Unemployment policies in six west-european countries

Transcription

1 ECPR General Conference in Reykjavik, August 2011 Unemployment policies in six west-european countries Political actors positions in the policy space First draft 15. August 2011 Abstract This paper analyzes how political conflict in labour market policy is structured across six European countries (Denmark, France, Germany, Italy, United Kingdom and Switzerland), by applying a Multidimensional Scaling (MDS) analysis on the preference data of all major political stakeholders, i.e. unions, employers organisations, parties, state bodies and administration, as well as research institutes and charities active in this policy field. In fact, as a consequence of the globalisation of labour markets, new social risks that create new differentiations between winners and losers of globalisation can be expected to arise and thus change the preference structure of political actors constituencies. The findings show that beyond the traditional class cleavage that shapes the contention over left or right policies, a second dimension, dealing with re-commodification policies appears. Even though this more recent conflict dimension concerning activation, or, liberalization policies, is not uncorrelated to the left-right divide, it proposes some useful analytical tools for distinguishing in detail the preferences of actors belonging to a same group or party family. Finally, it can be shown that labour market specific conflict structure not only depends on political constituencies, but is also influenced by the welfare state regime characteristics which determine the distinct problem pressures and pre-structure preferences. In fact, in continental welfare states, political actors throughout the whole political spectrum agree upon policies trying to reduce the insider/outsider divide, as for instance short-time work, whereas especially in Liberal and Nordic countries, a general consensus on the need to reduce public spending appears. Author Flavia Fossati NCCR Democracy University Zurich Affolternstrasse Zurich ffossati@nccr-democracy.uzh.ch Phone: +41 (0)

2 1. Introduction This paper analyzes how the political conflict in labour market policy is structured across six European countries (Denmark, France, Germany, Italy, United Kingdom and Switzerland) by applying an actor-centred policy network approach (see Knoke et al. 1996; Kriesi and Jegen 2001; Häusermann 2006, 2010a and 2010b). This approach allows me to focus on the policy preferences of all major political stakeholders which are active in a specific policy field, giving a more complete picture of the substantive contents of the debate and of policy requests. In fact, it can be expected that in times of increasing economic pressure on the welfare state (Pierson 1996 and 2001), of major reforms of unemployment benefit provision (see for instance Trickey 2001; Trickey and Walker 2001; Graziano 2007; Hinrichs 201), of increasing international competition on the labour market and under the influence of policy suggestions from supra-national institutions, such as the European Union with the Lisbon Agenda and the European Employment Strategy (Lahusen 2009; López-Santana 2006), political conflict changes. It might result that labour market policies are no longer characterized only by the traditional class cleavage or are bound to develop to become mere valence issues, as argued by Kitschelt (1994) 1. Instead, it can be hypothesized that the political conflict over labour market policies undergoes a pluralisation or a re-interpretation as a consequence of new employment related risks arising (Bonoli 2009a; Taylor-Gooby 2005b: 19ff.) and the different policy solutions adopted to face these specific disadvantages. Consequently, it is plausible that the left-right political axis is no longer able to fully capture the differing political preferences of the political stakeholders. Hence, in this paper I try to analyze whether additional conflict dimensions dividing the winners and losers of globalisation arise, and thus change political alliances, leading to a transformation of the policy-specific issue of space configuration. A similar approach on the political space structure 2, with a special focus on coalition building, has already been taken by Häusermann (2006, 2010a and 1 Kitschelt argues that [p]olitical conflict in advanced capitalism will become increasingly a cultural conflict about political parameters constraining citizens choices of life-styles and actual quality of life. Whereas the economy becomes a pure salience issue where parties agree in principle on objectives and even on the basic methods of policy-making but compete on the competence to realize such objectives, communitarian conflicts constitute directional issues on which parties stake out alternative visions of society. (Kitschelt 1994: 298, emphasis added). 2 The analysis of political space has a long tradition in political science. In fact, this is an effective instrument to map the overall salience and the policy preferences changes of both voters and parties (Kitschelt 1994; Kriesi et al and 2008; Laver and Sergenti 2010). In general, the analysis of the political space aims at describing the overall structure of the political space, and at describing the cleavages that shape political competition. This kind of analysis has been undertaken mainly using either party manifestos (Laver and Sergenti 2010) or other documents related to electoral campaigns, such as press releases or newspaper articles when analyzing party preferences; or on survey data, when analyzing the structure of voter preferences (Kriesi et al and 2008). The most recent conceptualisations of political space consider the party and voter preferences to be two-dimensional and to be structured along an economic and a cultural axis (Kitschelt 1994; Kriesi et al. 2006; Kriesi et al. 2008). The approach used in this paper aims at mapping differences not in the general political space but only in a specific policy field in order to discern changes in policy preferences which could be overlooked in a more generalized approach, which includes only party preferences (see Kriesi et al and 2008). 1

3 2010b), who finds that pension reforms in continental welfare states are potentially structured by four different conflict lines 3. Hence similarly to Häusermann (2010b), my approach to political space structure will focus on the question of whether a pluralisation of political cleavages is arising, or whether the political contest is still mainly determined by the traditional de-commodification axis, i.e. the left-right axis and its corresponding state-market orientation (Esping-Andersen 1998). However, the analytical focus of this paper is slightly different in the sense that I analyze policy positions during times of ordinary policy debates and not in the context of specific reform endeavours or in times of electoral campaign, where strategic positioning or bargaining is more likely. As a consequence, mapping political preferences in times of ordinary politics allows the study to capture the political preferences of the relevant actors in a more general and unbiased way than when analyzing these in times of punctuation, such as electoral campaigns or specific reform processes (True et al. 2007). Secondly, a specific interest is laid on detecting whether country-level variables, such as the welfare state regime, the unemployment insurance regimentation or the problem pressure, influence the political conflict in different national settings. In fact, in the design of the present study special attention has been paid to selecting countries belonging to different welfare state regimes and with different political contexts, allowing for an evaluation of the context influence 4. In sum, the present contribution will focus on answering how the changing context conditions of current labour markets influence the structure of policy preferences across Western Europe. In order to address this research question, this paper will first map the labour market policy conflict dimension and hence the actors preferences across six European countries. Such a mapping of preferences allows me to analyze whether beyond to the traditional left-right dimension a second cleavage appears, and if so, how these dimensions vary across countries. The decision to analyze unemployment or more generally labour market policies, was taken because it is the policy field most importantly affected by the consequences of a globalized economy, changing labour market structures and the emergence of new social risks (Bonoli 2005, 2006 and 2009a; Rueda 2005 and 2006; Taylor Gooby 2005). Furthermore, labour market policies are located at the border between the economic and the welfare state domain and thus represent a pivotal institution shaping citizens perspectives and chances. In fact, labour market policies not only regulate areas such as working hours or employment conditions but also pertain to the domain of welfare insurance, i.e. unemployment, pension or sickness benefits, or to areas such as labour market reintroduction, i.e. activation policies. 3 In detail, the four conflict lines concern: insurance, targeting, recalibration and (less importantly) capitalisation. The author then argues that differing preferences on these cleavages permit novel types of cross-class coalitions. 4 Denmark has been chosen as a representative of the Social-democratic welfare state; United-Kingdom as liberal; France, Germany and Switzerland as Continental and finally, Italy as south-european type. 2

4 Notably, many of the major reforms that have taken place in European welfare states over the last few decades pertain at least partially to the labour market domain and are characterized by the introduction of social investment policies, and the concession of out-of work benefits only under the condition of a rigorous commitment to activities related to work re-insertion such as in Germany, Denmark or the United Kingdom (Daguerre 2007; Bonoli 2009a; Trickey 2001; Trickey and Walker 2001; Taylor-Gooby 2005a). Alternatively, as in Italy and to some extent in Germany, liberalizing reforms related, for instance, to loosening hire and fire legislation and the creation of increasing shares of atypical working relations, have been introduced (Molina and Rhodes 2007; Aust and Bönker 2005: 34, 42). In this sense, the labour market policy area is a good case to test whether, apart from the left-right (or the state-market) dimension, a new conflict line has arisen, which deals with different ways to re-commodificate the labour force by either social investment (compare Gingrich and Ansell 2011) or by liberalising measures. The present contribution addresses these questions by means of a Multidimensional scaling analysis (MDS) of the policy preference data of the major political players active in labour market policy in six European countries. Apart from parties, unions and employers organisations, these stakeholders also include government actors and the administration, charities, social movements and research institutes 5. The paper is structured as follows: first, an issue-specific political space concept will be introduced which incorporates the challenges labour market policies face nowadays as a consequence of the economic crisis, globalisation and new social needs. Second, the macro-level variables, which are hypothesized to influence the country-specific differences, are assessed. Fourth, data and methods are presented and are followed by a general, and then country specific discussion of the results. Lastly, the similarities and differences between the political actors positioning and thus the configuration of the political space in the six countries will be discussed comparatively, followed by the conclusions drawn from this. 2. Theory and hypotheses 2.1 The labour market policy conflict axes: left-right versus liberalisation-activation The argument proposed here is that several changes in the national and international labour market context, as well as structural shifts, such as increasing female labour market participation, atypical work-biographies and new forms of precarious employment, give rise to new social needs and hence 5 See tables 2 and 3 in the appendix. 3

5 induce, in addition to the traditional left-right political conflict over de-commodification 6 (Esping- Andersen 1998 and 1996), an additional contest over policies able to meet these new challenges. These policy measures aiming principally at labour force re-commodification, i.e. strengthening the employees reliance on the labour market for income, focus on either work- or education-oriented activation or implement liberal policies, such as deregulating working hours regimentations, to increase opportunities to (re-)enter the labour market and thus reduce public welfare spending or engagement in creating jobs. This more recent conflict dimension is hypothesized to oppose the winners and the losers of globalisation in terms of sectoral employment, skills, and contract type. Hence, depending on the positioning of the political actors constituencies in the labour market, not only different preferences for redistributive policies should result, but also differences with respect to measures addressing the new economic, productive and social challenges. Political conflicts emerge from and are shaped by the political and economic context of the labour market. In particular, four factors may be considered to to affect the changes in political preference structures as a consequence of the ongoing process of globalisation: i) structural changes within the labour force, ii) welfare state reforms, iii) the reduction of the political manoeuvring space with respect to traditional redistributive policies and iv) supra-national policy recommendations. The most recent and pivotal changes that can be expected to influence political actors preferences, arising as a consequence of the ongoing globalisation process of the labour market, which changes the framework conditions of international competition, are firstly, structural shifts in the labour force,. In fact, globalisation goes hand in hand with the demand for an increasingly highly skilled and geographically mobile labour force (Hall and Soskice 2001) and thus changes the social needs that nowadays comprehend, for instance, single parenthood, increasing female labour market participation and the consequent problems of conciliating work and care activities, or the erosion of standard employment relations (Bonoli 2005, 2006 and 2009a; Berton et al. 2009; Schmid 2011; Weishaupt and Lack 2011). Especially these atypical employment relations, which comprehend for instance part-time, fixed-term, or temporary employment, contribute significantly to the alteration of the social risk structure in modern labour markets and thus also the demand for re-commodifying social policy options (Bonoli 2005, 2006 and 2009a; Rueda 2005 and 2006; Häusermann 2010b). Secondly, this internationalisation process also has a major influence on the economic manoeuvring space of (especially left) governments. Currently, mature welfare states spending has been steadily increasing, and this pressure on public debt is exacerbated by declining productivity and rising unemployment figures, which undermine the financing structures of the regimes (Gingrich and Ansell 2011; Iversen and Wren 1998). These sources of pressure, which have been further aggravated by 6 Esping-Anderson (1998) defines de-commodification as social policies, which were adopted by the welfare states to ensure workers pay against traditional risks, i.e. sickness, injury and unemployment. Hence, de-commodifying policies were introduced to decrease the dependence of employees and their families on the (labour) market. 4

6 the recent economic crisis, reduce the possibility for traditional left de-commodifying policies and thus pave the way for new policy solutions (Giddens 2000). Thirdly, and as a consequence of the structural changes, the changing requirements on the labour force and the reduced political feasibility of traditional redistributive policies, over the last few decades, several European countries have made important reforms in welfare state provision to reduce unemployment and welfare state dependency in a more efficient and enduring manner (Iversen and Cusack 2000). Overall these framework conditions make it necessary to look for novel policy solutions such as the Third way approach, developed as an alternative to the traditional left policy propositions, especially in the United States and in the United Kingdom, mainly by means of introducing social investment policies (Giddens 2000; Jensen 2009: 33ff.). Alternatively, especially in continental welfare states, we increasingly observe attempts to introduce liberalising reform or to flexibilise the labour market, as in Italy. Fourthly, the influence of supra-national policy recommendations formulated to meet the new economic environment with which European states are confronted should not be neglected. In fact, the European Commission has launched a reform agenda to increase the competitiveness of European countries (Lisbon Agenda and European Employment Strategy [EES])(Ferrera and Gualmini 2004). This agenda, which relies on the so-called Open Method of Coordination (OMC), proposes several policy goals and indicates possible labour market reform trajectories in the light of the successful Nordic models, such as Sweden. Consequently, the EES influences national debates and opens the policy options spectrum of the member countries for these specific solutions (Lahusen 2009; López-Santana 2006; Blank and Schulze 2011). Also, Bonoli (2010) or López-Santana (2006) suggest that a policy-learning process is developing, which broadens and alters politicians perceptions of policy options, leading to a partial Europeanization of solutions. From these briefly discussed points it becomes evident that the framework conditions for labour market policies and the viable policy options have changed. New demands have arisen as a consequence of demographic and structural shifts, public budgets and thus governments manoeuvring space are constrained, and supra-national institutions are having ever more influence on national policy options. In such a context the main conflict that seems to arise deals with the different policy options in order to re-commodificate the labour force, as opposed to traditional redistributive policies, which in an era of permanent austerity (Pierson 1996 and 2001) are no longer viable measures. According to the discussion in the literature, there seem to be two alternatives: either implement activation or liberalisation policies. Both aim at ensuring competitivity, keeping unemployment and public deficit low, and avoiding long-term dependency. Accordingly, Gringrich and Ansell (2011) identify social investment strategies as complementary to traditional measures, or in other words as being a cure against neoliberal policies and an alternative to passive traditional social consumption policies (Bonoli 2009b; Jensen 2009: 27-28). In 5

7 fact, a consensus seems to have developed that activation policies are a very flexible policy instrument able to address very different (structural) problems encountered by different welfare state regimes (Trickey 2001: 250; Daguerre 2007). On the other hand, different approaches to liberalisation can be encountered. Hence, it can be predicted that a further conflict dimension, over different, either liberalising or activation solutions, to adapt national labour markets to the changes induced by globalisation might arise. To illustrate this additional cleavage in more detail it is useful to also consider the preferences and the risk profiles of their constituencies. Addressing these preferences, firstly we encounter the traditional conflict line, which splits labour and capital in terms of more/less state intervention and redistribution preferences (Esping-Anderson 1998). This dimension can be hypothesized to be still relevant especially for issues related to policy decisions such as eligibility conditions, targeting and duration of unemployment benefits and other (traditional) passive welfare state policies. Here we expect actors representing employees, i.e. labour force, to prefer state intervention and high welfare benefits, or according to Esping-Andersen s theory and more generally the power-resource approach 7, high decommodification levels (Esping- Andersen 1998). Instead, firms and employers organizations can be expected to propose foremost market friendly and deregulating policies. However, as argued by Oesch (2006), and as described above, the labour force and more generally social structures have undergone a pluralisation of preferences with respect to the traditional cleavages between the old class categories, and as a consequence of structural shifts induced by globalisation. In his book Redrawing the class map, the Oesch argues that these preference shifts depend on i) the work-logic and ii) the skill level of the employees (Kriesi ; Kitschelt and Rehm 2005; Häusermann 2010b: 9, 57). Already, Milner and Keohane (1996: 15) made a similar point indicating that the globalisation process changes the preferences of the labour force, dividing them into a group of winners and a group of losers of the internationalisation phenomena. The same conclusion has been reached also by Kriesi et al. (2006, 2008), who assessed globalisation as representing a critical juncture, which leads to changes in political preferences and thus to a reinterpretation of the cultural conflict dimensions of political competition 9. Thus, in the literature 7 The power-resource approach argues that the major player in the development of welfare states has been the political left in a coalition with unions (see Esping-Andersen 1998; Huber and Stephens 2001). 8 Oesch s (2006) theory is based on a previous conceptualization of the split in the middle class by Kriesi (1998). The latter identifies a split in cultural preferences within the middle class depending on their control of work, which results in a distinction between managers and sociocultural professionals. 9 They discussed these changes foremost related to the cultural conflict dimension, which they determined to result from a transformation of the libertarian axis (Kitschelt 1994) in terms of integration or demarcation with respect to immigration and EU issues. In fact, Kriesi et al. (2006 and 2008) and Bornschier (2010) assessed the new political cleavage as being mobilized foremost by the radical right parties and argue that globalisation affects the cultural perception of natives and increasingly divides those who profit from globalisation from those who lose. This distinction is driven foremost by cultural and identity-related variables such as preferences on immigration and Europeanization. (Kriesi et al and 2008; Häusermann 2010b). 6

8 there seems to be a broad consensus on the causes changing political constituencies, however, rather different variables are used to qualitatively identify the dynamic according to which these changes occur. In this contribution I would like to rely on the stance taken by Kriesi (1998) and Kriesi et al. (2006 and 2008: 4-6), who argue that the winners of globalisation are those who profit from increasing market liberalisation because they still retain good exit options, i.e. are geographically mobile and skilled. As a consequence, it can be deduced that losers profit more from social investment, which enable them to acquire new skills to compete in an international economic context 10. Accordingly, the variables explaining the preferences of the constituency can be supposed to depend on the specific labour market profile. In detail, the exit options in the labour market can be explained in terms of i) skill level 11 (Rueda 2005 and 2006), ii) sectoral affiliation (Manow 2001; Kriesi et al. 2008: 4-6) and iii) occupation/ contract type. In sum, two distinct conflict axes in the modern labour market policy should result: firstly, the traditional class cleavage between labour and capital (de-commodification axis). Secondly, as a result of globalisation we have a conflict line dividing the labour force in winners and losers in terms of what policies they need to ensure their jobs in a globalised and competitive market. Let me note at this point that if we were to rely only on the left-right dimension to assess political actors preferences, we would simply observe unions, charities and left-wing parties opposing the employers organisations and right-wing parties. Instead, introducing an additional conflict line, different policy positions within these separate groups can be identified and analyzed, and this not only in terms of being more left or more right. < Table 1> As shown in Table 1, referring to the first conflict dimension, for employees the priority is to retain access to the labour market or to the necessary benefit to cover lacking income from traditional risks, such as sickness, injury or unemployment (Esping-Andersen 1998). In fact, the labour force has largely been assessed to be on the losers side in contrast to capital which is more mobile, (Iversen and Wren 1998) and thus the labour force greatly needs de-commodifying measures. On the contrary, for the firms and the capital the main focus is on increasing their competitivity and containing both public spending and the costs of the labour force. As a result, they plead for marginal welfare benefits and less state intervention in the market. 10 A similar argument has been made by Wueest and Fossati (2011). In a slightly different context the authors identified the political preferences of constituencies, as being relevant for which kind of problem frames are raised in labour market policy by specific political actors. In this case the differentiation between left and right, as well as winners and losers, was applied in order to explain whether social protection, social justice, prosperity, freedom, intervention or national protection frames were preferred and thus raised more frequently in national debates. 11 Skill level also has an impact on preferences for specific activation, in fact, depending on their work- or human capital orientation; different constituencies will prefer rather the one or the other, as will be shown below. 7

9 However, considering accurately the preferences of labour and capital, i.e. left and right, in the current globalized context, we find divergences in their preferences in the second dimension. Firstly, workers with high educational qualifications have an advantage with respect to those possessing low or obsolete skills. In globalized employment settings, highly-skilled people can consequently be hypothesized to prefer liberalizing measures as compared to financing-intensive activation measures. Furthermore, activation measures are often criticized as applying rather unspecific training, such as for instance job application training, which for high skilled employees are generally obsolete (compare Trickey 2001: 268ff.; Daguerre 2007). Thus by tendency, skilled workers retain several exit options and can be expected to be favourable to liberalising policies, whereas unskilled personnel fear unemployment and thus prefer social investment policies. The same mechanism should also be applicable to the employers organisations, which represent firms and sectors needing mainly highly skilled personnel. These organisations should also strongly favour liberalizing measures rather than activation policies which are deemed to be ineffective and costly. Instead, firms relying on low-skilled workers should be less decidedly against activation measures because their work-force could profit from further training. Moreover, with respect to skills, an additional differentiation becomes relevant. In fact, activation measures can be differentiated with respect to their main focus, which either stresses work or education activities (Kriesi et al. 2011a; Daguerre 2007). In continental welfare states in particular, the unemployed, members of the female work-force, or school leavers wishing to enter the labour market, might predominantly favour activation measures focusing on work-first measures such as internships and temporary or social contracts, to gain access to labour market in the first place. Instead, employees, especially those with low or obsolete skills might favour education and training activities, which can also be pursued while working. Secondly, in general there are sectors which profit from increasing liberalisation, such as the financial sector or export-oriented firms, whereas farmers and small business in particular, can be expected to prefer activation policies when it comes to re-commodifying the labour force. Divergent preferences can be also expected to characterize the public versus the private sector. While the public sector is generally sheltered from international competition and employees and their representatives should be less prone to favour liberalizing measures, in the private and profit-oriented sectors instead cost reduction and liberalisation should be favoured more strongly. Finally, while people working in typical, i.e. fulltime and long-term work relations, do not fear increasing international competition and liberalisation, people working on atypical contracts prefer social investment programmes, to retain or re-gain access to the labour market in the case of job loss due to restructuration or difficulties in their firm. Also with respect to contract type, the distinction between activation preferences once again becomes relevant. In fact, atypical contractees in 8

10 particular might prefer work-oriented activation options, while those in stable, long-term workingrelations should prefer further training Hypotheses on the space configuration and political actors preferences Based on the theoretical discussion, a set of partially contrasting hypotheses can be formulated. The first hypothesis (H1) is that the political preference configuration of political constituencies is two-dimensional. Along the traditional left-right 12 axis the conflict which concerns foremost the regulation of the role and the engagement of the state and the scope of redistribution policies (universal or targeted) develops (Esping-Andersen 1998). On the vertical axis, which concerns re-commodification strategies, instead a dualism between activation policies and the liberalization measures can be identified. In other words, an outdifferenciation of the economic conflict dimensions, as a reaction to the complexity of new social risks in modern welfare states should be observable. < Figure 1> By combining the two axes, the one representing the traditional de-commodifying policies and the other the re-commodifying policy options, four ideal-typical preferences can be derived: i) Flexicurity, ii) Social, iii) Liberal or iv) Deservingness policy alternatives. The Flexicurity approach is characterized by left-oriented passive measures entailing high decommodification, in combination with liberalising re-commodificaiton policies (such as hire and fire policies, etc.), and corresponds to the preferences of left winners. The Social approach, which corresponds to the preferences of left losers, is characterized by heavy de- and re-commodification policies, with a special focus on work-reinsertion policies. On the right political side, actors representing the right losers are placed within the Deservingness quadrant. This approach includes a rather strong orientation towards activation measures but combining these with marginal benefits and with strong deservingness criteria, for instance showing adequate efforts in job-hunting or attending periodical colloquia at jobcentres. This model thus corresponds to either work-first or workfare policy solutions. Finally, in the Liberal model we have traditionally right and marginal benefit structures (needs-tested benefits) combined with strongly liberalizing policies. This combination should suit especially high-skilled internationally mobile workforce members, i.e. right winners, who do not fear unemployment. In this approach the role of the state is constrained to providing help to self-help. 12 I hypothesize that the left-right dualism is not just overcome as sometimes suggested by theoreticians of the third way (Giddens 2000) but is still crucial in economic politics; or declined to a mere valence issue, as argued by Kitschelt (1994: 258). 9

11 Alternatively, it could be expected that the additional conflict over issues concerning recommodification strategies is, at least partially, integrated or embedded (Kriesi et al and 2008) in the traditional left-right dimension (H2). A similar approach has been proposed by Bonoli (2011) to describe and analyze active labour market policies. The author identifies left and activation policies as being structured on a continuum moving from protection to investment and finally to recommodification. The implication would be that the political space is not (clearly) two-dimensional but that instead a partial re-interpretation of the traditional left-right axis in the labour market domain occurs. Thus, this second correlated axis, would add explanatory power foremost in terms of highlighting additional relevant issues in this policy domain, and in differentiating more clearly the stance political actors take on these within actor groups or party families. In this case a left-activation oriented position could oppose a right-liberalizing pole, as shown in Figure 2. < Figure 2> The next hypothesis (H3) concerns the preference structure of political actors constituencies and their precise placement in the political space, as summarized in Figure 3 for the case the first hypothesis (H1) is underpinned. <Figure 3> Unions, which generally represent foremost the interests of the insiders of the labour market (Rueda 2005 and 2006), can be expected to be placed on the left political spectrum and keep favouring heavily de-commodifiyng, traditional left policies. However, for what concerns the second axis (re-commodification) arisen as an adaption to globalisation related challenges, they can be supposed to have different preferences depending on the characteristics of the labour force they represent. Unions representing highly skilled salaried members operating in internationalized economic sectors will favour the Flexicurity model and agree on further liberalisation of the labour market, because, as Kriesi (1998) argues they do not fear international competition and they retain several exit options. Furthermore, when asked for the specific preferences on activation policies, unions of high-skilled personnel can be expected to prefer investment in human capital rather than work-related activities. On the contrary, unions which represent foremost the interests of the traditional left clientele, the industry blue collar workers or of the sectors protected from globalisation (farming, small business) should be strongly in favour of traditional policies guaranteeing their income and helping them to avoid unemployment or precarisation. Moreover, on the re-commodification axis, blue-collar workers unions will prefer, for instance, measures such as short-time work or subsidized jobs, as compared to further training (human capital approach), or even to liberalizing measures situated on 10

12 the other extreme of the conflict line. Such unions will also prefer less individualized activation solutions than unions representing high-skilled for whom standard activation courses might be obsolete, and focus on social network preservation (Daguerre 2007) and labour market access reintroduction by means of work-related strategies (as work-first approaches). Similarly, typical losers, as for instance people on atypical or precarious contracts, such as short-term employment, part-time employment or subordinated employment, or the workforce with obsolete/low skills and the female workforce 13 will prefer the Social approach, and with respect to activation particularly those measures with work-first or work-related activities. Unions of public employees will instead have a moderate position on both the dimensions since their members are challenged neither by globalisation nor by traditional labour market risks, due to their working on long-term and stable contracts. Consequently, they can be expected to be placed between the Social and the Flexicurity approach. Concerning specifically the preferences for activation policies, these can be supposed to prefer human capital stimulation as compared to workfirst or workfare approaches, whose primary goal is to reduce welfare state dependency. Charities and organisations representing the unemployed are the main representatives of the hardcore of losers, who for instance do not find adequate representation in the party system because left parties tend to focus on the winners rather than the losers preferences (Krieis et al and 2008). These can be supposed to favour left policies and activating measures, and mainly work-first strategies. In fact, these represent individuals who have lost access to labour market policy, and thus are primarily interested in being re-introduced into the social network and into the labour market in order to avoid welfare state dependence. These, according to Rueda (2005), typical outsiders, have different policy preferences than insiders, who are much less concerned about liberalising reforms. This is consistent with the literature that supposes parties to take the part of the winners in economic terms while neglecting the losers such as the unemployed, workers in atypical situations etc. (Rueda 2006: 386). In fact, such people are being mobilized by new social movements such as the precari-movement in Italy but have little impact on the party system because they lack political leverage and access (Baglioni et al. 2008). Employer s organisations are hypothesized to favour right policies, but depending on the sectoral or skill composition of their workforce, either the Liberal, or the Deservingness model. Administration and government can be expected to act according to their ideological tradition (these will discussed in more detail below). In Liberal countries (UK) a Liberal model orientation can be 13 As we will see below, this will especially be the case in Continental and Southern welfare sates 11

13 hypothesized, in Nordic countries (DK) a Flexicurity approach and in Continental welfare states, such as France and Italy, a Social model should prevail (Daguerre 2007). For Switzerland the situation is more complex, since this welfare state represents a mixed type somewhere between the continental and the liberal model and thus the positioning of the administration and government might be more ambivalent. Finally, turning to political parties, several differences can be identified. On the left we find broadly speaking three types of parties: traditional and moderate social-democratic parties, who represent mainly the left-winners; furthermore, there are new more radical left parties mobilizing the losers of globalisation and finally, the greens. As Rueda (2006: 386) suggests the traditional social democratic parties are no longer concentrating on labour market losers but by tendency strive to the political middle and by consequence have renounced representing their traditional blue-collar clientele on the far left (Kitschelt 1994) but have grown their influence in specific winner categories, such as the high-skilled sociocultural professionals (Kriesi 1998). By consequence, they can also be expected to favour liberal recommodification measures (Flexicurity), since their members retain the necessary exit options and are not challenged by increasing globalisation. The same should apply also to Green parties, which can be expected to favour the Flexicurity approach. To some extent the traditional social-democratic clientele and the new group of individuals at social risk, such as unemployed and atypical workers, are instead mobilized by populist left parties (such as Die Linke in Germany) which prefer the Social approach. However, generally and as discussed above, the representation of low-skilled, atypical employees working in highly challenged sectors will be monopolized by actors with marginal political influence such as specific unions, new social movements and charities. Moderate and populist right parties can be expected to prefer the Deservingness model which combines marginal passive benefits with work-first or workfarist activation solutions to recommodify the labour force. Instead, liberal parties favour the liberal model combining liberal recommodification measures and marginal passive benefits. After having discussed in detail the theoretical placement of political actors (depending on the preferences of their constituencies), in the case where the political space is revealed as being clearly two-dimensional, attention should be briefly paid to the case in which instead H2 is more appropriate. In this case we can expect actors from Q1 to migrate to Q3 and those of Q4 to shift to Q2, converging on a diagonal preference formation, as illustrated below. <Figure 4> 12

14 In this case, even though the hypothesized second dimension results as not being entirely independent from the traditional class-conflict axis, it would still be useful in terms of distinguishing the political positioning within particular actor families or groups, which otherwise would be far less precisely identifiable. As an example, with respect to unions, even in the case where they were placed in Q3, it should still be possible to differentiate between those representing high-skilled and low-skilled personnel in terms of preferences for liberalisation, respectively, activation measures. After having developed a general framework explaining the positioning of political actors in the labour market policy space, in the following section the theoretical ground will be laid to develop hypotheses dealing with the influence of macro level variables on the development of the country specific conflict lines. 2.2 Theory: the influence of welfare state regimes on political conflict structure in labour market policy In the following the theoretical framework assessing the influence of the welfare state regime type on the country specific labour market structure will be briefly described 14. The insights, which are deducible from the conceptual framework developed by Thelen (2011) can be transposed also to national specific conflict line development. In fact, in a different context, the author argues that modern democracies, which were found to differentiate into two categories of Liberal Market Economies (LMEs) and Coordinated Market Economies (CMEs) according to the Varieties of Capitalism theory by Hall and Soskice (2001), are undergoing different types of adaptations to globalisation. Thelen (2011) argues that these specific reform trajectories which are characterized by changes in commitment to social equality and/ or coordination depends on the regime characteristics of the countries 15. Similarly, I take the stance that the institutional framework influences the preference structures across different countries (compare also Aust and Bönker 2005: 48-49), leading to very different policy reactions or as Thelen (2011: 22) put it reform trajectories to respond to the pressure of new social needs arising in a context of increasingly globalized labour markets. 14 Due to space constrains this theoretical elaboration will concentrate on the aspects most relevant in the context of this paper, and thus especially consider only those hypotheses whose arguments could be operationalized with the present dataset. Therefore, overall the account of the influence of welfare state regimes may at times appear incomplete and lack a diversified discussion. 15 She argues that depending on the institutional framework of the LME or CME countries different approaches towards liberalizing policy measures are taken. Nationally coordinated systems, such as the Scandinavian countries implement liberalizing reforms which are accompanied by the concept of Embedded flexibilisation, i.e. flexibilization measures are introduced but the new needs arising are absorbed by high de-commodification measures. The Industry or sectorally coordinated systems, which are for instance found in Germany or France, follow a dualisation path. And finally the LME countries, such as the United Kingdom follow a deregulation path, increasing flexibilisation and reducing benefits overall. 13

15 Overall, labour market policies can be expected to be heavily influenced by the welfare state regime 16 present in a country. In fact, depending on their difference with respect to decommodification practices 17 they have different effects on social stratification, as well as on the specific welfare mix (Esping-Andersen 1998) and consequently, are able to explain the differences in employment structure inherent challenges and thus determine, at least to some extent, the different sources of problem pressures. Hence, it can be expected that depending on the welfare state system, political actors tend to agree on specific policy measures as being necessary/ inevitable in dealing with regime-specific challenges, or on the other hand they might cohesively disagree with calls to implement or tighten measures which are deemed to already be implemented sufficiently. Hence, using Table 2 below, which summarizes the main characteristics of the different welfare state regimes, I will try to assess which kind of policies differ with respect to general (dis)agreement amongst the political elites of the different countries studied with respect to specific labour market policies. <Table 2> In Nordic welfare state regimes benefits are very universal and committed to reducing social stratification by means of redistribution (Esping-Andersen 1998). Generally, these countries managed the transition to a service economy very successfully, especially because of the early commitment to social investment policies, being thus able for instance to integrate the female labour force into the service sector, and thus increasing the overall employment participation share. In fact, these welfare regimes are among those with the highest spending on childcare to allow mothers to remain in the labour market process. Moreover, by the early introduction of social investment policies, with a special focus also on educational policies, these regimes had already started decades ago with a generalized up-skilling of the labour force, which raised its international competitivity (Taylor-Gooby 2005a; Taylor-Gooby 2005b; Timonen 2005). Also the introduction of part-time work and the liberal hiring and firing practices do not pose social problems because embedded within a well functioning security net. Thus, in the area of labour market supply, i.e. (un)employment problems related to new social needs, no major pressure can be expected in these regimes. Whereas questions inherent to the labour force structure are mostly unproblematic as a consequence of the cost-intensive social policies and especially in the current economic austerity context, there is concern about the increasingly high public deficit and whether tax and employment rates are high enough to finance the system (Timonen 2005). These economic considerations have 16 In the literature between three or four welfare state models can be distinguished depending on whether the south- European type is considered as an additional type next to the continental (Esping-Andersen 1998; Castles 1995; Ferrera 1996). 17 More specifically these concern eligibility rules, level, duration and generosity of the benefits. 14

16 already pushed some countries to drastically reduce benefits, as for instance the recent Danish unemployment reform which reduced the duration of benefits from four to two years. In sum, in social-democratic welfare state regimes, labour market problems should not be pressing since the workforce is highly skilled and benefits are generous. Instead, major problems can be expected to result from high public debt and spending on social policy (Taylor-Gooby 2005; Timonen 2005). In sum, the hypothesis H4a can be derived, that political actors in general might be concerned about public finances, but still strongly favour activation policies, which so far have proved very successful. H4a: In social-democratic welfare states, deficit reduction and activating policies can be expected to be issues on which most political actors agree. The second type of welfare state describes the liberal countries, as for instance the United Kingdom. There, according to the underlying liberalisation and market-reliance philosophy, benefits for the unemployed are kept low and the eligibility rules are strongly based on deservingness criteria. One result of the transformation to a service economy is the increase in the share of low-skilled jobs, or as Esping-Andersen called them junk-jobs, but also an up-skilling at the very other end of the social structure (Esping-Andersen 1998: ; Bonoli 2009: 37). These low-qualified jobs, which often also rely on atypical contract types, can be expected to increase the marginalization of already underprivileged groups such as unqualified immigrants or women (Taylor-Gooby and Larsen 2005). Hence, in liberal welfare states, low educational qualifications are generally associated with a high risk for unemployment and low wage (Taylor-Gooby and Larsen 2005: 58). As a consequence, for instance the United Kingdom developed welfare instruments, the so-called New Deals, which aimed at targeting those groups in particular which were at risk of unemployment. These include foremost young people, lone parents and low-skilled employees. The strategy adopted to stimulate self-reliance and prevent welfare-state dependency was to introduce work-first, or even workfare policies, whose motto is make work pay (Taylor-Gooby and Larsen 2005; Trickey 2001; Trickey and Walker 2001). These strongly work-related activation strategies are also supposed to reduce the amount of public spending and especially welfare state dependency, as expressed very clearly by the English government in several official documents (DWP 2010a and 2010b; HM Treasury 2010). Hence, even more so in the current era of austerity, we can expect liberal countries to additionally reduce public spending and contemporaneously restrict benefit eligibility criteria to reduce public deficit. The governments framing, will make reference to the philosophical commitment to (market-) liberalism and individual self-reliance. 15