of activation informs not only employment policies (in the narrow sense of that word), but also education, industrial and taxation policies, which in turn are recast and re-coordinated so as to promote the making of active subjects fit to serve the ‘active society’ (Dean 1995) or the ‘competition state’ (Cerny 1997). In brief, a major significance of the EES is its ability to make Denmark (and possibly other member states) question, scrutinize, plan and measure the ability of its employment policies to promote activation.

To pinpoint the normalizing significance of the EES, it may be worth developing Henning Jørgensen's karaoke analogy (see above): the major point about karaoke is not only that the attempt to sing like others rarely succeeds, but that it makes you refrain from singing other songs! Or, in the context of EES, that it makes Denmark and possibly other member states refrain from formulating employment policies based on a different problematic. The appeal of Karaoke is that we are all able to sing along even if we are not able to sing the same notes and remember all the words of a particular song. Likewise, the strength of the EES is that it urges member state to pursue activation even if they do it ‘their way’, i.e., according to historically institutionalized employment systems and organizational setups.

The analogy may even be drawn a bit further: In the same way that karaoke did not invent the song that participants try to sing, the EES did not invent active employment policies. Thus, what is new about the EES is not that it deals with activation, but that it urges member states to reproduce and possibly update employment policies in which measures promoting activation in one way or another is compulsory. By implication, the major danger of the EES is that it may cut off any other problematic that could serve as the basis for governing employment issues differently. The fact that it is so difficult in Denmark to be against the principle of activation, though one may dispute the exact way in which this is done, indicates that this is a danger that ought to be taken seriously.

Peter Triantafillou, PhD, Associate Professor at the Department of Society and Globalization, Roskilde University is interested in the games of power, government and freedom in the fields of employment, human resource management, and new public management. He has published several journal articles on these topics.