How regimes shape preferences. A study of political actors’ labour market policy preferences in flexicurity and dualizing countries

How regimes shape preferences. A study of political actors’ labour market policy preferences in... Abstract Political actors do not operate in a vacuum; rather, it is safe to assume that their preferences are influenced by the institutional context in which they operate. By means of novel interview data, which was collected in flexicurity countries, i.e. Denmark and Switzerland, and in dualizing countries, i.e. France, Germany and Italy, I investigate the preferences of parties, unions, state bodies, employers and social movement organizations towards traditional and activation strategies. I find that the institutional context indeed shapes preferences. The results reveal, for instance, that state bodies reject increasing activation efforts in flexicurity countries but support it in dualizing countries. Moreover, in line with previous research, social democratic parties are found to cater to the interests of insiders by endorsing the expansion of traditional measures in dualizing countries, while focusing on outsiders’ interests by preferring the expansion of activation in flexicurity countries. 1. Introduction Political actors do not operate in a vacuum; rather, it is safe to assume that their policy preferences are shaped by the institutional context in which they operate. However, much of the comparative political economy literature continues to assume constant actor preferences regardless of the context. This article takes a different perspective, and it shows how institutional factors shape the positions that political actors take on labour market policy. Specifically, I analyse actors’ preferences in relation to the main strategies adopted to fight unemployment and contrast traditional versus activation approaches. The objective of the analysis is to show how these preferences are affected by the institutional setup, namely, the labour market regime. Preferences on traditional strategies concern support for the expansion/retrenchment of measures such as passive unemployment benefits and employment protection legislation (EPL). Preferences on activation refer to—support for the expansion/retrenchment of different active labour market policies (ALMPs). These include ‘enabling’ and ‘demanding’ measures, such as training and counselling on the one hand and negative incentives and sanctions on the other (Eichhorst and Konle-Seidl, 2008; Bonoli 2013; Fossati, 2017). While accounting for both types of preferences, I concentrate on attitudes towards ALMPs because of the importance these measures gained over the past decades. In fact, activation seems to have affected the entire welfare state, transforming the scope of its provision from a ‘securing’ to an ‘enabling’ institution (Clasen and Clegg, 2011; Bonoli, 2013). In the light of this massive welfare reorientation, it is surprising that little effort has been made to assess political actors’ activation preferences in a systematic and direct way, as this could improve the understanding of current reform processes (Rueda, 2007; Nelson 2013; Tepe and Vanhuysse, 2013; Vlandas, 2013). Specifically, I argue that the institutional context, i.e. whether a country has developed a flexicurity or dualizing labour market regime, influences actors’ preferences because some positions become more/less legitimate, functional and salient depending on this institution. For instance, in regimes with generous benefits, it is dysfunctional for social democratic parties to advocate increases in replacement rates; rather, they should focus on more salient issues such as the expansion of ALMPs to address increasing unemployment. Labour market institutions also influence preferences because the policies implemented are the reference points for political actors and shape their understandings of what constitutes a particular strategy (Immergut, 1998; Larsen, 2007). Put differently, when political actors in a dualizing country refer to activation, they have in mind a different combination of policies than do actors in flexicurity countries. Finally, institutions influence outcomes, which in turn affect preferences. This is especially the case for the quality of the jobs that jobseekers enrolled in ALMPs are likely to obtain. If these jobs are bound to be predominantly precarious, as is the case in dualizing countries, social democrats and unions might reject the expansion of measures that swiftly push workers back into uncertainty. In contrast, if the jobs are seen as adequate, pro-labour actors might be less opposed to such a policy. Concisely, I maintain that to better understand political actors’ preferences, we should analyse them as the product of a specific institutional constellation. My contribution to the literature is three-fold. First, I expand the analysis beyond the ‘usual suspects’ and study all political actors’ preferences and, thus, the political conflict structure that characterizes the labour market policy domain (unlike, e.g. Rueda, 2007; Tepe and Vanhuysse 2013). Thereby, the political conflict axes are identified by aggregating actors’ preferences and the saliency of specific policies; then, these axes are used to display the political space in which the actors are located (Kriesi et al., 2008; Häusermann, 2010). This procedure allows the mapping of actors’ preferences and determining their potential1 for value-based coalitions in two flexicurity countries (Denmark and Switzerland) and three dualizing countries (France, Germany and Italy). Secondly, I demonstrate the importance of context-specific analyses by showing that actors belonging to the same party family or actor group may indeed have different preferences in different regimes. Thirdly, I innovate by using a more direct measure of preferences consisting of interview data (Knoke et al., 1996) rather than inference based on government spending patterns (e.g. Bonoli, 2013; Tepe and Vanhuysse, 2013; Vlandas, 2013) and/or the preferences of the actors’ constituencies (Rueda, 2007). These data have the advantage of being less likely to be distorted because they capture value-based preferences as net of (strategic) politics and package deals (Immergut, 1998; Sabatier and Jenkins-Smith, 1999). The article proceeds as follows: in the theoretical section, I set out and describe the labour market policies and institutions to which political actors refer when forming their preferences. In the empirical analysis, I construct what amounts empirically to a two-dimensional labour market space onto which I map the political actors’ positions.2 Then, I analyse the potential for coalitions based on the similarity of actors’ preferences. The last section summarizes the findings and explores avenues for further research. 2. Preferences and conflict structure of traditional and activation policies Traditional measures to fight unemployment include passive benefits and EPL. The positions of political actors with respect to these measures are structured along the labour–capital antagonism, and they can be synthesized as an axis with one side concerning preferences for generous policies and the other side concerning support for lean policies (Korpi, 1983; Esping-Andersen, 1990; Kitschelt, 1994). Mainly, social democratic parties and unions have proposed policies to reduce social inequality by insuring blue-collar workers against traditional industrial risks (Esping-Andersen, 1990). Conversely, political actors on the right propose liberal market solutions to reduce state intervention and constrain universalistic and redistributive welfare state spending and labour market regulation. Although this conflict still structures the labour market domain, it is no longer the only one that does so—as has been shown for other policy fields (Kriesi et al., 2008; Häusermann, 2010; Fossati and Häusermann, 2014). Recently, traditional policies have proven to offer suboptimal protection for an increasing share of outsiders, including atypical and unemployed workers with low or obsolete skills (Bonoli, 2005; Rueda, 2007). Moreover, the adverse economic conditions, including the lower growth levels that are typical of post-industrial economies, and the pressure placed on welfare states by sociodemographic changes preclude the possibility of meeting outsiders’ needs by simply increasing decommodification efforts (Pierson, 1998, 2001). Consequently, governments had to seek alternative approaches, leading to what is known as an activation turn. In fact, over the last three decades, following a supranational consensus guided by organizations such as the European Union, with its European Employment Strategy (EES), a range of policies that actively promote labour market reintegration have been introduced across Europe and beyond (Torfing, 1999; Gilbert, 2002; Bonoli and Natali, 2012; Bonoli, 2013). Bonoli (2013) argues that ALMPs can be distinguished based on the degree to which they invest in human capital and the degree to which they ‘push’ the unemployed swiftly back into employment by means of conditionality (pro-market orientation) (cf. Eichhorst and Konle-Seidl, 2008; Fossati, 2017). In this regard, he distinguishes four types of ALMPs: (a) occupational policies, which are characterized by a low human capital orientation and low pro-market orientation (public jobs); (b) incentive reinforcement measures, which have a strong pro-market orientation but no human capital component (sanctioning and incentives); (c) upskilling instruments, which have a strong human capital orientation and a strong pro-employment orientation (job-related training); and (d) employment assistance measures, which are characterized by a medium pro-market component and a weak human capital component (placement services, counselling). Generally, a combination of these strategies is applied. Empirically, however, some countries focus more on human capital investment (traditionally, social democratic countries), while others stress swift re-employment (traditionally, liberal countries). The interesting question is thus how this institutional variation influences political actors’ preferences. 3. How labour market regimes affect actors’ preferences I argue that the institutional legacies that place countries within a specific labour market regime differentially influence actors’ (activation) preferences3; thereby, three mechanisms apply. First, as argued by Steinmo (1989, p. 502), the ‘… institutional framework provides the context (emphasis added) in which groups and individuals interpret their self-interest and thereby define their policy preferences’. Thus, preferences (and values) have only context-specific substantial meanings (Steinmo, 1989, p. 502; Immergut, 1998). In other words, preferences for/against a specific measure are reactions to policies that have already been implemented. For instance, liberal parties should be more supportive of EPL liberalization in dualized countries than in flexicurity countries, where regulations are already very lenient. However, if we ignored institutional differences and considered the extent to which liberal parties agree with additional EPL liberalization and found that they ‘strongly agree’ in dualizing countries but just ‘agree’ in flexicurity countries, we could wrongly assume that these parties take a more moderate stance in flexicurity countries. In a similar vein, the labour market characteristics of a regime (de-)legitimize or make specific preferences more (less) salient. In fact, political actors can be expected to focus on social problems or on policies that need reform rather than engaging in issues that are already sufficiently addressed, such as labour market liberalization in flexicurity or insider protection in dualizing countries. Secondly, institutions provide anchor points and frames of reference for political actors’ policy definitions. They interpret a strategy based on what is implemented in their country and ponder whether adding a particular policy matches, reinforces or contradicts the pre-existing framework. For instance, rational employer organizations should have different ALMP preferences in flexicurity countries, where ALMPs provide a larger pool of well-qualified candidates, whereas in dualizing countries, non-upskilled unemployed are merely pushed back into the job market. Thereby, the institutional effect unfolds also based on the specific framing of policies, i.e. the way in which problems and solutions are assessed and communicated (Entman, 1993; Larsen, 2007; see Wüest and Fossati (2014) for a framing analysis of labour market discourse). In liberal countries, unemployment is generally framed as a consequence of behavioural shortcomings (Daguerre, 2007). In such a context, it is plausibly more difficult to support human capital investments than negative incentives (e.g. conditionality). In traditionally social democratic countries, unemployment is often considered to have structural causes, such as a mismatch between demand and supply, which easily explains skill-based policy preferences and responses (Daguerre, 2007). The unstable working conditions and the high unemployment risk of outsiders in dualizing countries, instead are likely to push political actors to defend activation measures catering insiders rather than investing in outsider’s human capital that might be lost again in the event of a future layoff. Finally, the institutionally pre-structured social outcomes likely affect actors’ preferences (Steinmo, 1989). In the labour market domain, the quality of jobs that unemployed obtain matters. If a well-intentioned policy produces suboptimal outcomes and results in precarious jobs, political actors on the left may distance themselves from such measures even if these would correspond to their ideological roots. Concisely, following Lowi’s (1972) logic, I propose that policies shape preferences and that— although this is beyond the scope of the present article—preferences matter because they are at the core of politics. 4. The characteristics of flexicurity and dualizing labour market regimes To examine the effect of the institutional context on political actors’ preferences, I focus on Denmark and Switzerland as representative flexicurity4 countries and France, Germany and Italy as representatives dualizing countries.5 As suggested by Thelen (2012), this theoretical distinction is more precise and less static than the welfare regimes or varieties of capitalism approaches (Esping-Andersen, 1990; Hall and Soskice, 2001). In fact, it accounts for reactions to the liberalization pressures by distinguishing among countries that try to preserve social equality and those that respond with dualization or even deregulation (e.g. the UK, the USA). I carefully chose the cases in order to maximize comparability within a regime by grouping countries that implemented similar policies (Table 1). Table 1. Active and passive (traditional) labour market policy efforts   ALMPs   Traditional labour market policies     Spending on training as a percentage of GDPa  Spending on ALMPs as a percentage of GDP per percentage of unemploymentb  ALMP spending per unemployed personc  Automatic ALMP spending adjustmentd  Net unemployment benefit replacement rate for a single persone  Net unemployment benefit replacement rate for one- earner couple with two childrene  Flexicurity              Switzerland  0.20  0.18  2.3  Yes  0.71  0.83  Denmark  0.66  0.33  4.2  Yes  0.55  0.62  Dualizing              Germany  0.27  0.09  1.8  No  0.60  0.72  France  0.37  0.11  1.5  No  0.69  0.70  Italy  0.15  0.06  0.5  No  0.63  0.73    ALMPs   Traditional labour market policies     Spending on training as a percentage of GDPa  Spending on ALMPs as a percentage of GDP per percentage of unemploymentb  ALMP spending per unemployed personc  Automatic ALMP spending adjustmentd  Net unemployment benefit replacement rate for a single persone  Net unemployment benefit replacement rate for one- earner couple with two childrene  Flexicurity              Switzerland  0.20  0.18  2.3  Yes  0.71  0.83  Denmark  0.66  0.33  4.2  Yes  0.55  0.62  Dualizing              Germany  0.27  0.09  1.8  No  0.60  0.72  France  0.37  0.11  1.5  No  0.69  0.70  Italy  0.15  0.06  0.5  No  0.63  0.73  Sources: a OECD (2016a), figures for 2010; b Bonoli (2013, p. 30), figures for 2007; c Tepe and Vanhuysse (2013); d Duell et al. (2010); e van Vliet and Caminada (2012), figures for 2009. Table 1. Active and passive (traditional) labour market policy efforts   ALMPs   Traditional labour market policies     Spending on training as a percentage of GDPa  Spending on ALMPs as a percentage of GDP per percentage of unemploymentb  ALMP spending per unemployed personc  Automatic ALMP spending adjustmentd  Net unemployment benefit replacement rate for a single persone  Net unemployment benefit replacement rate for one- earner couple with two childrene  Flexicurity              Switzerland  0.20  0.18  2.3  Yes  0.71  0.83  Denmark  0.66  0.33  4.2  Yes  0.55  0.62  Dualizing              Germany  0.27  0.09  1.8  No  0.60  0.72  France  0.37  0.11  1.5  No  0.69  0.70  Italy  0.15  0.06  0.5  No  0.63  0.73    ALMPs   Traditional labour market policies     Spending on training as a percentage of GDPa  Spending on ALMPs as a percentage of GDP per percentage of unemploymentb  ALMP spending per unemployed personc  Automatic ALMP spending adjustmentd  Net unemployment benefit replacement rate for a single persone  Net unemployment benefit replacement rate for one- earner couple with two childrene  Flexicurity              Switzerland  0.20  0.18  2.3  Yes  0.71  0.83  Denmark  0.66  0.33  4.2  Yes  0.55  0.62  Dualizing              Germany  0.27  0.09  1.8  No  0.60  0.72  France  0.37  0.11  1.5  No  0.69  0.70  Italy  0.15  0.06  0.5  No  0.63  0.73  Sources: a OECD (2016a), figures for 2010; b Bonoli (2013, p. 30), figures for 2007; c Tepe and Vanhuysse (2013); d Duell et al. (2010); e van Vliet and Caminada (2012), figures for 2009. Classifying Switzerland and Denmark as flexicurity countries is legitimate because they are similar in terms of traditional and—in particular—activation measures (Viebrock and Clasen 2009, Bonoli, 2013; see Shahidiet al. (2016) for a similar typology). First, as shown in Table 1, in the domain of activation policies, investment in ALMPs as a share of gross domestic product (GDP) for each percentage point of unemployment is very high at 0.18 and 0.33, respectively (Bonoli, 2013; Tepe and Vanhuysse, 2013). Moreover, the main pillars of successful activation consist of human capital rehabilitation and strict demand for a proactive employment search (Gerfin and Lechner 2002; Madsen, 2002; Duell et al., 2010; Knotz, 2015). Unsurprisingly, in Switzerland, 38% of jobseekers participated in training measures in 2009 (Duell et al., 2010; cf. Gerfin and Lechner, 2002). Finally, these are the only countries where ALMP expenditures increase automatically with unemployment levels (Duell et al., 2010). In other words, these countries follow a human capital fostering activation strategy (Barbier and Ludwig-Mayerhofer, 2004; Bonoli, 2013). With respect to traditional labour market policies, both countries have extremely flexible labour markets and loose EPL (Madsen, 2002; Viebrock and Clasen 2009; Clasen and Clegg, 2011). The EPL index produced by the Organisation for Economic Co-operation and Development (OECD) for all categories of workers amounts to merely 2.1 points for Denmark and 1.6 point for Switzerland in 2010 (the OECD average is 2.4 of 6 points). In contrast to dualizing countries, this flexibility also applies to workers with regular contracts (cf. Venn, 2009, p. 8; OECD, 2013; Svalund, 2013). A strong social partnership enables this level of flexibility because it compensates all workers with solid protection (Wilthagen, 2002), providing universal and generous unemployment coverage (columns 4 and 5). In 2009, the replacement rate (as a percentage of the previous wage) of a one-earner couple with two children in Switzerland amounted to 82% and to 73% in Denmark (van Vliet and Caminada, 2012). Secondly, I analyse Germany, France and Italy as representative dualization regimes (Palier, 2010). As shown in Table 1, in these countries, there is conspicuously less ALMP effort than in flexicurity countries, and it is characterized mainly by occupational strategies,6 i.e. pushing the employed back into the labour market or placing them in short-time work whilst failing to systematically invest in upskilling (Torfing, 1999; Barbier and Ludwig-Mayerhofer, 2004; Daguerre, 2007; Clasen and Clegg, 2011; Bonoli, 2013, p. 93ff.). In terms of traditional policies, dualizing countries are characterized by the existence of two parallel labour markets (Lindbeck and Snower, 1988; Emmenegger et al., 2012). Strong EPL regulates the primary labour market and creates favourable employment conditions. In contrast, the secondary labour market follows purely economic reasoning and offers precarious employment, loose hire-and-fire legislation and marginal—if any—welfare benefits (Venn, 2009; Palier, 2010). In these countries, precariousness is a serious problem (Berton et al. 2009; Emmenegger et al., 2012). For instance, there is a high incidence (approximately 49% in Italy and 20% in Germany) of non-voluntarily part-time work (compared to 7% in Switzerland and 15% in Denmark) (OECD, 2016b). Moreover, welfare provision is generous for insiders; however, the existence of two segregated labour markets and social support systems undermines attempts to reduce social inequality in these countries compared with flexicurity countries (Berton et al., 2009; Palier, 2010; Thelen, 2012). In sum, Switzerland and Denmark are flexibility countries because both respond to the challenges of liberalization by preserving generous and encompassing passive benefits, granting flexible labour markets and addressing labour market challenges through human capital-focused ALMPs. I compare them with three dualizing countries with generous but not encompassing passive benefits. Strictly regulated labour markets are at their core, but there is unpredictability at the margin, and they lag behind in terms of activation by implementing—if at all—pro-employment-focused ALMPs (Thelen, 2012; see also Rueda, 2007; Eichhorst and Konle-Seidl, 2008; Viebrock and Clasen 2009; Clasen and Clegg, 2011; Bonoli, 2013; Shahidi et al., 2016). In the next section, I formulate hypotheses about the effects of these institutional characteristics on political actors’ preferences, which in turn, constitute the basis for political actors’ coalition potential. 5. Hypotheses: political actors’ preferences in flexicurity and dualizing regimes As previously argued, it is important to keep in mind that actors’ preferences are shaped by their institutional frames of reference, i.e. by the strategies that are characteristic of a regime, the problems that are salient or insufficiently addressed, and the anticipation of job market-related outcomes concerning their constituencies. However, as I formulate expectations for very different actor types, I take into account additional theoretical considerations such as the actors’ ideological roots, i.e. stable traits that represent their core policy beliefs (Sabatier and Jenkins-Smith, 1999), and the interests of the constituency to which they cater (e.g. Rueda, 2007; Häusermann, 2010). Communist parties traditionally take a pro-labour stance with respect to labour–capital antagonism. Following Rueda (2005), I generally expect these organizations to focus on expanding traditional policies rather than on asking workers to adapt to the demands of the post-industrial labour market by taking part in ALMPs in any circumstance. In terms of regime differences, communist parties should be particularly sceptical of activation in dualizing countries because in austere times and with increasing flexibilization pressure, the depletion of insider privileges is looming. Moreover, in these countries, ALMPs do not aim at upskilling or, thus, at improving the labour market chances of the unemployed worker; rather, they push them back into precarious jobs. In dualizing countries, communist parties should focus on preserving the privileges of their traditional constituency, contemporaneously expanding these passive benefits to outsiders where possible. Conversely, I expect communists to favour activation in flexicurity countries because jobs are less precarious and ALMPs focus on upskilling. From a purely ideological perspective, social democratic parties should favour the expansion of both traditional measures and ALMPs to address the needs of both their traditional clientele (blue-collar workers) and contemporary vulnerable workers (outsiders). However, as argued in insider–outsider theory, policy legacies and austerity push social democratic parties to make strategic choices (Rueda, 2007). Consequently, I expect that in flexicurity regimes, social democrats advocate for the expansion of training-based ALMPs because both insiders and outsiders benefit from such policies and because insiders’ traditional interests are adequately accommodated (Emmenegger, 2009). Conversely, in dualized countries, traditional protections for insiders are highly endangered, and the interests of insiders and outsiders diverge because of strong EPL, insider fear of wage competition by outsiders, and precarious job expectations (Saint-Paul, 1998). Thus, social democrats support traditional policies to accommodate their unionized and more politically active core constituency rather than increasing activation effort (Bonoli, 2005; Rueda, 2007). I expect green parties to hold preferences similar to those of social democratic parties on the traditional conflict dimension (Kitschelt, 1994). The main difference is that the constituencies of green parties are younger and generally work in high-skilled occupations (Oesch, 2006; Geering and Häusermann, 2013). Consequently, these individuals are less likely to become unemployed, and if so, they are interested in re-entering the labour market swiftly. I thus expect green parties to favour the expansion of ALMPs, particularly in flexicurity countries where they benefit from upskilling. Liberal parties are mainly elected by highly skilled business-friendly individuals who favour strong market logic (Oesch, 2006; Häusermann, 2010). These parties focus on reducing spending on traditional policies to decrease their constituents’ tax burden and to increase the competitiveness of the labour market (Rueda, 2005). Whilst they are generally less likely to be affected by unemployment than the constituencies of leftist parties (Oesch, 2006), liberal parties might favour ALMPs as means to increase the employability of the workforce, especially in flexicurity countries where ALMPs are human capital centred. In contrast, in dualizing countries, liberal party support for activation should be weaker because occupational measures do not increase the expected productivity of workers. Right-wing and conservative parties are characterized by preferences for subsidiarity and self-reliance; hence, they are likely to reject the expansion of traditional policies (Huber and Stephens, 2001; Myles and Quadagno, 2002). As argued by Rueda (2007), their constituency comprises upscale groups such as (upper) middle class voters, business-friendly individuals, and employers. Particularly in flexicurity countries, they should be sceptical of increasing already substantial ALMP efforts and more likely to advocate for retrenchment on this axis. In contrast, in dualizing countries, they might be more receptive to increasing activation to foster the swift reintegration of the unemployed and, thus, decrease public deficits and increase the supply of cheap labour (outsiders). I expect that populist parties, which traditionally represent blue-collar workers (insiders), to hold the business-friendly positions of right-wing parties because by only advocating mainstream labour market policy positions are these organizations able to participate in government whilst remaining faithful to their extremist immigration policies (their core issue) (cf. Afonso, 2015). In other words, to pursue their office-seeking strategy, they adapt their preferences to those of the conservative mainstream. In terms of ideology, like social democratic parties, unions should advocate for the expansion of effort on both dimensions and thus fulfil their role as representatives of all workers (Esping-Andersen, 1990). The expectation differs if we lend credit to insider–-outsider theory, which equates union and social democratic preferences and ties them to the institutional context (Rueda, 2007; Gordon, 2015). According to this theory, unions should focus on the core interests of their members and prefer the expansion of traditional policies, particularly in dualized countries with strong ELP (Arnd, 2013). Conversely, they should be supportive of ALMPs expansion in flexicurity countries, where the preferences of insiders and outsiders converge owing to the increased unemployment risk of insiders (Saint-Paul, 1998; Emmenegger, 2009). With respect to traditional policies, employer organizations prioritize budgetary rigour over welfare expenditures (Esping-Andersen, 1990). Moreover, in terms of business interests, these organizations support ALMPs if they increase the employability of jobseekers and thus the pool of well-qualified candidates (Swank and Martin, 2001). Their support for expanding ALMPs should be strongest in flexicurity countries, where the benefits of human capital intervention and gains in productivity are highest. In contrast, in dualizing countries, where the policy mix is oriented towards occupational strategies, the expansion of ALMPs does not benefit employer organizations to the same extent, which is likely to reduce their support for such an expansion. State bodies and administrations should generally support the expansion of activation because they are in close contact and are thus influenced by supranational consensus (e.g. EES) that combines a neoliberal perspective on budgetary balance aimed at decreasing spending on traditional policies and the idea that activation is essential to reducing unemployment (Daguerre and Taylor-Gooby, 2004; Fleckenstein, 2008; Stiller and van Gerven, 2012). I expect that state bodies and administrations in dualizing countries that do not (yet) meet EES recommendations support the expansion of ALMPs more strongly than flexicurity countries where well-functioning ALMP schemes may lead a ceiling on support. Summarizing the hypotheses regarding activation preferences, in flexicurity countries, I expect social democratic, green and liberal parties, unions and employer organizations to favour increasing ALMP effort and right-wing parties to favour decreasing ALMP effort. In dualizing countries, I expect state bodies, administrations and right-wing parties to support the expansion of activation, whereas social democrats, communist parties and unions should prefer a decrease in activation. Finally, the expectations for social movement organizations (SMOs) are unclear because they depend on the nature of the initiative; therefore, their positioning is an empirical question. 6. Operationalization and methods To analyse political actors’ labour market policy preferences, I rely on novel interview data collected in Denmark, France, Germany, Italy and Switzerland in the autumn of 2010. The data set is based on semi-structured telephone interviews with major policymakers7 active in this domain, i.e. parties, unions, employer organizations, state bodies, administrations and SMOs.8 In the advocacy coalition and network analysis literature, the notion is that in the policymaking process, all experts in the policy field should be analysed to determine the potential for coalitions in terms of overlapping preferences (Knoke et al., 1996; Sabatier and Weible, 2007). Whilst the relevance of parties, social partners and state bodies are uncontested, that of SMOs is less straightforward. I argue that these organizations, although they have less political leverage, still play an important role during consultation processes as providers of external expertise and as representatives of labour market outsiders who are not usually unionized and who would not be represented otherwise. Moreover, they intervene as agenda setters or as co-organizers of demonstrations to raise awareness of neglected social problems (Knoke et al., 1996). The interviews were held between October and December 2010. It goes without saying that this period was characterized by a deep economic crisis. The advantage of analysing labour market policy preferences in this setting is that labour market issues were on the daily agenda and that political actors had strong incentives to take well-considered positions that were in line with their constituencies’ interests, thus rendering position shifts difficult to justify. Elite interview data are ideally suited to the analysis of political actors’ preferences because they allow their assessment independently of possibly idiosyncratic, very specific and/or limited political reform processes (e.g. Häusermann, 2010). Moreover, the precision of the questions enables the accurate operationalization of different ALMP types, whereas data that rely on extremely broad categories, such as the Manifesto Data Collection, do not. A strength of these data is that it captures both political actors’ preferences (a position measure) and their perceptions of the relative importance of particular policy measures (salience measure9). This allows for the construction of an indicator that weights an actor’s position10 by the salience of a measure.11 As shown in Table 2, I operationalize traditional conflict by means of ‘raising the minimum wage’, which is a state-led intervention to guarantee a decent living standard for workers, and ‘unemployment benefit reduction’, which captures benefit retrenchment. Two items operationalizing preferences for lower EPL (‘loosening hire-and-fire legislation’ and ‘increasing the flexibility of working hours’) are also included. Table 2. List of indicators Minimum wage  Raising the minimum wage.  Benefit reduction  Reducing unemployment benefits.  Working hours  Increasing the flexibility of working hours.  Hire-fire  Loosening hire-and-fire legislation.  Sanction  Imposing tougher sanctions for those who refuse to accept work that is deemed appropriate for them.  State job  Using state programmes to create jobs.  Training  Increasing retraining possibilities for the unemployed.  Short-time  Promoting short-time work—the ability of employers to reduce workers’ hours when orders are low.  Reintegration  Actively promoting reintegration into the labour market.  Minimum wage  Raising the minimum wage.  Benefit reduction  Reducing unemployment benefits.  Working hours  Increasing the flexibility of working hours.  Hire-fire  Loosening hire-and-fire legislation.  Sanction  Imposing tougher sanctions for those who refuse to accept work that is deemed appropriate for them.  State job  Using state programmes to create jobs.  Training  Increasing retraining possibilities for the unemployed.  Short-time  Promoting short-time work—the ability of employers to reduce workers’ hours when orders are low.  Reintegration  Actively promoting reintegration into the labour market.  Table 2. List of indicators Minimum wage  Raising the minimum wage.  Benefit reduction  Reducing unemployment benefits.  Working hours  Increasing the flexibility of working hours.  Hire-fire  Loosening hire-and-fire legislation.  Sanction  Imposing tougher sanctions for those who refuse to accept work that is deemed appropriate for them.  State job  Using state programmes to create jobs.  Training  Increasing retraining possibilities for the unemployed.  Short-time  Promoting short-time work—the ability of employers to reduce workers’ hours when orders are low.  Reintegration  Actively promoting reintegration into the labour market.  Minimum wage  Raising the minimum wage.  Benefit reduction  Reducing unemployment benefits.  Working hours  Increasing the flexibility of working hours.  Hire-fire  Loosening hire-and-fire legislation.  Sanction  Imposing tougher sanctions for those who refuse to accept work that is deemed appropriate for them.  State job  Using state programmes to create jobs.  Training  Increasing retraining possibilities for the unemployed.  Short-time  Promoting short-time work—the ability of employers to reduce workers’ hours when orders are low.  Reintegration  Actively promoting reintegration into the labour market.  Following Bonoli’s (2013) definition, I capture occupational ALMPs by means of ‘the use of state programmes to create jobs’ and in terms of subsidies to employers to employ individuals through a ‘promotion of short-time work’ (Sacchi et al., 2011). Secondly, incentive reinforcement is operationalized as ‘increasing sanctions when an unemployed person refuses a job, which is deemed appropriate’ (Trickey and Walker, 2001; Kemmerling and Bruttel, 2006; Clasen and Clegg, 2011; Fossati, 2017). I measure attitudes towards upskilling as ‘increasing of training effort’ and employment assistance by means of the ‘increasing reintegration effort’ measure. To analyse and display preferences, conflict dimensions and coalition potential in an accessible way, I conduct a regime-specific factor analysis (varimax rotation) on the nine indicators presented in Table 2. This analysis provides insight into whether there is regime-specific variation in the conflict dimensions that are identifiable in terms of the strength and direction of the factor loadings. Then, I retain the individual actors’ factor scores in order to locate them within a multidimensional policy space (Figures 1 and 2). Figure 1. View largeDownload slide Actor configuration in the flexicurity countries (Denmark and Switzerland). Denmark: Parties: SD Social Democrat Party, Venstre Liberal Party, DF Dansk Folkepartis, KF Konservative Folketsparti, RG Red-Green Alliance, SF Socialist Folkeparti; Unions: AC Akademikernes Centralorganisation, FTF Confederation of Professionals, LO Confederation of Trade Unions; Employer organisations: DA Confederation of Danish Employers, DI Confederation of Danish Industry; Administrations: NLMA National Labour Market Authority, DEC Economic Council. Social movementorganisations: SFI Danish national centre for social research; ECLM Economic Council of labour movement (AE), CEPOS Conservative think-tank. Switzerland: Parties: Gruene Grüne Partei Schweiz, FDP Freisinning Demorkatische Partei, die Liberalen, SVP Schweizerische Volkspartei, CVP Christlichdemokratische Volkspartei, BDP Bürgerlich Demokratische Partei, SP Sozialdemokratische Partei; Unions: Unia Unia, KV Kaufmännischer Verband Schweiz, Syna Syna Arbeitslosen Kasse, AS Angestellte Schweiz, SGB Scherizerischer Gewerkschaftsbund, TS Travail.Suisse, Gewerkschaftsdachorganisation; Employer organisations: SBV Dachverband Schweizerischer Baumeisterverband, SGV Schweizerischer Gewerbeverband, SAV Schweizerischer, Arbeitgeberverband, Swissmem Swissmem; Administrations: BE Canton Bern, AG Canton Argau, SODK Conference of the Cantonal Social Ministers, SECO State Secretary for Economic Affairs. Social movement organisations: Caritas Caritas Switzerland; Attac Attac Switzerland; Kabba Kabba SMO on behalf of the unemployed, AvS Avenir Suisse. Figure 1. View largeDownload slide Actor configuration in the flexicurity countries (Denmark and Switzerland). Denmark: Parties: SD Social Democrat Party, Venstre Liberal Party, DF Dansk Folkepartis, KF Konservative Folketsparti, RG Red-Green Alliance, SF Socialist Folkeparti; Unions: AC Akademikernes Centralorganisation, FTF Confederation of Professionals, LO Confederation of Trade Unions; Employer organisations: DA Confederation of Danish Employers, DI Confederation of Danish Industry; Administrations: NLMA National Labour Market Authority, DEC Economic Council. Social movementorganisations: SFI Danish national centre for social research; ECLM Economic Council of labour movement (AE), CEPOS Conservative think-tank. Switzerland: Parties: Gruene Grüne Partei Schweiz, FDP Freisinning Demorkatische Partei, die Liberalen, SVP Schweizerische Volkspartei, CVP Christlichdemokratische Volkspartei, BDP Bürgerlich Demokratische Partei, SP Sozialdemokratische Partei; Unions: Unia Unia, KV Kaufmännischer Verband Schweiz, Syna Syna Arbeitslosen Kasse, AS Angestellte Schweiz, SGB Scherizerischer Gewerkschaftsbund, TS Travail.Suisse, Gewerkschaftsdachorganisation; Employer organisations: SBV Dachverband Schweizerischer Baumeisterverband, SGV Schweizerischer Gewerbeverband, SAV Schweizerischer, Arbeitgeberverband, Swissmem Swissmem; Administrations: BE Canton Bern, AG Canton Argau, SODK Conference of the Cantonal Social Ministers, SECO State Secretary for Economic Affairs. Social movement organisations: Caritas Caritas Switzerland; Attac Attac Switzerland; Kabba Kabba SMO on behalf of the unemployed, AvS Avenir Suisse. Figure 2. View largeDownload slide Actor configuration in dualizing countries (Germany, France and Italy). France: Parties: UMP Union pour le Mouvement Populaire (UMP), PS Parti Socialiste, FN Front National, PCF Parti Communiste Français, LO Lutte Ouvrière, NPA Nouveau Parti Anticapitaliste, FO Force Ouvrière, EELV Europe Ecologie Les Verts; Unions: Solidaires Solidaires, UNSA Union nationale des Syndicats Autonomes, CFECGC Confédération Française de l'Encadrement, CFDT Confédération Française Démocratique du Travail, CFTC Confédération Française des Travailleurs, CGT Confédération Générale du Travail, FSU Fédération Syndicale Unitaire; Employer organisations: CGPME Confédération Générale des Petites et Moyennes Entreprises, MEDEF Mouvement des Entreprises de France, UNAPL Confédération Interprofessionnelle des Professions Libérales, UPA Union Professionnelle Artisanale; Administrations: MDT Ministère du Travail. Social movement organisations: AC! Agir contre le Chômage!; APEIS Association Pour l’Emploi, l’Inofrmation et la Solidarité; CNPE Comité national des privés d'emploi CGT; MNCP Mouvement National des Chômeurs et Précaires; SNC Solidarité nouvelles face au chômage. Germany: Parties: Linke Die Linke, CDU/CSU Christlich Demokratische Union/Christlich Soziale Union, NPD Nationaldemokratische Partei Deutschlands, FDP Freie Demokratische Partei/Die Liberalen, SPD Sozialdemokratische Partei Deutschlands, Gruene Grüne Partei Deutschland; Unions: KGA Koordinierungsstelle Gewerkschaftlicher Arbeitslosengruppen, IGM Industriegewerkschaft Metall, Verdi Vereinte Dienstleistungsgewerkschaft, DGB Deutscher Gewerkschaftsbund; Employer organisations: BDA Bundesvereinigung der Deutschen Arbeitgeberverbände; Administrations: BKS Bundesvereinigung Kommunale Spitzenverbände, BMAS Bundesministerium für Arbeit und Soziales, BWM Bundeswirtschaftsministerium, BA Bundesagentur für Arbeit. Social movement organisations: Caritas Caritas Germany; EFD Erwerbslosenforum; INSM Initiative Neue Soziale Marktwirtschaft; PW Paritätischer Wohlfahrtsverband; Bertesl Bertelsmannstiftung; IAB Institut für Arbeitsmarkt und Berufsforschung. Italy: Parties: PD Partito Democratico, SEL Sinistra Ecologia e Libertà, UdC Unione di Centro, PdL Popolo della Libertà, Lega Lega Nord, IdV Italia dei Valori, PdCI Partito dei Comunisti Italiani, PRC Rifondazione Comunista;Unions: COBAS Confederazione dei Comitati di Base, UIL Unione Italiana del Lavoro, CISL Confederazione Italiana Sindacati dei Lavoratori, CIGL Confederazione Generale Italiana del Lavoro; Employer organisations: Confindustria Confindustria; Administrations: MdL Ministero del Lavoro, INPS Istituto Nazionale Previdenza Sociale. Social movement organisations: ISFOL Istituto Sviluppo della Formazione Professionale dei Lavoratori; ACLI Associazioni Cristiane Lavoratori Italiani (Patronato CISL); ALVP Movimento Associazione Lavoratori Vittime del Precariato; INCA Istituto Nazionale Confederale di Assistenza (patronato CIGL); ARCI Associazione di Promozione Culturale; RdC Rete della Conoscenza. Figure 2. View largeDownload slide Actor configuration in dualizing countries (Germany, France and Italy). France: Parties: UMP Union pour le Mouvement Populaire (UMP), PS Parti Socialiste, FN Front National, PCF Parti Communiste Français, LO Lutte Ouvrière, NPA Nouveau Parti Anticapitaliste, FO Force Ouvrière, EELV Europe Ecologie Les Verts; Unions: Solidaires Solidaires, UNSA Union nationale des Syndicats Autonomes, CFECGC Confédération Française de l'Encadrement, CFDT Confédération Française Démocratique du Travail, CFTC Confédération Française des Travailleurs, CGT Confédération Générale du Travail, FSU Fédération Syndicale Unitaire; Employer organisations: CGPME Confédération Générale des Petites et Moyennes Entreprises, MEDEF Mouvement des Entreprises de France, UNAPL Confédération Interprofessionnelle des Professions Libérales, UPA Union Professionnelle Artisanale; Administrations: MDT Ministère du Travail. Social movement organisations: AC! Agir contre le Chômage!; APEIS Association Pour l’Emploi, l’Inofrmation et la Solidarité; CNPE Comité national des privés d'emploi CGT; MNCP Mouvement National des Chômeurs et Précaires; SNC Solidarité nouvelles face au chômage. Germany: Parties: Linke Die Linke, CDU/CSU Christlich Demokratische Union/Christlich Soziale Union, NPD Nationaldemokratische Partei Deutschlands, FDP Freie Demokratische Partei/Die Liberalen, SPD Sozialdemokratische Partei Deutschlands, Gruene Grüne Partei Deutschland; Unions: KGA Koordinierungsstelle Gewerkschaftlicher Arbeitslosengruppen, IGM Industriegewerkschaft Metall, Verdi Vereinte Dienstleistungsgewerkschaft, DGB Deutscher Gewerkschaftsbund; Employer organisations: BDA Bundesvereinigung der Deutschen Arbeitgeberverbände; Administrations: BKS Bundesvereinigung Kommunale Spitzenverbände, BMAS Bundesministerium für Arbeit und Soziales, BWM Bundeswirtschaftsministerium, BA Bundesagentur für Arbeit. Social movement organisations: Caritas Caritas Germany; EFD Erwerbslosenforum; INSM Initiative Neue Soziale Marktwirtschaft; PW Paritätischer Wohlfahrtsverband; Bertesl Bertelsmannstiftung; IAB Institut für Arbeitsmarkt und Berufsforschung. Italy: Parties: PD Partito Democratico, SEL Sinistra Ecologia e Libertà, UdC Unione di Centro, PdL Popolo della Libertà, Lega Lega Nord, IdV Italia dei Valori, PdCI Partito dei Comunisti Italiani, PRC Rifondazione Comunista;Unions: COBAS Confederazione dei Comitati di Base, UIL Unione Italiana del Lavoro, CISL Confederazione Italiana Sindacati dei Lavoratori, CIGL Confederazione Generale Italiana del Lavoro; Employer organisations: Confindustria Confindustria; Administrations: MdL Ministero del Lavoro, INPS Istituto Nazionale Previdenza Sociale. Social movement organisations: ISFOL Istituto Sviluppo della Formazione Professionale dei Lavoratori; ACLI Associazioni Cristiane Lavoratori Italiani (Patronato CISL); ALVP Movimento Associazione Lavoratori Vittime del Precariato; INCA Istituto Nazionale Confederale di Assistenza (patronato CIGL); ARCI Associazione di Promozione Culturale; RdC Rete della Conoscenza. The reliability of the actor constellations is assessed by means of three robustness checks. First, I replicate the analyses using only parties, social partners and state bodies (excluding SMOs). Moreover, the models are re-estimated without the issue salience weightings and for each country separately. These additional analyses show that the actor constellations are stable.12 7. The policy space: context-specific traditional and activation preferences The results of the factor analysis presented in Table 313 show that in both regimes, the policy space is two-dimensional and mirrors support/rejection of the two main strategies to fight unemployment, namely, traditional and activation strategies. Table 3. Traditional and activation factors by regime   Flexicurity   Dualized   Items  Denmark and Switzerland   Germany†, France and Italy     Traditional  Activation  Traditional  Activation  Sanction  0.52  −0.23  0.58  −0.09  Benefit reduction  0.71  −0.07  0.51  −0.26  Working hours  0.46  −0.33  0.71  −0.13  Hire-fire  0.46  0.03  0.55  −0.16  Minimum wage  −0.53  0.22  −0.76  −0.05  State job  −0.31  0.61  −0.72  −0.02  Training  −0.11  0.48  −0.30  0.40  Short-time  0.00  0.59  0.20  0.35  Reintegration  −0.09  0.22  −0.02  0.66  Eigenvalue  2.12  1.21  2.71  1.95  N  40  40  67  67    Flexicurity   Dualized   Items  Denmark and Switzerland   Germany†, France and Italy     Traditional  Activation  Traditional  Activation  Sanction  0.52  −0.23  0.58  −0.09  Benefit reduction  0.71  −0.07  0.51  −0.26  Working hours  0.46  −0.33  0.71  −0.13  Hire-fire  0.46  0.03  0.55  −0.16  Minimum wage  −0.53  0.22  −0.76  −0.05  State job  −0.31  0.61  −0.72  −0.02  Training  −0.11  0.48  −0.30  0.40  Short-time  0.00  0.59  0.20  0.35  Reintegration  −0.09  0.22  −0.02  0.66  Eigenvalue  2.12  1.21  2.71  1.95  N  40  40  67  67  † Attac Germany was excluded from the sample. Factor loadings in bold highlight correlations of the indicators and the factor, which are higher than the minimum correlation of 0.35. Table 3. Traditional and activation factors by regime   Flexicurity   Dualized   Items  Denmark and Switzerland   Germany†, France and Italy     Traditional  Activation  Traditional  Activation  Sanction  0.52  −0.23  0.58  −0.09  Benefit reduction  0.71  −0.07  0.51  −0.26  Working hours  0.46  −0.33  0.71  −0.13  Hire-fire  0.46  0.03  0.55  −0.16  Minimum wage  −0.53  0.22  −0.76  −0.05  State job  −0.31  0.61  −0.72  −0.02  Training  −0.11  0.48  −0.30  0.40  Short-time  0.00  0.59  0.20  0.35  Reintegration  −0.09  0.22  −0.02  0.66  Eigenvalue  2.12  1.21  2.71  1.95  N  40  40  67  67    Flexicurity   Dualized   Items  Denmark and Switzerland   Germany†, France and Italy     Traditional  Activation  Traditional  Activation  Sanction  0.52  −0.23  0.58  −0.09  Benefit reduction  0.71  −0.07  0.51  −0.26  Working hours  0.46  −0.33  0.71  −0.13  Hire-fire  0.46  0.03  0.55  −0.16  Minimum wage  −0.53  0.22  −0.76  −0.05  State job  −0.31  0.61  −0.72  −0.02  Training  −0.11  0.48  −0.30  0.40  Short-time  0.00  0.59  0.20  0.35  Reintegration  −0.09  0.22  −0.02  0.66  Eigenvalue  2.12  1.21  2.71  1.95  N  40  40  67  67  † Attac Germany was excluded from the sample. Factor loadings in bold highlight correlations of the indicators and the factor, which are higher than the minimum correlation of 0.35. As expected, the results show that political actors interpret activation14 differently depending on the institutional context (see the factor loadings). In the flexicurity countries, the ‘creation of public job[s]’ is associated with activation because the state is strongly engaged in creating re-employment opportunities, whereas in the dualizing countries, state efforts to create jobs remain predominantly connected with the traditional dimension, as this is a leftist strategy to reduce unemployment in the southern welfare states (Esping-Andersen, 1990). The results also indicate that the activation dimension in flexicurity countries is determined foremost by policies that highlight the active role of the state in creating job and training opportunities. Somewhat surprisingly, short-time work also resonates with activation in flexicurity countries. However, this finding can be explained because such work is a way to preserve workers’ skills by retaining them in firms and is accordingly in line with the human capital approach. The loading of ‘training’ is comparatively low because of the broad agreement among almost all actors on this proposition. Briefly, the activation axis in the flexicurity countries gathers measures that are characterized by a state that activates its workforce by means of direct human capital training, on-the-job learning or skill preservation measures. Policies that are not part of this repertoire, i.e. strategies focused on swift reintegration, display lower factor loadings and are hence less central to the conflict dimension. In dualizing countries, the activation dimension is determined by strategies to reintegrate the unemployed quickly without investing in training or public employment programmes. Interestingly, in these countries, short-time work has an ambiguous meaning; in fact, it is associated with both activation and traditional strategies because governments implement this measure to keep insiders on the payroll in times of crisis—a sort of insurance policy (Sacchi et al., 2011). At the same time, this policy helps preserve skills. This ambiguity can be clearly identified because in the dualizing regime, short-time work also scores, though less strongly, on the traditional axis. In sum, I find that policies that are characteristic of a regime determine political preferences more strongly (i.e. have high factor loadings) and thus mirror the different ALMP legacies: a human capital orientation in flexicurity and an occupational orientation in dualizing regimes. 7.1 Actor’s preferences and potential for coalition in flexicurity countries After describing the axes characterizing labour market conflicts, let us now analyse the single actor’s positions and the way in which these translate into coalition potential. Using the actors’ factor scores, I position them in the space. In Figures 1 and 2, I represent parties by a solid circle; state bodies, by a hollow circle; unions, by a hollow circle with a dotted border; employer organizations, by a hollow circle with a segmented border; and SMOs, by a small dot. The shading of the symbols differs across countries. Finally, I define potential coalitions to correspond to the quadrants, which result by crossing the traditional and activation dimensions. Quadrant II, which I call the modern left coalition, comprises actors who are in favour of expansion on both the traditional and the human capital imprinted activation dimensions. In the flexicurity countries, this group comprises several large unions, namely, the Swiss Confederation of Trade Unions (SGB), the Swiss unions Travail Suisse, Unia, and Union of Professionals (KV) and the Danish Union Federation (LO). Moreover, it includes the coordinating organ of the Swiss Cantons (SODK), the Danish Socialist Popular Party (SF) and the Danish Red-Green Alliance. Swiss greens are in favour of ALMP expansion, but they take a neutral stance on traditional policies. These results give some support to the hypothesis that green parties belong to this coalition. Furthermore, as expected based on insider–outsider theory, the bulk of unions favours expansion on both traditional and activation dimensions—at least in flexicurity countries. Quadrant III contains the traditional left coalition, which is characterized by the endorsement of passive benefits and EPL and sceptical attitudes towards increasing activation. The greatest supporters of the expansion of traditional labour market policies are the SMOs Attac and Kabba and, less so, the Danish white-collar union FTF. Whilst still part of the traditional left coalition, the Swiss union Syna; the Danish union AC; Swiss state actors, namely, the cantons of Aargau and Berne; the State Secretary for Economic Affairs (SECO); its Danish pendant NLMA; and the Danish Folkeparti (DF) adopt a moderate stance on the traditional dimension. Interestingly, the Swiss employer organization SAV is also part of this coalition, and it supports the status quo on the traditional dimension and is moderately against increasing activation efforts. These findings lend support to the hypothesis that state bodies and administrations in flexicurity countries should be sceptical of additional ALMP investments, as the implemented measures already work well. They are economically motivated to contain public spending and thus reject additional activation. The third-way coalition (Quadrant I) comprises a group of actors that share preferences for increased activation effort combined with a more or less market liberal stance. Specifically, Swiss (borderline) and Danish social democratic parties, Swiss greens (borderline), Swiss white-collar employees (AS) and, consistent with expectations, the largest Swiss employer organization (SGV) belong to this coalition. In fact, SGV combines a moderately positive stance on activation with a clearly market liberal position on the traditional axis. In sum, these organizations are strongly in favour of regime-specific activation strategies, but they simultaneously support the status quo or a reduction in passive welfare effort. Finally, in this group, the Swiss Christian Democratic Party (CVP) is an outlier, as it supports substantial liberalization and above average strengthening of ALMPs. This composition sustains the hypothesis that where welfare states address traditional risks efficiently and create attractive jobs for the previously unemployed, social democratic parties turn towards groups that are newly at risk. They do so by advocating the status quo or adopting a moderately liberal stance on the traditional axis but clearly supporting the expansion of human capital based ALMPs. Moreover, the results contradict the hypothesis that state bodies and administrations in flexicurity countries endorse increasing activation efforts because of their contact with supranational bodies. In contrast, I find that these organizations are oriented towards preserving the status quo on the traditional dimensions and rejecting additional expansion of activation polices. The traditional right coalition (Quadrant IV), which favours retrenchment on both dimensions, mainly includes employer organizations and conservative parties, i.e. the Swiss People’s Party (SVP) and the Danish Conservative party (KF). The Danish employer organization DI and the Swiss SBV share an extreme position in which they strongly reject additional activation, while the union Swissmem and the Danish DA have a very liberal stance on the traditional axis. These results contradict my expectation that employer organizations should support activation more strongly in flexicurity countries owing to the high standards of their retraining policies. Finally, there is a moderate subgroup of traditional right actors, namely, the Danish Economic Council (DEC), the Swiss conservative democrats (BDP), the Swiss liberal party (FDP), and the Danish liberal party (Venstre), which in terms of traditional polices, favour the status quo. Overall, the actor constellation suggests that in countries where welfare state benefits are encompassing, activation policies are human capital centred, and where the interests of insiders and outsiders converge owing to flexible labour market regimentation and acceptable jobs, unions favour expansion on both dimensions and social democratic parties cater to outsiders. In contrast, state bodies, employer organizations, and conservative and liberal parties are sceptical of additional activation. 7.2 Actor’s preferences and potential coalitions in dualizing countries As shown in Figure 2, in dualizing countries, unions are spread across both left coalitions. In the modern left coalition, we find several major unions, namely, the German IGM, DGB and Verdi; the French UNSA and CFECGC; and the Italian Christian democratic union (CISL). Contrary to my expectations, (former) communists (PCF in France, PdCI and PRC in Italy and die Linke in Germany) also cluster in this quadrant. These parties clearly favour increasing activation by addressing outsiders’ interests rather than focusing solely on their traditional constituency. Similar to the finding for the flexicurity countries, Figure 2 shows that the greens in Germany (Grüne), Italy (SEL) and France (EELV) favour expansion on both dimensions. Thus, they are not less supportive of activation in dualizing countries even though this strategy is less focused on human capital enhancement in such countries than in flexicurity countries and jobseekers are at risk of being relegated to precarious jobs. In the traditional left cluster, we find the German and the French socialist parties (SPD and PS), the left-oriented Italia dei Valori (IdV), the centrist UdC in Italy and the other unions (the Italian grass-roots organized COBAS and the UIL and the French CGT, Solidaires, FSU and CFDT). Only the Italian confederation of trade unions (CIGL) and the French CFTC belong to the third-way coalition and hence strongly support additional activation measures whilst taking a neutral stance on the traditional axis. These results corroborate previous studies and the expectation that in dualizing labour markets, social democrats stick to the traditional solution of strengthening policies to address insiders’ needs whilst neglecting outsiders’ interests in increasing activation effort. State bodies and administrations were expected to belong to the third-way coalition. Figure 2 shows that this hypothesis can be corroborated for the French Ministry of Work (MdT), the German Ministry of Social Affairs (BMAS), and the Italian state bodies, i.e. the Italian Ministry of Work (MdL) and the Ministry for Social Insurance (INPS). Conversely, the German Ministry for Economic Affairs (BWM) and the Ministry for Work (BAA) are situated in the right coalition, which in the light of their liberal-leaning Hartz IV reform, might not be surprising. Interestingly, German socialists (SPD), who initiated these reforms under Schröder, also tend towards a centrist position on the traditional axis in comparison with their Italian and French counterparts, who take positions further to the left. Moreover, state bodies share a moderate stance on the traditional axis but differ widely on the activation dimension. It appears that in dualizing countries, organizations that address social affairs and work are heavily influenced by the third-way consensus at the supranational level. This finding is plausible because these actors address the need for welfare reform. Conversely, in Germany, since major activation reforms were successfully implemented, the BWM and the BAA are more focused on balancing budgets than on increasing state involvement in either dimension. In the dualizing countries, the right coalition is extremely divided on the traditional dimension, but it has a homogeneous stance in refusing to increase occupational activation efforts. These right-oriented actors seem to disagree foremost on whether employment should be liberalized further. This issue has been particularly pushed by employer organizations and government parties in dualizing countries to counteract the rigidities of continental welfare states’ labour markets. The hypothesis that conservative government parties belong the third-way coalition is thus only partially corroborated. In fact, both the German CSU/CDU and the Italian PdL are part of the right coalition, whilst the French UMP, Lega and Italian PD are in the third-way coalition that endorses the expansion of ALMPs. In sum, I find that social democratic parties are part of the traditional left coalition catering to insiders and that greens and communists favour expansion on both dimensions and thus cater to both insiders and outsiders (modern left coalition). On the other extreme of the traditional dimension, employer associations and conservative political parties prefer the retrenchment of both traditional and activation policies. State bodies and administrations in Germany, where ALMPs have already been massively reformed, are part of the traditional right coalition. Conversely, in Italy and France, the administration supports third-way strategies by following the supranational consensus on the re-commodification of the labour force. 8. Conclusion In this article, I argue that political actors’ preferences are best understood as a product of the institutional context in which they are embedded. In fact, labour market regimes are the frames of reference against which actors form their preferences by shaping the way in which political problems are interpreted and adequate policy strategies are sought. The main objective of this article has been to show how the core players in the domain of labour market policy position themselves vis-à-vis the main policies used to fight unemployment, namely, traditional policies (EPL and passive benefits) and activation policies (training, reintegration measures and sanctions). The empirical analyses produced three main findings. First, the factor analyses show that the labour market policy space is characterized by two regime-specific conflict axes. Interestingly, in flexicurity countries, in line with their human capital ALMP tradition, the policies that determine the activation dimension are related to training and linked to strong engagement by the state in creating employment opportunities. In contrast, dualizing countries focus on reintegrating unemployed workers into the labour market without a focus on upskilling (occupational strategies). Secondly, I show that actors’ preferences and thus their potential coalitions also depend on the institutional setup. In dualizing countries, where pressure to adapt ALMPs to the EES guidelines is highest, I find that state bodies and administrations are more supportive of expansion on the activation axis. In contrast, in flexicurity countries, where these schemes are already encompassing and functioning well, they seem to think there is no need to burden the welfare budget by expanding these policies. Accordingly, state bodies are in the traditional right coalition, which supports retrenchment on both dimensions. The results also show that unions—particularly in flexicurity countries—communist parties and greens in both regimes embrace modern left policy strategies, favouring expansion on both the activation and traditional dimensions. Conversely, many of the large unions in dualizing countries have joined forces with social democratic parties to protect insiders’ interests; they reject the expansion of activation policies but support the expansion of traditional policies. I also find that, in both regimes, right-oriented parties and employer organizations have the potential to coalesce; they are against expansion on both the traditional and activation dimensions. Finally, focusing on social democratic parties, this new data set allows me to corroborate the findings of Rueda (2007) and show that these parties concentrate on labour market insiders in dualizing countries (a traditional left strategy) but extend their mobilization to outsiders in flexicurity countries (a third-way strategy). In sum, political actors’ preferences, and thus their potential coalitions, differ depending on the institutional design. What are the implications of these findings? My results challenge the widespread practice in political economy to compare preferences, coalitions and conflict patterns across countries without accounting for the intuitional peculiarities that shape political action. This likely gives rise to regime-specific cross-class alliance potentials: in flexicurity countries, state bodies and administrations might plausibly build powerful alliances with employers and conservative parties to retrench state efforts, whereas in dualizing countries, state bodies may join forces with unions and left opposition parties to address outsiders’ needs. My results also indicate that the composition of the conflict structure in a specific policy field does not always refer to the same content. The factor analyses show that the traditional and activation dimensions differ across regimes, a result that calls for more sophisticated analyses of political preferences in general and of the validity of non-specific indexes, such as the left-right dimension, in particular. By studying the basic conflict lines that structure political preferences concerning labour market policy, I unveil the fundamental preferences of a variety of actors and hence lay the groundwork for understanding and studying policymaking in this domain. Future research could focus on coalition formation in phases, where strategic alliances gain importance beyond mere ideological commonalities, or trace coalitions over longer periods or in other contexts. Supplementary material Supplementary material is available at Socio-Economic Review Journal online. Funding This research has been supported by the NCCR Democracy 21, Project 11, which is financed by the Swiss National Science Foundation. Footnotes 1 Political actors’ preferences determine their potential to form coalitions with other actors based on common ideological and/or value-based ground. These preferences and coalitions constitute the conflict structure in a policy field (cf. Kriesi et al., 2008; Häusermann, 2010). 2 I use ‘position’ and ‘preference’ interchangeably because I merely analyse value-based or ideology-based positions and do not include strategic motives, which could lead to a distinction between real coalitions and expressed preferences (Immergut, 1998). 3 Institutional legacies and actors’ preferences are, to some extent, endogenous since preferences articulated in a political process shape the institutional settings, and vice versa. Here, I analyse actors’ preferences but not the politics that may follow from expressing these attitudes. 4 Flexicurity is a debated concept. I follow Whiltagen and Tros (2004) and define flexicurity as ‘[t]he pursuit of a balance between “flexibility” and “security” …’. This scheme supports weaker groups and not merely labour market insiders, and it results from the coordination effort of social partnership and the state, allowing the synchronization of social and economic policy. As argued by Viebrock and Clasen (2009), flexicurity countries such as Denmark rely on flexible labour markets, generous unemployment support and activation (cf. Madsen, 2002). 5 To analyse how specific labour market characteristics affect political actors’ preferences, I group countries with similar characteristics into labour market regime types. I refrain from analysing political actors’ preferences by country because I am not interested in the effect of possibly idiosyncratic country-specific institutional settings on preferences. 6 The third approach identified in the literature is referred to as ‘work-first’ activation. This model, which is characteristic of liberal countries, stresses the need to swiftly reintroduce workers into the labour market, principally by means of conditionality and negative incentives. 7 Originally, 132 political actors were contacted, and 109 interviews were obtained (refer to Tables A1 and A2 in the Online Appendix for the details on data collection and response rates). 8 Interview partners were chosen as representatives of the major organizations that are involved in the field of unemployment policy. The relevance of the organizations was cross-checked with two experts per country and validated by means of media analyses. See Tables A3 and A4 in the Online Appendix for the complete list of actors. 9 To capture fundamental political conflicts, it is essential to focus on salient policies because political actors have a stance on all issues, but they judge them differently in terms of relevance. Accordingly, I calculate the indicator by multiplying the standardized salience and position for each actor. This procedure attributes less weight to measures that actors consider less relevant. An actor’s position on a given policy measure was gauged on a scale ranging from 1 to 5 (strongly disagree to strongly agree). To operationalize a measure’s salience, the respondents indicated the most important measure, the three most important measures, and the three least important measures. The resulting salience indicator assigns three points to the most important measure, two points to the other two important measures, zero points to the three least important measures and one point to the remaining measures. 10 See Table A5 in the Online Appendix for the question wording. 11 Missing cases were recoded as neutral both in position and salience; however, they represent only between 2% and 5% of cases, see Tables A6 and A7 in the Online Appendix for descriptive statistics and correlations. 12 The results are not shown but are available upon request. 13 See Tables A8 and A9 and Figures A1–A5 in the Online Appendix for country-specific solutions. 14 The findings suggest that there is also variation in the understanding of traditional policies, but for reasons of space, I focus on activation preferences. However, the fact that regime specificities exist on the traditional axis allows for the generalization of the argument that institutional legacies influence political actors’ preferences. Acknowledgements The author would like to thank Silja Häusermann, Hanspeter Kriesi, Giuliano Bonoli, Ruth Beckmann, Fabienne Liechti, Desmond King, Alexandre Afonso, Laurent Bernhard, Regula Hänggli, four anonymous reviewers and the editors of Socio-Economic Review for very helpful and constructive comments. The author would also thank Laurent Bernhard, Regula Hänggli and Kirsty Stone-Wyler for excellent cooperation during the data collection process. References Afonso A. ( 2015) ‘ Choosing Whom to Betray: Populist Right-Wing Parties, Welfare State Reforms and the Trade-Off between Office and Votes’, European Political Science Review , 7, 271– 292. DOI: 10.1017/S1755773914000125. Google Scholar CrossRef Search ADS   Arnd C. ( 2013) The Electoral Consequences of Third Way Welfare State Reforms: Social Democracy’s Transformation and Its Political Costs , Amsterdam, Amsterdam University Press. Google Scholar CrossRef Search ADS   Barbier J., Ludwig-Mayerhofer W. ( 2004) ‘ Introduction’, European Societies , 6, 423– 436. DOI: 10.1080/1461669042000275845. Google Scholar CrossRef Search ADS   Berton F., Richiardi M., Sacchi S. ( 2009) Flex-Insecurity: Perché in Italia la flessibilità Diventa precarietà , Bologna, Il Mulino. Bonoli G. ( 2005) ‘ The Politics of the New Social Policies: Providing Coverage against New Social Risks in Mature Welfare States’, Policy and Politics , 33, 431– 449. DOI: 10.1332/0305573054325765. Google Scholar CrossRef Search ADS   Bonoli G. ( 2013) The Origins of Active Social Policy. Labour Market and Childcare Policies in a Comparative Perspective , Oxford, Oxford University Press. Google Scholar CrossRef Search ADS   Bonoli G., Natali D. ( 2012) The Politics of the ‘New’ Welfare State: Analysing Reforms in Western Europe , Oxford, Oxford University Press. Clasen J., Clegg D. (ed.) ( 2011) Regulating the Risk of Unemployment. National Adaptations to Post-Industrial Labour Markets in Europe , Oxford, Oxford University Press. Google Scholar CrossRef Search ADS   Daguerre A. ( 2007) Active Labour Market Policies and Welfare Reform. Europe and the US in Comparative Perspective , Houndmills/NY, Palgrave Macmillan. Google Scholar CrossRef Search ADS   Daguerre A., Taylor-Gooby P. ( 2004) ‘ Neglecting Europe: Explaining the Predominance of American Ideas in New Labour's Welfare Policies Since 1997’, Journal of European Social Policy , 14, 25– 39. DOI: 10.1177/0958928704039786. Google Scholar CrossRef Search ADS   Duell N., Tergeist P., Bazant U., Cimper S. ( 2010) Activation Policies in Switzerland, OECD Social Employment and Migration Working Papers, accessed at www.oecd.org/els/workingpapers on July 16, 2016. Eichhorst W., Konle-Seidl R. ( 2008) Contingent Convergence: A Comparative Analysis of Activation Policies, IZA Discussion Paper Series, No. 3905, Bonn, Institute for the Study of Labor (IZA). Emmenegger P. ( 2009) ‘ Barriers to Entry: Insider/Outsider Politics and the Political Determinants of Job Security Regulations’, Journal of European Social Policy , 19, 131– 146. DOI: 10.1177/0958928708101866. Google Scholar CrossRef Search ADS   Emmenegger P., Häusermann S., Palier B., Seeleib-Kaiser M. (eds) ( 2012) The Age of Dualisation. The Changing Face of Inequality in Deindustrializing Societies , Oxford, Oxford University Press. Google Scholar CrossRef Search ADS   Entman R. M. ( 1993) ‘ Framing: Toward Clarification of a Fractured Paradigm’, Journal of Communication , 43, 51– 58. DOI: 10.1111/j.1460-2466.1993.tb01304.x. Google Scholar CrossRef Search ADS   Esping-Andersen G. ( 1990) Three Worlds of Welfare State Capitalism , Princeton, Princeton University Press. Fleckenstein T. ( 2008) ‘ Restructuring Welfare for the Unemployed: The Hartz Legislation in Germany’, Journal of European Social Policy , 18, 177– 188. DOI: 10.1177/0958928707087593. Google Scholar CrossRef Search ADS   Fossati F. ( 2017) ‘Who Wants Demanding Active Labour Market Policies? Public Attitudes towards Policies That Put Pressure on the Unemployed’, Journal of Social Policy , 13, 1– 21. DOI: 10.1017/S0047279417000216. Google Scholar CrossRef Search ADS   Fossati F., Häusermann S. ( 2014) ‘ Social Policy Preferences and Party Choice in the 2011 Swiss Elections’, Swiss Political Science Review’ , 20, 590– 611. DOI: 10.1111/spsr.12115. Google Scholar CrossRef Search ADS   Geering D., Häusermann S. ( 2013) ‘Changing Party Electorates and Economic Realignment. Explaining Party Positions on Labor Market Policy in Western Europe’, paper presented at the ECPR General Conference, Bordeaux, France, September 4–7. Gerfin M., Lechner M. ( 2002) ‘ A Microeconometric Evaluation of the Active Labour Market Policy in Switzerland’, Economic Journal , 112, 854– 893. DOI: 10.1111/1468-0297.00072. Google Scholar CrossRef Search ADS   Gilbert N. ( 2002) Transformation of the Welfare State. The Silent Surrender of Public Responsibility , Oxford, Oxford University Press. Google Scholar CrossRef Search ADS   Gordon J. C. ( 2015) ‘ Protecting the Unemployed: Varieties of Unionism and the Evolution of Unemployment Benefits and Active Labor Market Policy in the Rich Democracies’, Socio-Economic Review , 13, 79– 99. DOI: 10.1093/ser/mwu029. Google Scholar CrossRef Search ADS   Hall P. A., Soskice D. (eds) ( 2001) Varieties of Capitalism , Oxford, Oxford University Press. Google Scholar CrossRef Search ADS   Häusermann S. ( 2010) The Politics of Welfare State Reform in Continental Europe. Modernization in Hard Times , Cambridge, Cambridge University Press. Google Scholar CrossRef Search ADS   Huber E., Stephens J. ( 2001) Development and Crisis of the Welfare State. Parties and Policies in Global Markets , Chicago, University of Chicago Press. Google Scholar CrossRef Search ADS   Immergut E. M. ( 1998) ‘ The Theoretical Core of the New Institutionalism’, Politics and Society , 26, 5– 34. DOI: 10.1177/0032329298026001002. Google Scholar CrossRef Search ADS   Kemmerling A., Bruttel O. ( 2006) ‘“ New Politics” in German Labour Market Policy? The Implications of the Recent Hartz Reforms for the German Welfare State’, West European Politics , 29, 90– 112. DOI: 10.1080/01402380500389273. Google Scholar CrossRef Search ADS   Kitschelt H. ( 1994) The Transformation of European Social Democracy , Cambridge, Cambridge University Press. Google Scholar CrossRef Search ADS   Knoke D., Pappi F. U., Broadbent J., Tsujinaka Y. ( 1996) Comparing Policy Networks. Labor Politics in the U.S., Germany, and Japan , New York, Cambridge University Press. Google Scholar CrossRef Search ADS   Knotz C. ( 2015) The Politics of Unemployment Benefit Conditionality , Lund, Lunds Universitet. Korpi W. ( 1983) The Democratic Class Struggle , London, Routledge. Kriesi H., Grande E., Lachat R., Dolezal M., Bornschier S., Frey T. ( 2008) West European Politics in the Age of Globalization , Cambridge, Cambridge University Press. Google Scholar CrossRef Search ADS   Larsen C. A. ( 2007) ‘ The Institutional Logic of Welfare Attitudes. How Welfare Regimes Influence Public Support’, Comparative Political Studies , 41, 145– 168. Google Scholar CrossRef Search ADS   Lindbeck A., Snower D. J. ( 1988) The Insider-Outsider Theory of Employment and Unemployment , Cambridge, MIT Press. Lowi T. J. ( 1972) ‘ Four Systems of Policy, Politics, and Choice’, Public Administration Review , 32, 298– 310. DOI: 10.2307/974990. Google Scholar CrossRef Search ADS   Madsen P. ( 2002) ‘The Danish Model of Flexicurity: A Paradise with Some Snakes’. In Sarfati H., Bonoli G. (eds) Labour Market and Social Protection Reforms in International Perspective , Aldershot, Ashgate, pp. 243– 265. Myles J., Quadagno J. ( 2002) ‘ Political Theory of the Welfare State’, Politics & Society , 76, 34– 57. Nelson M. ( 2013) ‘ Making Markets with Active Labor Market Policies: the Influence of Political Parties, Welfare State Regimes, and Economic Change on Spending on Different Types of Policies’, European Political Science Review , 5, 255– 277. DOI: 10.1017/S1755773912000148. Google Scholar CrossRef Search ADS   Oesch D. ( 2006) Redrawing the Class Map. Stratification and Institutions in Britain , Germany, Sweden and Switzerland, Houndsmill/NY, Palgrave Macmillan. Google Scholar CrossRef Search ADS   Organisation for Economic Co-operation and Development (OECD) ( 2013) ‘Employment Protection Indicator, OECD Employment and Labour Market Statistics (Database)’, accesed at http://www.oecd.org/employment/emp on February 2, 2015. Organisation for Economic Co-operation and Development (OECD) ( 2016a) ‘Public Expenditures and Participation Stocks in LMP’, accessed at stats.oecd.org on February 24, 2016. Organisation for Economic Co-operation and Development Organisation for Economic Co-operation and Development (OECD) ( 2016b) ‘The Incidence of Involuntary Part-Time Workers’, accessed at https://stats.oecd.org/Index.aspx?DataSetCode=TEMP_I# on February 24, 2016. Palier B. (ed.) ( 2010) A Long Goodbye to Bismarck? The Politics of Welfare State Reform in Continental Europe , Amsterdam, Amsterdam University Press. Google Scholar CrossRef Search ADS   Pierson P. ( 1998) ‘ Irresistible Forces, Immovable Objects: Post-Industrial Welfare States Confront Permanent Austerity’, Journal of European Public Policy , 5, 539– 560. DOI: 10.1080/13501769880000011. Google Scholar CrossRef Search ADS   Pierson P. ( 2001) ‘Coping with Permanent Austerity. Welfare State Restructuring in Affluent Democracies’. In Pierson P. (ed.) The New Politics of the Welfare State , Oxford, Oxford University Press, pp. 410– 459. Google Scholar CrossRef Search ADS   Rueda D. ( 2005) ‘ Insider-Outsider Politics in Industrialised Democracies: The Challenge to Social Democratic Parties’, American Political Science Review , 99, 61– 74. Google Scholar CrossRef Search ADS   Rueda D. ( 2007) Social Democracy Inside Out. Partisanship and Labor Market Policy in Industrialized Democracies , Oxford, Oxford University Press. Google Scholar CrossRef Search ADS   Sabatier P. A., Jenkins-Smith H. C. ( 1999) ‘The Advocacy Coalition Framework: An Assessment’. In Sabatier P. A. (ed.) Theories of the Policy Process , Boulder, Westview and Publishing House, pp. 117– 166. Sabatier P. A., Weible C. M. ( 2007) ‘The Advocacy Coalition Framework. Innovations and Clarifications’. In Sabatier P. A. (ed.) Theories of the Policy Process , Cambridge, Westview Press, pp. 189– 220. Sacchi S., Pancaldi F., Arisi C. ( 2011) ‘ The Economic Crisis as a Trigger of Convergence? Short-Time Work in Italy, Germany and Austria’, Social Policy & Administration , 45, 465– 487. DOI: 10.1111/j.1467-9515.2011.00785.x. Google Scholar CrossRef Search ADS   Saint-Paul G. ( 1998) ‘ A Framework for Analyzing the Political Support for Active Labor Market Policy’, Journal of Public Economics , 67, 151– 165. DOI: 10.1016/S0047-2727(97)00069-8. Google Scholar CrossRef Search ADS   Shahidi F. V., Moortel D. de, Muntaner C., Davis O., Siddiqi A. ( 2016) ‘ Do Flexicurity Policies Protect Workers from the Adverse Health Consequences of Temporary Employment? A Cross-National Comparative Analysis’, SSM - Population Health , 2, 674– 682. DOI: 10.1016/j.ssmph.2016.09.005. Google Scholar CrossRef Search ADS   Steinmo S. ( 1989) ‘ Political Institutions and Tax Policy in the United States, Sweden, and Britain’, World Politics , 41, 500– 535. DOI: 10.2307/2010528. Google Scholar CrossRef Search ADS   Stiller S., van Gerven M. ( 2012) ‘ The European Employment Strategy and National Core Executives: Impacts on Activation Reforms in the Netherlands and Germany’, Journal of European Social Policy , 22, 118– 132. DOI: 10.1177/0958928711433652. Google Scholar CrossRef Search ADS   Svalund J. ( 2013) ‘ Labor Market Institutions, Mobility, and Dualization in the Nordic Countries’, Nordic Journal of Working Life Studies , 3. DOI: 10.19154/njwls.v3i1.2523. Swank D., Martin C. J. ( 2001) ‘ Employers and the Welfare State’, Comparative Political Studies , 34, 889– 923. DOI: 10.1177/0010414001034008003. Google Scholar CrossRef Search ADS   Tepe M., Vanhuysse P. ( 2013) ‘ Parties, Unions and Activation Strategies: The Context-Dependent Politics of Active Labour Market Policy Spending’, Political Studies , 61, 480– 504. DOI: 10.1111/j.1467-9248.2012.00996.x. Google Scholar CrossRef Search ADS   Thelen K. ( 2012) ‘ Varieties of Capitalism: Trajectories of Liberalization and the New Politics of Social Solidarity’, Annual Review of Political Science , 15, 137– 159. DOI: 10.1146/annurev-polisci-070110-122959. Google Scholar CrossRef Search ADS   Torfing J. ( 1999) ‘ Towards a Schumpeterian Workfare Post National Regime: Path-Shaping and Path-Dependency in Danish Welfare State Reform’, Economy and Society , 28, 369– 402. DOI: 10.1080/03085149900000010. Google Scholar CrossRef Search ADS   Trickey H., Walker R. ( 2001) ‘Steps to Compulsion within British Labour Market Policies’. In Lødemel I., Trickey H. (eds) An Offer You Can't Refuse. Workfare in International Perspective , Bristol, The Policy Press, pp. 181– 213. van Vliet O., Caminada K. ( 2012) ‘Unemployment Replacement Rates Dataset among 34 Welfare States, 1971–2009. An Update, Extension and Modification of the Scrugg’s Welfare State Entitlement Data Set’. Neujobs Special Report, No. 2. Venn D. ( 2009) Legislation, Collective Bargaining and Enforcement: Updating the OECD Employment Protection Indicators, Employment and Migration Working Paper No. 89. Viebrock E., Clasen J. ( 2009) Flexicurity – A State-Of-the-Art Review. Working Papers on the Reconciliation of Work and Welfare in Europe, No. 1, RECWOWE Publication, Edinburgh, Dissemination and Dialogue Centre. Vlandas T. ( 2013) ‘ Mixing Apples with Oranges? Partisanship and Active Labour Market Policies in Europe’, Journal of European Social Policy , 23, 3– 20. DOI: 10.1177/0958928712463161. Google Scholar CrossRef Search ADS   Whiltagen T., Tros F. ( 2004) ‘“The Concept of Flexicurity”. A New Approach to Regulating Employment and Labour Markets’, Transfer, European Review of Labour and Research , 10, S. 166– 186. Google Scholar CrossRef Search ADS   Wilthagen T. ( 2002) ‘Managing Social Risks with Transitional Labour Markets’. In Mosley H., O’Reilly J., Schönmann K. (eds) Labour Markets, Gender and Institutional Change. Essays in Honour of Guenther Schmid . Cheltenham, Edward Elgar, pp. 264– 289. Wüest B., Fossati F. ( 2014) ‘ Quantitative Discursive Institutionalism: A Comparison of Labour Market Policy Discourse across Western Europe’, Journal of European Public Policy , 22, 708– 730. Google Scholar CrossRef Search ADS   © The Author 2017. Published by Oxford University Press and the Society for the Advancement of Socio-Economics. All rights reserved. For Permissions, please email: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Socio-Economic Review Oxford University Press

How regimes shape preferences. A study of political actors’ labour market policy preferences in flexicurity and dualizing countries

Socio-Economic Review , Volume Advance Article – Nov 2, 2017

Loading next page...
 
/lp/ou_press/how-regimes-shape-preferences-a-study-of-political-actors-labour-bXGPMNzZYM
Publisher
Oxford University Press
Copyright
© The Author 2017. Published by Oxford University Press and the Society for the Advancement of Socio-Economics. All rights reserved. For Permissions, please email: journals.permissions@oup.com
ISSN
1475-1461
eISSN
1475-147X
D.O.I.
10.1093/ser/mwx040
Publisher site
See Article on Publisher Site

Abstract

Abstract Political actors do not operate in a vacuum; rather, it is safe to assume that their preferences are influenced by the institutional context in which they operate. By means of novel interview data, which was collected in flexicurity countries, i.e. Denmark and Switzerland, and in dualizing countries, i.e. France, Germany and Italy, I investigate the preferences of parties, unions, state bodies, employers and social movement organizations towards traditional and activation strategies. I find that the institutional context indeed shapes preferences. The results reveal, for instance, that state bodies reject increasing activation efforts in flexicurity countries but support it in dualizing countries. Moreover, in line with previous research, social democratic parties are found to cater to the interests of insiders by endorsing the expansion of traditional measures in dualizing countries, while focusing on outsiders’ interests by preferring the expansion of activation in flexicurity countries. 1. Introduction Political actors do not operate in a vacuum; rather, it is safe to assume that their policy preferences are shaped by the institutional context in which they operate. However, much of the comparative political economy literature continues to assume constant actor preferences regardless of the context. This article takes a different perspective, and it shows how institutional factors shape the positions that political actors take on labour market policy. Specifically, I analyse actors’ preferences in relation to the main strategies adopted to fight unemployment and contrast traditional versus activation approaches. The objective of the analysis is to show how these preferences are affected by the institutional setup, namely, the labour market regime. Preferences on traditional strategies concern support for the expansion/retrenchment of measures such as passive unemployment benefits and employment protection legislation (EPL). Preferences on activation refer to—support for the expansion/retrenchment of different active labour market policies (ALMPs). These include ‘enabling’ and ‘demanding’ measures, such as training and counselling on the one hand and negative incentives and sanctions on the other (Eichhorst and Konle-Seidl, 2008; Bonoli 2013; Fossati, 2017). While accounting for both types of preferences, I concentrate on attitudes towards ALMPs because of the importance these measures gained over the past decades. In fact, activation seems to have affected the entire welfare state, transforming the scope of its provision from a ‘securing’ to an ‘enabling’ institution (Clasen and Clegg, 2011; Bonoli, 2013). In the light of this massive welfare reorientation, it is surprising that little effort has been made to assess political actors’ activation preferences in a systematic and direct way, as this could improve the understanding of current reform processes (Rueda, 2007; Nelson 2013; Tepe and Vanhuysse, 2013; Vlandas, 2013). Specifically, I argue that the institutional context, i.e. whether a country has developed a flexicurity or dualizing labour market regime, influences actors’ preferences because some positions become more/less legitimate, functional and salient depending on this institution. For instance, in regimes with generous benefits, it is dysfunctional for social democratic parties to advocate increases in replacement rates; rather, they should focus on more salient issues such as the expansion of ALMPs to address increasing unemployment. Labour market institutions also influence preferences because the policies implemented are the reference points for political actors and shape their understandings of what constitutes a particular strategy (Immergut, 1998; Larsen, 2007). Put differently, when political actors in a dualizing country refer to activation, they have in mind a different combination of policies than do actors in flexicurity countries. Finally, institutions influence outcomes, which in turn affect preferences. This is especially the case for the quality of the jobs that jobseekers enrolled in ALMPs are likely to obtain. If these jobs are bound to be predominantly precarious, as is the case in dualizing countries, social democrats and unions might reject the expansion of measures that swiftly push workers back into uncertainty. In contrast, if the jobs are seen as adequate, pro-labour actors might be less opposed to such a policy. Concisely, I maintain that to better understand political actors’ preferences, we should analyse them as the product of a specific institutional constellation. My contribution to the literature is three-fold. First, I expand the analysis beyond the ‘usual suspects’ and study all political actors’ preferences and, thus, the political conflict structure that characterizes the labour market policy domain (unlike, e.g. Rueda, 2007; Tepe and Vanhuysse 2013). Thereby, the political conflict axes are identified by aggregating actors’ preferences and the saliency of specific policies; then, these axes are used to display the political space in which the actors are located (Kriesi et al., 2008; Häusermann, 2010). This procedure allows the mapping of actors’ preferences and determining their potential1 for value-based coalitions in two flexicurity countries (Denmark and Switzerland) and three dualizing countries (France, Germany and Italy). Secondly, I demonstrate the importance of context-specific analyses by showing that actors belonging to the same party family or actor group may indeed have different preferences in different regimes. Thirdly, I innovate by using a more direct measure of preferences consisting of interview data (Knoke et al., 1996) rather than inference based on government spending patterns (e.g. Bonoli, 2013; Tepe and Vanhuysse, 2013; Vlandas, 2013) and/or the preferences of the actors’ constituencies (Rueda, 2007). These data have the advantage of being less likely to be distorted because they capture value-based preferences as net of (strategic) politics and package deals (Immergut, 1998; Sabatier and Jenkins-Smith, 1999). The article proceeds as follows: in the theoretical section, I set out and describe the labour market policies and institutions to which political actors refer when forming their preferences. In the empirical analysis, I construct what amounts empirically to a two-dimensional labour market space onto which I map the political actors’ positions.2 Then, I analyse the potential for coalitions based on the similarity of actors’ preferences. The last section summarizes the findings and explores avenues for further research. 2. Preferences and conflict structure of traditional and activation policies Traditional measures to fight unemployment include passive benefits and EPL. The positions of political actors with respect to these measures are structured along the labour–capital antagonism, and they can be synthesized as an axis with one side concerning preferences for generous policies and the other side concerning support for lean policies (Korpi, 1983; Esping-Andersen, 1990; Kitschelt, 1994). Mainly, social democratic parties and unions have proposed policies to reduce social inequality by insuring blue-collar workers against traditional industrial risks (Esping-Andersen, 1990). Conversely, political actors on the right propose liberal market solutions to reduce state intervention and constrain universalistic and redistributive welfare state spending and labour market regulation. Although this conflict still structures the labour market domain, it is no longer the only one that does so—as has been shown for other policy fields (Kriesi et al., 2008; Häusermann, 2010; Fossati and Häusermann, 2014). Recently, traditional policies have proven to offer suboptimal protection for an increasing share of outsiders, including atypical and unemployed workers with low or obsolete skills (Bonoli, 2005; Rueda, 2007). Moreover, the adverse economic conditions, including the lower growth levels that are typical of post-industrial economies, and the pressure placed on welfare states by sociodemographic changes preclude the possibility of meeting outsiders’ needs by simply increasing decommodification efforts (Pierson, 1998, 2001). Consequently, governments had to seek alternative approaches, leading to what is known as an activation turn. In fact, over the last three decades, following a supranational consensus guided by organizations such as the European Union, with its European Employment Strategy (EES), a range of policies that actively promote labour market reintegration have been introduced across Europe and beyond (Torfing, 1999; Gilbert, 2002; Bonoli and Natali, 2012; Bonoli, 2013). Bonoli (2013) argues that ALMPs can be distinguished based on the degree to which they invest in human capital and the degree to which they ‘push’ the unemployed swiftly back into employment by means of conditionality (pro-market orientation) (cf. Eichhorst and Konle-Seidl, 2008; Fossati, 2017). In this regard, he distinguishes four types of ALMPs: (a) occupational policies, which are characterized by a low human capital orientation and low pro-market orientation (public jobs); (b) incentive reinforcement measures, which have a strong pro-market orientation but no human capital component (sanctioning and incentives); (c) upskilling instruments, which have a strong human capital orientation and a strong pro-employment orientation (job-related training); and (d) employment assistance measures, which are characterized by a medium pro-market component and a weak human capital component (placement services, counselling). Generally, a combination of these strategies is applied. Empirically, however, some countries focus more on human capital investment (traditionally, social democratic countries), while others stress swift re-employment (traditionally, liberal countries). The interesting question is thus how this institutional variation influences political actors’ preferences. 3. How labour market regimes affect actors’ preferences I argue that the institutional legacies that place countries within a specific labour market regime differentially influence actors’ (activation) preferences3; thereby, three mechanisms apply. First, as argued by Steinmo (1989, p. 502), the ‘… institutional framework provides the context (emphasis added) in which groups and individuals interpret their self-interest and thereby define their policy preferences’. Thus, preferences (and values) have only context-specific substantial meanings (Steinmo, 1989, p. 502; Immergut, 1998). In other words, preferences for/against a specific measure are reactions to policies that have already been implemented. For instance, liberal parties should be more supportive of EPL liberalization in dualized countries than in flexicurity countries, where regulations are already very lenient. However, if we ignored institutional differences and considered the extent to which liberal parties agree with additional EPL liberalization and found that they ‘strongly agree’ in dualizing countries but just ‘agree’ in flexicurity countries, we could wrongly assume that these parties take a more moderate stance in flexicurity countries. In a similar vein, the labour market characteristics of a regime (de-)legitimize or make specific preferences more (less) salient. In fact, political actors can be expected to focus on social problems or on policies that need reform rather than engaging in issues that are already sufficiently addressed, such as labour market liberalization in flexicurity or insider protection in dualizing countries. Secondly, institutions provide anchor points and frames of reference for political actors’ policy definitions. They interpret a strategy based on what is implemented in their country and ponder whether adding a particular policy matches, reinforces or contradicts the pre-existing framework. For instance, rational employer organizations should have different ALMP preferences in flexicurity countries, where ALMPs provide a larger pool of well-qualified candidates, whereas in dualizing countries, non-upskilled unemployed are merely pushed back into the job market. Thereby, the institutional effect unfolds also based on the specific framing of policies, i.e. the way in which problems and solutions are assessed and communicated (Entman, 1993; Larsen, 2007; see Wüest and Fossati (2014) for a framing analysis of labour market discourse). In liberal countries, unemployment is generally framed as a consequence of behavioural shortcomings (Daguerre, 2007). In such a context, it is plausibly more difficult to support human capital investments than negative incentives (e.g. conditionality). In traditionally social democratic countries, unemployment is often considered to have structural causes, such as a mismatch between demand and supply, which easily explains skill-based policy preferences and responses (Daguerre, 2007). The unstable working conditions and the high unemployment risk of outsiders in dualizing countries, instead are likely to push political actors to defend activation measures catering insiders rather than investing in outsider’s human capital that might be lost again in the event of a future layoff. Finally, the institutionally pre-structured social outcomes likely affect actors’ preferences (Steinmo, 1989). In the labour market domain, the quality of jobs that unemployed obtain matters. If a well-intentioned policy produces suboptimal outcomes and results in precarious jobs, political actors on the left may distance themselves from such measures even if these would correspond to their ideological roots. Concisely, following Lowi’s (1972) logic, I propose that policies shape preferences and that— although this is beyond the scope of the present article—preferences matter because they are at the core of politics. 4. The characteristics of flexicurity and dualizing labour market regimes To examine the effect of the institutional context on political actors’ preferences, I focus on Denmark and Switzerland as representative flexicurity4 countries and France, Germany and Italy as representatives dualizing countries.5 As suggested by Thelen (2012), this theoretical distinction is more precise and less static than the welfare regimes or varieties of capitalism approaches (Esping-Andersen, 1990; Hall and Soskice, 2001). In fact, it accounts for reactions to the liberalization pressures by distinguishing among countries that try to preserve social equality and those that respond with dualization or even deregulation (e.g. the UK, the USA). I carefully chose the cases in order to maximize comparability within a regime by grouping countries that implemented similar policies (Table 1). Table 1. Active and passive (traditional) labour market policy efforts   ALMPs   Traditional labour market policies     Spending on training as a percentage of GDPa  Spending on ALMPs as a percentage of GDP per percentage of unemploymentb  ALMP spending per unemployed personc  Automatic ALMP spending adjustmentd  Net unemployment benefit replacement rate for a single persone  Net unemployment benefit replacement rate for one- earner couple with two childrene  Flexicurity              Switzerland  0.20  0.18  2.3  Yes  0.71  0.83  Denmark  0.66  0.33  4.2  Yes  0.55  0.62  Dualizing              Germany  0.27  0.09  1.8  No  0.60  0.72  France  0.37  0.11  1.5  No  0.69  0.70  Italy  0.15  0.06  0.5  No  0.63  0.73    ALMPs   Traditional labour market policies     Spending on training as a percentage of GDPa  Spending on ALMPs as a percentage of GDP per percentage of unemploymentb  ALMP spending per unemployed personc  Automatic ALMP spending adjustmentd  Net unemployment benefit replacement rate for a single persone  Net unemployment benefit replacement rate for one- earner couple with two childrene  Flexicurity              Switzerland  0.20  0.18  2.3  Yes  0.71  0.83  Denmark  0.66  0.33  4.2  Yes  0.55  0.62  Dualizing              Germany  0.27  0.09  1.8  No  0.60  0.72  France  0.37  0.11  1.5  No  0.69  0.70  Italy  0.15  0.06  0.5  No  0.63  0.73  Sources: a OECD (2016a), figures for 2010; b Bonoli (2013, p. 30), figures for 2007; c Tepe and Vanhuysse (2013); d Duell et al. (2010); e van Vliet and Caminada (2012), figures for 2009. Table 1. Active and passive (traditional) labour market policy efforts   ALMPs   Traditional labour market policies     Spending on training as a percentage of GDPa  Spending on ALMPs as a percentage of GDP per percentage of unemploymentb  ALMP spending per unemployed personc  Automatic ALMP spending adjustmentd  Net unemployment benefit replacement rate for a single persone  Net unemployment benefit replacement rate for one- earner couple with two childrene  Flexicurity              Switzerland  0.20  0.18  2.3  Yes  0.71  0.83  Denmark  0.66  0.33  4.2  Yes  0.55  0.62  Dualizing              Germany  0.27  0.09  1.8  No  0.60  0.72  France  0.37  0.11  1.5  No  0.69  0.70  Italy  0.15  0.06  0.5  No  0.63  0.73    ALMPs   Traditional labour market policies     Spending on training as a percentage of GDPa  Spending on ALMPs as a percentage of GDP per percentage of unemploymentb  ALMP spending per unemployed personc  Automatic ALMP spending adjustmentd  Net unemployment benefit replacement rate for a single persone  Net unemployment benefit replacement rate for one- earner couple with two childrene  Flexicurity              Switzerland  0.20  0.18  2.3  Yes  0.71  0.83  Denmark  0.66  0.33  4.2  Yes  0.55  0.62  Dualizing              Germany  0.27  0.09  1.8  No  0.60  0.72  France  0.37  0.11  1.5  No  0.69  0.70  Italy  0.15  0.06  0.5  No  0.63  0.73  Sources: a OECD (2016a), figures for 2010; b Bonoli (2013, p. 30), figures for 2007; c Tepe and Vanhuysse (2013); d Duell et al. (2010); e van Vliet and Caminada (2012), figures for 2009. Classifying Switzerland and Denmark as flexicurity countries is legitimate because they are similar in terms of traditional and—in particular—activation measures (Viebrock and Clasen 2009, Bonoli, 2013; see Shahidiet al. (2016) for a similar typology). First, as shown in Table 1, in the domain of activation policies, investment in ALMPs as a share of gross domestic product (GDP) for each percentage point of unemployment is very high at 0.18 and 0.33, respectively (Bonoli, 2013; Tepe and Vanhuysse, 2013). Moreover, the main pillars of successful activation consist of human capital rehabilitation and strict demand for a proactive employment search (Gerfin and Lechner 2002; Madsen, 2002; Duell et al., 2010; Knotz, 2015). Unsurprisingly, in Switzerland, 38% of jobseekers participated in training measures in 2009 (Duell et al., 2010; cf. Gerfin and Lechner, 2002). Finally, these are the only countries where ALMP expenditures increase automatically with unemployment levels (Duell et al., 2010). In other words, these countries follow a human capital fostering activation strategy (Barbier and Ludwig-Mayerhofer, 2004; Bonoli, 2013). With respect to traditional labour market policies, both countries have extremely flexible labour markets and loose EPL (Madsen, 2002; Viebrock and Clasen 2009; Clasen and Clegg, 2011). The EPL index produced by the Organisation for Economic Co-operation and Development (OECD) for all categories of workers amounts to merely 2.1 points for Denmark and 1.6 point for Switzerland in 2010 (the OECD average is 2.4 of 6 points). In contrast to dualizing countries, this flexibility also applies to workers with regular contracts (cf. Venn, 2009, p. 8; OECD, 2013; Svalund, 2013). A strong social partnership enables this level of flexibility because it compensates all workers with solid protection (Wilthagen, 2002), providing universal and generous unemployment coverage (columns 4 and 5). In 2009, the replacement rate (as a percentage of the previous wage) of a one-earner couple with two children in Switzerland amounted to 82% and to 73% in Denmark (van Vliet and Caminada, 2012). Secondly, I analyse Germany, France and Italy as representative dualization regimes (Palier, 2010). As shown in Table 1, in these countries, there is conspicuously less ALMP effort than in flexicurity countries, and it is characterized mainly by occupational strategies,6 i.e. pushing the employed back into the labour market or placing them in short-time work whilst failing to systematically invest in upskilling (Torfing, 1999; Barbier and Ludwig-Mayerhofer, 2004; Daguerre, 2007; Clasen and Clegg, 2011; Bonoli, 2013, p. 93ff.). In terms of traditional policies, dualizing countries are characterized by the existence of two parallel labour markets (Lindbeck and Snower, 1988; Emmenegger et al., 2012). Strong EPL regulates the primary labour market and creates favourable employment conditions. In contrast, the secondary labour market follows purely economic reasoning and offers precarious employment, loose hire-and-fire legislation and marginal—if any—welfare benefits (Venn, 2009; Palier, 2010). In these countries, precariousness is a serious problem (Berton et al. 2009; Emmenegger et al., 2012). For instance, there is a high incidence (approximately 49% in Italy and 20% in Germany) of non-voluntarily part-time work (compared to 7% in Switzerland and 15% in Denmark) (OECD, 2016b). Moreover, welfare provision is generous for insiders; however, the existence of two segregated labour markets and social support systems undermines attempts to reduce social inequality in these countries compared with flexicurity countries (Berton et al., 2009; Palier, 2010; Thelen, 2012). In sum, Switzerland and Denmark are flexibility countries because both respond to the challenges of liberalization by preserving generous and encompassing passive benefits, granting flexible labour markets and addressing labour market challenges through human capital-focused ALMPs. I compare them with three dualizing countries with generous but not encompassing passive benefits. Strictly regulated labour markets are at their core, but there is unpredictability at the margin, and they lag behind in terms of activation by implementing—if at all—pro-employment-focused ALMPs (Thelen, 2012; see also Rueda, 2007; Eichhorst and Konle-Seidl, 2008; Viebrock and Clasen 2009; Clasen and Clegg, 2011; Bonoli, 2013; Shahidi et al., 2016). In the next section, I formulate hypotheses about the effects of these institutional characteristics on political actors’ preferences, which in turn, constitute the basis for political actors’ coalition potential. 5. Hypotheses: political actors’ preferences in flexicurity and dualizing regimes As previously argued, it is important to keep in mind that actors’ preferences are shaped by their institutional frames of reference, i.e. by the strategies that are characteristic of a regime, the problems that are salient or insufficiently addressed, and the anticipation of job market-related outcomes concerning their constituencies. However, as I formulate expectations for very different actor types, I take into account additional theoretical considerations such as the actors’ ideological roots, i.e. stable traits that represent their core policy beliefs (Sabatier and Jenkins-Smith, 1999), and the interests of the constituency to which they cater (e.g. Rueda, 2007; Häusermann, 2010). Communist parties traditionally take a pro-labour stance with respect to labour–capital antagonism. Following Rueda (2005), I generally expect these organizations to focus on expanding traditional policies rather than on asking workers to adapt to the demands of the post-industrial labour market by taking part in ALMPs in any circumstance. In terms of regime differences, communist parties should be particularly sceptical of activation in dualizing countries because in austere times and with increasing flexibilization pressure, the depletion of insider privileges is looming. Moreover, in these countries, ALMPs do not aim at upskilling or, thus, at improving the labour market chances of the unemployed worker; rather, they push them back into precarious jobs. In dualizing countries, communist parties should focus on preserving the privileges of their traditional constituency, contemporaneously expanding these passive benefits to outsiders where possible. Conversely, I expect communists to favour activation in flexicurity countries because jobs are less precarious and ALMPs focus on upskilling. From a purely ideological perspective, social democratic parties should favour the expansion of both traditional measures and ALMPs to address the needs of both their traditional clientele (blue-collar workers) and contemporary vulnerable workers (outsiders). However, as argued in insider–outsider theory, policy legacies and austerity push social democratic parties to make strategic choices (Rueda, 2007). Consequently, I expect that in flexicurity regimes, social democrats advocate for the expansion of training-based ALMPs because both insiders and outsiders benefit from such policies and because insiders’ traditional interests are adequately accommodated (Emmenegger, 2009). Conversely, in dualized countries, traditional protections for insiders are highly endangered, and the interests of insiders and outsiders diverge because of strong EPL, insider fear of wage competition by outsiders, and precarious job expectations (Saint-Paul, 1998). Thus, social democrats support traditional policies to accommodate their unionized and more politically active core constituency rather than increasing activation effort (Bonoli, 2005; Rueda, 2007). I expect green parties to hold preferences similar to those of social democratic parties on the traditional conflict dimension (Kitschelt, 1994). The main difference is that the constituencies of green parties are younger and generally work in high-skilled occupations (Oesch, 2006; Geering and Häusermann, 2013). Consequently, these individuals are less likely to become unemployed, and if so, they are interested in re-entering the labour market swiftly. I thus expect green parties to favour the expansion of ALMPs, particularly in flexicurity countries where they benefit from upskilling. Liberal parties are mainly elected by highly skilled business-friendly individuals who favour strong market logic (Oesch, 2006; Häusermann, 2010). These parties focus on reducing spending on traditional policies to decrease their constituents’ tax burden and to increase the competitiveness of the labour market (Rueda, 2005). Whilst they are generally less likely to be affected by unemployment than the constituencies of leftist parties (Oesch, 2006), liberal parties might favour ALMPs as means to increase the employability of the workforce, especially in flexicurity countries where ALMPs are human capital centred. In contrast, in dualizing countries, liberal party support for activation should be weaker because occupational measures do not increase the expected productivity of workers. Right-wing and conservative parties are characterized by preferences for subsidiarity and self-reliance; hence, they are likely to reject the expansion of traditional policies (Huber and Stephens, 2001; Myles and Quadagno, 2002). As argued by Rueda (2007), their constituency comprises upscale groups such as (upper) middle class voters, business-friendly individuals, and employers. Particularly in flexicurity countries, they should be sceptical of increasing already substantial ALMP efforts and more likely to advocate for retrenchment on this axis. In contrast, in dualizing countries, they might be more receptive to increasing activation to foster the swift reintegration of the unemployed and, thus, decrease public deficits and increase the supply of cheap labour (outsiders). I expect that populist parties, which traditionally represent blue-collar workers (insiders), to hold the business-friendly positions of right-wing parties because by only advocating mainstream labour market policy positions are these organizations able to participate in government whilst remaining faithful to their extremist immigration policies (their core issue) (cf. Afonso, 2015). In other words, to pursue their office-seeking strategy, they adapt their preferences to those of the conservative mainstream. In terms of ideology, like social democratic parties, unions should advocate for the expansion of effort on both dimensions and thus fulfil their role as representatives of all workers (Esping-Andersen, 1990). The expectation differs if we lend credit to insider–-outsider theory, which equates union and social democratic preferences and ties them to the institutional context (Rueda, 2007; Gordon, 2015). According to this theory, unions should focus on the core interests of their members and prefer the expansion of traditional policies, particularly in dualized countries with strong ELP (Arnd, 2013). Conversely, they should be supportive of ALMPs expansion in flexicurity countries, where the preferences of insiders and outsiders converge owing to the increased unemployment risk of insiders (Saint-Paul, 1998; Emmenegger, 2009). With respect to traditional policies, employer organizations prioritize budgetary rigour over welfare expenditures (Esping-Andersen, 1990). Moreover, in terms of business interests, these organizations support ALMPs if they increase the employability of jobseekers and thus the pool of well-qualified candidates (Swank and Martin, 2001). Their support for expanding ALMPs should be strongest in flexicurity countries, where the benefits of human capital intervention and gains in productivity are highest. In contrast, in dualizing countries, where the policy mix is oriented towards occupational strategies, the expansion of ALMPs does not benefit employer organizations to the same extent, which is likely to reduce their support for such an expansion. State bodies and administrations should generally support the expansion of activation because they are in close contact and are thus influenced by supranational consensus (e.g. EES) that combines a neoliberal perspective on budgetary balance aimed at decreasing spending on traditional policies and the idea that activation is essential to reducing unemployment (Daguerre and Taylor-Gooby, 2004; Fleckenstein, 2008; Stiller and van Gerven, 2012). I expect that state bodies and administrations in dualizing countries that do not (yet) meet EES recommendations support the expansion of ALMPs more strongly than flexicurity countries where well-functioning ALMP schemes may lead a ceiling on support. Summarizing the hypotheses regarding activation preferences, in flexicurity countries, I expect social democratic, green and liberal parties, unions and employer organizations to favour increasing ALMP effort and right-wing parties to favour decreasing ALMP effort. In dualizing countries, I expect state bodies, administrations and right-wing parties to support the expansion of activation, whereas social democrats, communist parties and unions should prefer a decrease in activation. Finally, the expectations for social movement organizations (SMOs) are unclear because they depend on the nature of the initiative; therefore, their positioning is an empirical question. 6. Operationalization and methods To analyse political actors’ labour market policy preferences, I rely on novel interview data collected in Denmark, France, Germany, Italy and Switzerland in the autumn of 2010. The data set is based on semi-structured telephone interviews with major policymakers7 active in this domain, i.e. parties, unions, employer organizations, state bodies, administrations and SMOs.8 In the advocacy coalition and network analysis literature, the notion is that in the policymaking process, all experts in the policy field should be analysed to determine the potential for coalitions in terms of overlapping preferences (Knoke et al., 1996; Sabatier and Weible, 2007). Whilst the relevance of parties, social partners and state bodies are uncontested, that of SMOs is less straightforward. I argue that these organizations, although they have less political leverage, still play an important role during consultation processes as providers of external expertise and as representatives of labour market outsiders who are not usually unionized and who would not be represented otherwise. Moreover, they intervene as agenda setters or as co-organizers of demonstrations to raise awareness of neglected social problems (Knoke et al., 1996). The interviews were held between October and December 2010. It goes without saying that this period was characterized by a deep economic crisis. The advantage of analysing labour market policy preferences in this setting is that labour market issues were on the daily agenda and that political actors had strong incentives to take well-considered positions that were in line with their constituencies’ interests, thus rendering position shifts difficult to justify. Elite interview data are ideally suited to the analysis of political actors’ preferences because they allow their assessment independently of possibly idiosyncratic, very specific and/or limited political reform processes (e.g. Häusermann, 2010). Moreover, the precision of the questions enables the accurate operationalization of different ALMP types, whereas data that rely on extremely broad categories, such as the Manifesto Data Collection, do not. A strength of these data is that it captures both political actors’ preferences (a position measure) and their perceptions of the relative importance of particular policy measures (salience measure9). This allows for the construction of an indicator that weights an actor’s position10 by the salience of a measure.11 As shown in Table 2, I operationalize traditional conflict by means of ‘raising the minimum wage’, which is a state-led intervention to guarantee a decent living standard for workers, and ‘unemployment benefit reduction’, which captures benefit retrenchment. Two items operationalizing preferences for lower EPL (‘loosening hire-and-fire legislation’ and ‘increasing the flexibility of working hours’) are also included. Table 2. List of indicators Minimum wage  Raising the minimum wage.  Benefit reduction  Reducing unemployment benefits.  Working hours  Increasing the flexibility of working hours.  Hire-fire  Loosening hire-and-fire legislation.  Sanction  Imposing tougher sanctions for those who refuse to accept work that is deemed appropriate for them.  State job  Using state programmes to create jobs.  Training  Increasing retraining possibilities for the unemployed.  Short-time  Promoting short-time work—the ability of employers to reduce workers’ hours when orders are low.  Reintegration  Actively promoting reintegration into the labour market.  Minimum wage  Raising the minimum wage.  Benefit reduction  Reducing unemployment benefits.  Working hours  Increasing the flexibility of working hours.  Hire-fire  Loosening hire-and-fire legislation.  Sanction  Imposing tougher sanctions for those who refuse to accept work that is deemed appropriate for them.  State job  Using state programmes to create jobs.  Training  Increasing retraining possibilities for the unemployed.  Short-time  Promoting short-time work—the ability of employers to reduce workers’ hours when orders are low.  Reintegration  Actively promoting reintegration into the labour market.  Table 2. List of indicators Minimum wage  Raising the minimum wage.  Benefit reduction  Reducing unemployment benefits.  Working hours  Increasing the flexibility of working hours.  Hire-fire  Loosening hire-and-fire legislation.  Sanction  Imposing tougher sanctions for those who refuse to accept work that is deemed appropriate for them.  State job  Using state programmes to create jobs.  Training  Increasing retraining possibilities for the unemployed.  Short-time  Promoting short-time work—the ability of employers to reduce workers’ hours when orders are low.  Reintegration  Actively promoting reintegration into the labour market.  Minimum wage  Raising the minimum wage.  Benefit reduction  Reducing unemployment benefits.  Working hours  Increasing the flexibility of working hours.  Hire-fire  Loosening hire-and-fire legislation.  Sanction  Imposing tougher sanctions for those who refuse to accept work that is deemed appropriate for them.  State job  Using state programmes to create jobs.  Training  Increasing retraining possibilities for the unemployed.  Short-time  Promoting short-time work—the ability of employers to reduce workers’ hours when orders are low.  Reintegration  Actively promoting reintegration into the labour market.  Following Bonoli’s (2013) definition, I capture occupational ALMPs by means of ‘the use of state programmes to create jobs’ and in terms of subsidies to employers to employ individuals through a ‘promotion of short-time work’ (Sacchi et al., 2011). Secondly, incentive reinforcement is operationalized as ‘increasing sanctions when an unemployed person refuses a job, which is deemed appropriate’ (Trickey and Walker, 2001; Kemmerling and Bruttel, 2006; Clasen and Clegg, 2011; Fossati, 2017). I measure attitudes towards upskilling as ‘increasing of training effort’ and employment assistance by means of the ‘increasing reintegration effort’ measure. To analyse and display preferences, conflict dimensions and coalition potential in an accessible way, I conduct a regime-specific factor analysis (varimax rotation) on the nine indicators presented in Table 2. This analysis provides insight into whether there is regime-specific variation in the conflict dimensions that are identifiable in terms of the strength and direction of the factor loadings. Then, I retain the individual actors’ factor scores in order to locate them within a multidimensional policy space (Figures 1 and 2). Figure 1. View largeDownload slide Actor configuration in the flexicurity countries (Denmark and Switzerland). Denmark: Parties: SD Social Democrat Party, Venstre Liberal Party, DF Dansk Folkepartis, KF Konservative Folketsparti, RG Red-Green Alliance, SF Socialist Folkeparti; Unions: AC Akademikernes Centralorganisation, FTF Confederation of Professionals, LO Confederation of Trade Unions; Employer organisations: DA Confederation of Danish Employers, DI Confederation of Danish Industry; Administrations: NLMA National Labour Market Authority, DEC Economic Council. Social movementorganisations: SFI Danish national centre for social research; ECLM Economic Council of labour movement (AE), CEPOS Conservative think-tank. Switzerland: Parties: Gruene Grüne Partei Schweiz, FDP Freisinning Demorkatische Partei, die Liberalen, SVP Schweizerische Volkspartei, CVP Christlichdemokratische Volkspartei, BDP Bürgerlich Demokratische Partei, SP Sozialdemokratische Partei; Unions: Unia Unia, KV Kaufmännischer Verband Schweiz, Syna Syna Arbeitslosen Kasse, AS Angestellte Schweiz, SGB Scherizerischer Gewerkschaftsbund, TS Travail.Suisse, Gewerkschaftsdachorganisation; Employer organisations: SBV Dachverband Schweizerischer Baumeisterverband, SGV Schweizerischer Gewerbeverband, SAV Schweizerischer, Arbeitgeberverband, Swissmem Swissmem; Administrations: BE Canton Bern, AG Canton Argau, SODK Conference of the Cantonal Social Ministers, SECO State Secretary for Economic Affairs. Social movement organisations: Caritas Caritas Switzerland; Attac Attac Switzerland; Kabba Kabba SMO on behalf of the unemployed, AvS Avenir Suisse. Figure 1. View largeDownload slide Actor configuration in the flexicurity countries (Denmark and Switzerland). Denmark: Parties: SD Social Democrat Party, Venstre Liberal Party, DF Dansk Folkepartis, KF Konservative Folketsparti, RG Red-Green Alliance, SF Socialist Folkeparti; Unions: AC Akademikernes Centralorganisation, FTF Confederation of Professionals, LO Confederation of Trade Unions; Employer organisations: DA Confederation of Danish Employers, DI Confederation of Danish Industry; Administrations: NLMA National Labour Market Authority, DEC Economic Council. Social movementorganisations: SFI Danish national centre for social research; ECLM Economic Council of labour movement (AE), CEPOS Conservative think-tank. Switzerland: Parties: Gruene Grüne Partei Schweiz, FDP Freisinning Demorkatische Partei, die Liberalen, SVP Schweizerische Volkspartei, CVP Christlichdemokratische Volkspartei, BDP Bürgerlich Demokratische Partei, SP Sozialdemokratische Partei; Unions: Unia Unia, KV Kaufmännischer Verband Schweiz, Syna Syna Arbeitslosen Kasse, AS Angestellte Schweiz, SGB Scherizerischer Gewerkschaftsbund, TS Travail.Suisse, Gewerkschaftsdachorganisation; Employer organisations: SBV Dachverband Schweizerischer Baumeisterverband, SGV Schweizerischer Gewerbeverband, SAV Schweizerischer, Arbeitgeberverband, Swissmem Swissmem; Administrations: BE Canton Bern, AG Canton Argau, SODK Conference of the Cantonal Social Ministers, SECO State Secretary for Economic Affairs. Social movement organisations: Caritas Caritas Switzerland; Attac Attac Switzerland; Kabba Kabba SMO on behalf of the unemployed, AvS Avenir Suisse. Figure 2. View largeDownload slide Actor configuration in dualizing countries (Germany, France and Italy). France: Parties: UMP Union pour le Mouvement Populaire (UMP), PS Parti Socialiste, FN Front National, PCF Parti Communiste Français, LO Lutte Ouvrière, NPA Nouveau Parti Anticapitaliste, FO Force Ouvrière, EELV Europe Ecologie Les Verts; Unions: Solidaires Solidaires, UNSA Union nationale des Syndicats Autonomes, CFECGC Confédération Française de l'Encadrement, CFDT Confédération Française Démocratique du Travail, CFTC Confédération Française des Travailleurs, CGT Confédération Générale du Travail, FSU Fédération Syndicale Unitaire; Employer organisations: CGPME Confédération Générale des Petites et Moyennes Entreprises, MEDEF Mouvement des Entreprises de France, UNAPL Confédération Interprofessionnelle des Professions Libérales, UPA Union Professionnelle Artisanale; Administrations: MDT Ministère du Travail. Social movement organisations: AC! Agir contre le Chômage!; APEIS Association Pour l’Emploi, l’Inofrmation et la Solidarité; CNPE Comité national des privés d'emploi CGT; MNCP Mouvement National des Chômeurs et Précaires; SNC Solidarité nouvelles face au chômage. Germany: Parties: Linke Die Linke, CDU/CSU Christlich Demokratische Union/Christlich Soziale Union, NPD Nationaldemokratische Partei Deutschlands, FDP Freie Demokratische Partei/Die Liberalen, SPD Sozialdemokratische Partei Deutschlands, Gruene Grüne Partei Deutschland; Unions: KGA Koordinierungsstelle Gewerkschaftlicher Arbeitslosengruppen, IGM Industriegewerkschaft Metall, Verdi Vereinte Dienstleistungsgewerkschaft, DGB Deutscher Gewerkschaftsbund; Employer organisations: BDA Bundesvereinigung der Deutschen Arbeitgeberverbände; Administrations: BKS Bundesvereinigung Kommunale Spitzenverbände, BMAS Bundesministerium für Arbeit und Soziales, BWM Bundeswirtschaftsministerium, BA Bundesagentur für Arbeit. Social movement organisations: Caritas Caritas Germany; EFD Erwerbslosenforum; INSM Initiative Neue Soziale Marktwirtschaft; PW Paritätischer Wohlfahrtsverband; Bertesl Bertelsmannstiftung; IAB Institut für Arbeitsmarkt und Berufsforschung. Italy: Parties: PD Partito Democratico, SEL Sinistra Ecologia e Libertà, UdC Unione di Centro, PdL Popolo della Libertà, Lega Lega Nord, IdV Italia dei Valori, PdCI Partito dei Comunisti Italiani, PRC Rifondazione Comunista;Unions: COBAS Confederazione dei Comitati di Base, UIL Unione Italiana del Lavoro, CISL Confederazione Italiana Sindacati dei Lavoratori, CIGL Confederazione Generale Italiana del Lavoro; Employer organisations: Confindustria Confindustria; Administrations: MdL Ministero del Lavoro, INPS Istituto Nazionale Previdenza Sociale. Social movement organisations: ISFOL Istituto Sviluppo della Formazione Professionale dei Lavoratori; ACLI Associazioni Cristiane Lavoratori Italiani (Patronato CISL); ALVP Movimento Associazione Lavoratori Vittime del Precariato; INCA Istituto Nazionale Confederale di Assistenza (patronato CIGL); ARCI Associazione di Promozione Culturale; RdC Rete della Conoscenza. Figure 2. View largeDownload slide Actor configuration in dualizing countries (Germany, France and Italy). France: Parties: UMP Union pour le Mouvement Populaire (UMP), PS Parti Socialiste, FN Front National, PCF Parti Communiste Français, LO Lutte Ouvrière, NPA Nouveau Parti Anticapitaliste, FO Force Ouvrière, EELV Europe Ecologie Les Verts; Unions: Solidaires Solidaires, UNSA Union nationale des Syndicats Autonomes, CFECGC Confédération Française de l'Encadrement, CFDT Confédération Française Démocratique du Travail, CFTC Confédération Française des Travailleurs, CGT Confédération Générale du Travail, FSU Fédération Syndicale Unitaire; Employer organisations: CGPME Confédération Générale des Petites et Moyennes Entreprises, MEDEF Mouvement des Entreprises de France, UNAPL Confédération Interprofessionnelle des Professions Libérales, UPA Union Professionnelle Artisanale; Administrations: MDT Ministère du Travail. Social movement organisations: AC! Agir contre le Chômage!; APEIS Association Pour l’Emploi, l’Inofrmation et la Solidarité; CNPE Comité national des privés d'emploi CGT; MNCP Mouvement National des Chômeurs et Précaires; SNC Solidarité nouvelles face au chômage. Germany: Parties: Linke Die Linke, CDU/CSU Christlich Demokratische Union/Christlich Soziale Union, NPD Nationaldemokratische Partei Deutschlands, FDP Freie Demokratische Partei/Die Liberalen, SPD Sozialdemokratische Partei Deutschlands, Gruene Grüne Partei Deutschland; Unions: KGA Koordinierungsstelle Gewerkschaftlicher Arbeitslosengruppen, IGM Industriegewerkschaft Metall, Verdi Vereinte Dienstleistungsgewerkschaft, DGB Deutscher Gewerkschaftsbund; Employer organisations: BDA Bundesvereinigung der Deutschen Arbeitgeberverbände; Administrations: BKS Bundesvereinigung Kommunale Spitzenverbände, BMAS Bundesministerium für Arbeit und Soziales, BWM Bundeswirtschaftsministerium, BA Bundesagentur für Arbeit. Social movement organisations: Caritas Caritas Germany; EFD Erwerbslosenforum; INSM Initiative Neue Soziale Marktwirtschaft; PW Paritätischer Wohlfahrtsverband; Bertesl Bertelsmannstiftung; IAB Institut für Arbeitsmarkt und Berufsforschung. Italy: Parties: PD Partito Democratico, SEL Sinistra Ecologia e Libertà, UdC Unione di Centro, PdL Popolo della Libertà, Lega Lega Nord, IdV Italia dei Valori, PdCI Partito dei Comunisti Italiani, PRC Rifondazione Comunista;Unions: COBAS Confederazione dei Comitati di Base, UIL Unione Italiana del Lavoro, CISL Confederazione Italiana Sindacati dei Lavoratori, CIGL Confederazione Generale Italiana del Lavoro; Employer organisations: Confindustria Confindustria; Administrations: MdL Ministero del Lavoro, INPS Istituto Nazionale Previdenza Sociale. Social movement organisations: ISFOL Istituto Sviluppo della Formazione Professionale dei Lavoratori; ACLI Associazioni Cristiane Lavoratori Italiani (Patronato CISL); ALVP Movimento Associazione Lavoratori Vittime del Precariato; INCA Istituto Nazionale Confederale di Assistenza (patronato CIGL); ARCI Associazione di Promozione Culturale; RdC Rete della Conoscenza. The reliability of the actor constellations is assessed by means of three robustness checks. First, I replicate the analyses using only parties, social partners and state bodies (excluding SMOs). Moreover, the models are re-estimated without the issue salience weightings and for each country separately. These additional analyses show that the actor constellations are stable.12 7. The policy space: context-specific traditional and activation preferences The results of the factor analysis presented in Table 313 show that in both regimes, the policy space is two-dimensional and mirrors support/rejection of the two main strategies to fight unemployment, namely, traditional and activation strategies. Table 3. Traditional and activation factors by regime   Flexicurity   Dualized   Items  Denmark and Switzerland   Germany†, France and Italy     Traditional  Activation  Traditional  Activation  Sanction  0.52  −0.23  0.58  −0.09  Benefit reduction  0.71  −0.07  0.51  −0.26  Working hours  0.46  −0.33  0.71  −0.13  Hire-fire  0.46  0.03  0.55  −0.16  Minimum wage  −0.53  0.22  −0.76  −0.05  State job  −0.31  0.61  −0.72  −0.02  Training  −0.11  0.48  −0.30  0.40  Short-time  0.00  0.59  0.20  0.35  Reintegration  −0.09  0.22  −0.02  0.66  Eigenvalue  2.12  1.21  2.71  1.95  N  40  40  67  67    Flexicurity   Dualized   Items  Denmark and Switzerland   Germany†, France and Italy     Traditional  Activation  Traditional  Activation  Sanction  0.52  −0.23  0.58  −0.09  Benefit reduction  0.71  −0.07  0.51  −0.26  Working hours  0.46  −0.33  0.71  −0.13  Hire-fire  0.46  0.03  0.55  −0.16  Minimum wage  −0.53  0.22  −0.76  −0.05  State job  −0.31  0.61  −0.72  −0.02  Training  −0.11  0.48  −0.30  0.40  Short-time  0.00  0.59  0.20  0.35  Reintegration  −0.09  0.22  −0.02  0.66  Eigenvalue  2.12  1.21  2.71  1.95  N  40  40  67  67  † Attac Germany was excluded from the sample. Factor loadings in bold highlight correlations of the indicators and the factor, which are higher than the minimum correlation of 0.35. Table 3. Traditional and activation factors by regime   Flexicurity   Dualized   Items  Denmark and Switzerland   Germany†, France and Italy     Traditional  Activation  Traditional  Activation  Sanction  0.52  −0.23  0.58  −0.09  Benefit reduction  0.71  −0.07  0.51  −0.26  Working hours  0.46  −0.33  0.71  −0.13  Hire-fire  0.46  0.03  0.55  −0.16  Minimum wage  −0.53  0.22  −0.76  −0.05  State job  −0.31  0.61  −0.72  −0.02  Training  −0.11  0.48  −0.30  0.40  Short-time  0.00  0.59  0.20  0.35  Reintegration  −0.09  0.22  −0.02  0.66  Eigenvalue  2.12  1.21  2.71  1.95  N  40  40  67  67    Flexicurity   Dualized   Items  Denmark and Switzerland   Germany†, France and Italy     Traditional  Activation  Traditional  Activation  Sanction  0.52  −0.23  0.58  −0.09  Benefit reduction  0.71  −0.07  0.51  −0.26  Working hours  0.46  −0.33  0.71  −0.13  Hire-fire  0.46  0.03  0.55  −0.16  Minimum wage  −0.53  0.22  −0.76  −0.05  State job  −0.31  0.61  −0.72  −0.02  Training  −0.11  0.48  −0.30  0.40  Short-time  0.00  0.59  0.20  0.35  Reintegration  −0.09  0.22  −0.02  0.66  Eigenvalue  2.12  1.21  2.71  1.95  N  40  40  67  67  † Attac Germany was excluded from the sample. Factor loadings in bold highlight correlations of the indicators and the factor, which are higher than the minimum correlation of 0.35. As expected, the results show that political actors interpret activation14 differently depending on the institutional context (see the factor loadings). In the flexicurity countries, the ‘creation of public job[s]’ is associated with activation because the state is strongly engaged in creating re-employment opportunities, whereas in the dualizing countries, state efforts to create jobs remain predominantly connected with the traditional dimension, as this is a leftist strategy to reduce unemployment in the southern welfare states (Esping-Andersen, 1990). The results also indicate that the activation dimension in flexicurity countries is determined foremost by policies that highlight the active role of the state in creating job and training opportunities. Somewhat surprisingly, short-time work also resonates with activation in flexicurity countries. However, this finding can be explained because such work is a way to preserve workers’ skills by retaining them in firms and is accordingly in line with the human capital approach. The loading of ‘training’ is comparatively low because of the broad agreement among almost all actors on this proposition. Briefly, the activation axis in the flexicurity countries gathers measures that are characterized by a state that activates its workforce by means of direct human capital training, on-the-job learning or skill preservation measures. Policies that are not part of this repertoire, i.e. strategies focused on swift reintegration, display lower factor loadings and are hence less central to the conflict dimension. In dualizing countries, the activation dimension is determined by strategies to reintegrate the unemployed quickly without investing in training or public employment programmes. Interestingly, in these countries, short-time work has an ambiguous meaning; in fact, it is associated with both activation and traditional strategies because governments implement this measure to keep insiders on the payroll in times of crisis—a sort of insurance policy (Sacchi et al., 2011). At the same time, this policy helps preserve skills. This ambiguity can be clearly identified because in the dualizing regime, short-time work also scores, though less strongly, on the traditional axis. In sum, I find that policies that are characteristic of a regime determine political preferences more strongly (i.e. have high factor loadings) and thus mirror the different ALMP legacies: a human capital orientation in flexicurity and an occupational orientation in dualizing regimes. 7.1 Actor’s preferences and potential for coalition in flexicurity countries After describing the axes characterizing labour market conflicts, let us now analyse the single actor’s positions and the way in which these translate into coalition potential. Using the actors’ factor scores, I position them in the space. In Figures 1 and 2, I represent parties by a solid circle; state bodies, by a hollow circle; unions, by a hollow circle with a dotted border; employer organizations, by a hollow circle with a segmented border; and SMOs, by a small dot. The shading of the symbols differs across countries. Finally, I define potential coalitions to correspond to the quadrants, which result by crossing the traditional and activation dimensions. Quadrant II, which I call the modern left coalition, comprises actors who are in favour of expansion on both the traditional and the human capital imprinted activation dimensions. In the flexicurity countries, this group comprises several large unions, namely, the Swiss Confederation of Trade Unions (SGB), the Swiss unions Travail Suisse, Unia, and Union of Professionals (KV) and the Danish Union Federation (LO). Moreover, it includes the coordinating organ of the Swiss Cantons (SODK), the Danish Socialist Popular Party (SF) and the Danish Red-Green Alliance. Swiss greens are in favour of ALMP expansion, but they take a neutral stance on traditional policies. These results give some support to the hypothesis that green parties belong to this coalition. Furthermore, as expected based on insider–outsider theory, the bulk of unions favours expansion on both traditional and activation dimensions—at least in flexicurity countries. Quadrant III contains the traditional left coalition, which is characterized by the endorsement of passive benefits and EPL and sceptical attitudes towards increasing activation. The greatest supporters of the expansion of traditional labour market policies are the SMOs Attac and Kabba and, less so, the Danish white-collar union FTF. Whilst still part of the traditional left coalition, the Swiss union Syna; the Danish union AC; Swiss state actors, namely, the cantons of Aargau and Berne; the State Secretary for Economic Affairs (SECO); its Danish pendant NLMA; and the Danish Folkeparti (DF) adopt a moderate stance on the traditional dimension. Interestingly, the Swiss employer organization SAV is also part of this coalition, and it supports the status quo on the traditional dimension and is moderately against increasing activation efforts. These findings lend support to the hypothesis that state bodies and administrations in flexicurity countries should be sceptical of additional ALMP investments, as the implemented measures already work well. They are economically motivated to contain public spending and thus reject additional activation. The third-way coalition (Quadrant I) comprises a group of actors that share preferences for increased activation effort combined with a more or less market liberal stance. Specifically, Swiss (borderline) and Danish social democratic parties, Swiss greens (borderline), Swiss white-collar employees (AS) and, consistent with expectations, the largest Swiss employer organization (SGV) belong to this coalition. In fact, SGV combines a moderately positive stance on activation with a clearly market liberal position on the traditional axis. In sum, these organizations are strongly in favour of regime-specific activation strategies, but they simultaneously support the status quo or a reduction in passive welfare effort. Finally, in this group, the Swiss Christian Democratic Party (CVP) is an outlier, as it supports substantial liberalization and above average strengthening of ALMPs. This composition sustains the hypothesis that where welfare states address traditional risks efficiently and create attractive jobs for the previously unemployed, social democratic parties turn towards groups that are newly at risk. They do so by advocating the status quo or adopting a moderately liberal stance on the traditional axis but clearly supporting the expansion of human capital based ALMPs. Moreover, the results contradict the hypothesis that state bodies and administrations in flexicurity countries endorse increasing activation efforts because of their contact with supranational bodies. In contrast, I find that these organizations are oriented towards preserving the status quo on the traditional dimensions and rejecting additional expansion of activation polices. The traditional right coalition (Quadrant IV), which favours retrenchment on both dimensions, mainly includes employer organizations and conservative parties, i.e. the Swiss People’s Party (SVP) and the Danish Conservative party (KF). The Danish employer organization DI and the Swiss SBV share an extreme position in which they strongly reject additional activation, while the union Swissmem and the Danish DA have a very liberal stance on the traditional axis. These results contradict my expectation that employer organizations should support activation more strongly in flexicurity countries owing to the high standards of their retraining policies. Finally, there is a moderate subgroup of traditional right actors, namely, the Danish Economic Council (DEC), the Swiss conservative democrats (BDP), the Swiss liberal party (FDP), and the Danish liberal party (Venstre), which in terms of traditional polices, favour the status quo. Overall, the actor constellation suggests that in countries where welfare state benefits are encompassing, activation policies are human capital centred, and where the interests of insiders and outsiders converge owing to flexible labour market regimentation and acceptable jobs, unions favour expansion on both dimensions and social democratic parties cater to outsiders. In contrast, state bodies, employer organizations, and conservative and liberal parties are sceptical of additional activation. 7.2 Actor’s preferences and potential coalitions in dualizing countries As shown in Figure 2, in dualizing countries, unions are spread across both left coalitions. In the modern left coalition, we find several major unions, namely, the German IGM, DGB and Verdi; the French UNSA and CFECGC; and the Italian Christian democratic union (CISL). Contrary to my expectations, (former) communists (PCF in France, PdCI and PRC in Italy and die Linke in Germany) also cluster in this quadrant. These parties clearly favour increasing activation by addressing outsiders’ interests rather than focusing solely on their traditional constituency. Similar to the finding for the flexicurity countries, Figure 2 shows that the greens in Germany (Grüne), Italy (SEL) and France (EELV) favour expansion on both dimensions. Thus, they are not less supportive of activation in dualizing countries even though this strategy is less focused on human capital enhancement in such countries than in flexicurity countries and jobseekers are at risk of being relegated to precarious jobs. In the traditional left cluster, we find the German and the French socialist parties (SPD and PS), the left-oriented Italia dei Valori (IdV), the centrist UdC in Italy and the other unions (the Italian grass-roots organized COBAS and the UIL and the French CGT, Solidaires, FSU and CFDT). Only the Italian confederation of trade unions (CIGL) and the French CFTC belong to the third-way coalition and hence strongly support additional activation measures whilst taking a neutral stance on the traditional axis. These results corroborate previous studies and the expectation that in dualizing labour markets, social democrats stick to the traditional solution of strengthening policies to address insiders’ needs whilst neglecting outsiders’ interests in increasing activation effort. State bodies and administrations were expected to belong to the third-way coalition. Figure 2 shows that this hypothesis can be corroborated for the French Ministry of Work (MdT), the German Ministry of Social Affairs (BMAS), and the Italian state bodies, i.e. the Italian Ministry of Work (MdL) and the Ministry for Social Insurance (INPS). Conversely, the German Ministry for Economic Affairs (BWM) and the Ministry for Work (BAA) are situated in the right coalition, which in the light of their liberal-leaning Hartz IV reform, might not be surprising. Interestingly, German socialists (SPD), who initiated these reforms under Schröder, also tend towards a centrist position on the traditional axis in comparison with their Italian and French counterparts, who take positions further to the left. Moreover, state bodies share a moderate stance on the traditional axis but differ widely on the activation dimension. It appears that in dualizing countries, organizations that address social affairs and work are heavily influenced by the third-way consensus at the supranational level. This finding is plausible because these actors address the need for welfare reform. Conversely, in Germany, since major activation reforms were successfully implemented, the BWM and the BAA are more focused on balancing budgets than on increasing state involvement in either dimension. In the dualizing countries, the right coalition is extremely divided on the traditional dimension, but it has a homogeneous stance in refusing to increase occupational activation efforts. These right-oriented actors seem to disagree foremost on whether employment should be liberalized further. This issue has been particularly pushed by employer organizations and government parties in dualizing countries to counteract the rigidities of continental welfare states’ labour markets. The hypothesis that conservative government parties belong the third-way coalition is thus only partially corroborated. In fact, both the German CSU/CDU and the Italian PdL are part of the right coalition, whilst the French UMP, Lega and Italian PD are in the third-way coalition that endorses the expansion of ALMPs. In sum, I find that social democratic parties are part of the traditional left coalition catering to insiders and that greens and communists favour expansion on both dimensions and thus cater to both insiders and outsiders (modern left coalition). On the other extreme of the traditional dimension, employer associations and conservative political parties prefer the retrenchment of both traditional and activation policies. State bodies and administrations in Germany, where ALMPs have already been massively reformed, are part of the traditional right coalition. Conversely, in Italy and France, the administration supports third-way strategies by following the supranational consensus on the re-commodification of the labour force. 8. Conclusion In this article, I argue that political actors’ preferences are best understood as a product of the institutional context in which they are embedded. In fact, labour market regimes are the frames of reference against which actors form their preferences by shaping the way in which political problems are interpreted and adequate policy strategies are sought. The main objective of this article has been to show how the core players in the domain of labour market policy position themselves vis-à-vis the main policies used to fight unemployment, namely, traditional policies (EPL and passive benefits) and activation policies (training, reintegration measures and sanctions). The empirical analyses produced three main findings. First, the factor analyses show that the labour market policy space is characterized by two regime-specific conflict axes. Interestingly, in flexicurity countries, in line with their human capital ALMP tradition, the policies that determine the activation dimension are related to training and linked to strong engagement by the state in creating employment opportunities. In contrast, dualizing countries focus on reintegrating unemployed workers into the labour market without a focus on upskilling (occupational strategies). Secondly, I show that actors’ preferences and thus their potential coalitions also depend on the institutional setup. In dualizing countries, where pressure to adapt ALMPs to the EES guidelines is highest, I find that state bodies and administrations are more supportive of expansion on the activation axis. In contrast, in flexicurity countries, where these schemes are already encompassing and functioning well, they seem to think there is no need to burden the welfare budget by expanding these policies. Accordingly, state bodies are in the traditional right coalition, which supports retrenchment on both dimensions. The results also show that unions—particularly in flexicurity countries—communist parties and greens in both regimes embrace modern left policy strategies, favouring expansion on both the activation and traditional dimensions. Conversely, many of the large unions in dualizing countries have joined forces with social democratic parties to protect insiders’ interests; they reject the expansion of activation policies but support the expansion of traditional policies. I also find that, in both regimes, right-oriented parties and employer organizations have the potential to coalesce; they are against expansion on both the traditional and activation dimensions. Finally, focusing on social democratic parties, this new data set allows me to corroborate the findings of Rueda (2007) and show that these parties concentrate on labour market insiders in dualizing countries (a traditional left strategy) but extend their mobilization to outsiders in flexicurity countries (a third-way strategy). In sum, political actors’ preferences, and thus their potential coalitions, differ depending on the institutional design. What are the implications of these findings? My results challenge the widespread practice in political economy to compare preferences, coalitions and conflict patterns across countries without accounting for the intuitional peculiarities that shape political action. This likely gives rise to regime-specific cross-class alliance potentials: in flexicurity countries, state bodies and administrations might plausibly build powerful alliances with employers and conservative parties to retrench state efforts, whereas in dualizing countries, state bodies may join forces with unions and left opposition parties to address outsiders’ needs. My results also indicate that the composition of the conflict structure in a specific policy field does not always refer to the same content. The factor analyses show that the traditional and activation dimensions differ across regimes, a result that calls for more sophisticated analyses of political preferences in general and of the validity of non-specific indexes, such as the left-right dimension, in particular. By studying the basic conflict lines that structure political preferences concerning labour market policy, I unveil the fundamental preferences of a variety of actors and hence lay the groundwork for understanding and studying policymaking in this domain. Future research could focus on coalition formation in phases, where strategic alliances gain importance beyond mere ideological commonalities, or trace coalitions over longer periods or in other contexts. Supplementary material Supplementary material is available at Socio-Economic Review Journal online. Funding This research has been supported by the NCCR Democracy 21, Project 11, which is financed by the Swiss National Science Foundation. Footnotes 1 Political actors’ preferences determine their potential to form coalitions with other actors based on common ideological and/or value-based ground. These preferences and coalitions constitute the conflict structure in a policy field (cf. Kriesi et al., 2008; Häusermann, 2010). 2 I use ‘position’ and ‘preference’ interchangeably because I merely analyse value-based or ideology-based positions and do not include strategic motives, which could lead to a distinction between real coalitions and expressed preferences (Immergut, 1998). 3 Institutional legacies and actors’ preferences are, to some extent, endogenous since preferences articulated in a political process shape the institutional settings, and vice versa. Here, I analyse actors’ preferences but not the politics that may follow from expressing these attitudes. 4 Flexicurity is a debated concept. I follow Whiltagen and Tros (2004) and define flexicurity as ‘[t]he pursuit of a balance between “flexibility” and “security” …’. This scheme supports weaker groups and not merely labour market insiders, and it results from the coordination effort of social partnership and the state, allowing the synchronization of social and economic policy. As argued by Viebrock and Clasen (2009), flexicurity countries such as Denmark rely on flexible labour markets, generous unemployment support and activation (cf. Madsen, 2002). 5 To analyse how specific labour market characteristics affect political actors’ preferences, I group countries with similar characteristics into labour market regime types. I refrain from analysing political actors’ preferences by country because I am not interested in the effect of possibly idiosyncratic country-specific institutional settings on preferences. 6 The third approach identified in the literature is referred to as ‘work-first’ activation. This model, which is characteristic of liberal countries, stresses the need to swiftly reintroduce workers into the labour market, principally by means of conditionality and negative incentives. 7 Originally, 132 political actors were contacted, and 109 interviews were obtained (refer to Tables A1 and A2 in the Online Appendix for the details on data collection and response rates). 8 Interview partners were chosen as representatives of the major organizations that are involved in the field of unemployment policy. The relevance of the organizations was cross-checked with two experts per country and validated by means of media analyses. See Tables A3 and A4 in the Online Appendix for the complete list of actors. 9 To capture fundamental political conflicts, it is essential to focus on salient policies because political actors have a stance on all issues, but they judge them differently in terms of relevance. Accordingly, I calculate the indicator by multiplying the standardized salience and position for each actor. This procedure attributes less weight to measures that actors consider less relevant. An actor’s position on a given policy measure was gauged on a scale ranging from 1 to 5 (strongly disagree to strongly agree). To operationalize a measure’s salience, the respondents indicated the most important measure, the three most important measures, and the three least important measures. The resulting salience indicator assigns three points to the most important measure, two points to the other two important measures, zero points to the three least important measures and one point to the remaining measures. 10 See Table A5 in the Online Appendix for the question wording. 11 Missing cases were recoded as neutral both in position and salience; however, they represent only between 2% and 5% of cases, see Tables A6 and A7 in the Online Appendix for descriptive statistics and correlations. 12 The results are not shown but are available upon request. 13 See Tables A8 and A9 and Figures A1–A5 in the Online Appendix for country-specific solutions. 14 The findings suggest that there is also variation in the understanding of traditional policies, but for reasons of space, I focus on activation preferences. However, the fact that regime specificities exist on the traditional axis allows for the generalization of the argument that institutional legacies influence political actors’ preferences. Acknowledgements The author would like to thank Silja Häusermann, Hanspeter Kriesi, Giuliano Bonoli, Ruth Beckmann, Fabienne Liechti, Desmond King, Alexandre Afonso, Laurent Bernhard, Regula Hänggli, four anonymous reviewers and the editors of Socio-Economic Review for very helpful and constructive comments. The author would also thank Laurent Bernhard, Regula Hänggli and Kirsty Stone-Wyler for excellent cooperation during the data collection process. References Afonso A. ( 2015) ‘ Choosing Whom to Betray: Populist Right-Wing Parties, Welfare State Reforms and the Trade-Off between Office and Votes’, European Political Science Review , 7, 271– 292. DOI: 10.1017/S1755773914000125. Google Scholar CrossRef Search ADS   Arnd C. ( 2013) The Electoral Consequences of Third Way Welfare State Reforms: Social Democracy’s Transformation and Its Political Costs , Amsterdam, Amsterdam University Press. Google Scholar CrossRef Search ADS   Barbier J., Ludwig-Mayerhofer W. ( 2004) ‘ Introduction’, European Societies , 6, 423– 436. DOI: 10.1080/1461669042000275845. Google Scholar CrossRef Search ADS   Berton F., Richiardi M., Sacchi S. ( 2009) Flex-Insecurity: Perché in Italia la flessibilità Diventa precarietà , Bologna, Il Mulino. Bonoli G. ( 2005) ‘ The Politics of the New Social Policies: Providing Coverage against New Social Risks in Mature Welfare States’, Policy and Politics , 33, 431– 449. DOI: 10.1332/0305573054325765. Google Scholar CrossRef Search ADS   Bonoli G. ( 2013) The Origins of Active Social Policy. Labour Market and Childcare Policies in a Comparative Perspective , Oxford, Oxford University Press. Google Scholar CrossRef Search ADS   Bonoli G., Natali D. ( 2012) The Politics of the ‘New’ Welfare State: Analysing Reforms in Western Europe , Oxford, Oxford University Press. Clasen J., Clegg D. (ed.) ( 2011) Regulating the Risk of Unemployment. National Adaptations to Post-Industrial Labour Markets in Europe , Oxford, Oxford University Press. Google Scholar CrossRef Search ADS   Daguerre A. ( 2007) Active Labour Market Policies and Welfare Reform. Europe and the US in Comparative Perspective , Houndmills/NY, Palgrave Macmillan. Google Scholar CrossRef Search ADS   Daguerre A., Taylor-Gooby P. ( 2004) ‘ Neglecting Europe: Explaining the Predominance of American Ideas in New Labour's Welfare Policies Since 1997’, Journal of European Social Policy , 14, 25– 39. DOI: 10.1177/0958928704039786. Google Scholar CrossRef Search ADS   Duell N., Tergeist P., Bazant U., Cimper S. ( 2010) Activation Policies in Switzerland, OECD Social Employment and Migration Working Papers, accessed at www.oecd.org/els/workingpapers on July 16, 2016. Eichhorst W., Konle-Seidl R. ( 2008) Contingent Convergence: A Comparative Analysis of Activation Policies, IZA Discussion Paper Series, No. 3905, Bonn, Institute for the Study of Labor (IZA). Emmenegger P. ( 2009) ‘ Barriers to Entry: Insider/Outsider Politics and the Political Determinants of Job Security Regulations’, Journal of European Social Policy , 19, 131– 146. DOI: 10.1177/0958928708101866. Google Scholar CrossRef Search ADS   Emmenegger P., Häusermann S., Palier B., Seeleib-Kaiser M. (eds) ( 2012) The Age of Dualisation. The Changing Face of Inequality in Deindustrializing Societies , Oxford, Oxford University Press. Google Scholar CrossRef Search ADS   Entman R. M. ( 1993) ‘ Framing: Toward Clarification of a Fractured Paradigm’, Journal of Communication , 43, 51– 58. DOI: 10.1111/j.1460-2466.1993.tb01304.x. Google Scholar CrossRef Search ADS   Esping-Andersen G. ( 1990) Three Worlds of Welfare State Capitalism , Princeton, Princeton University Press. Fleckenstein T. ( 2008) ‘ Restructuring Welfare for the Unemployed: The Hartz Legislation in Germany’, Journal of European Social Policy , 18, 177– 188. DOI: 10.1177/0958928707087593. Google Scholar CrossRef Search ADS   Fossati F. ( 2017) ‘Who Wants Demanding Active Labour Market Policies? Public Attitudes towards Policies That Put Pressure on the Unemployed’, Journal of Social Policy , 13, 1– 21. DOI: 10.1017/S0047279417000216. Google Scholar CrossRef Search ADS   Fossati F., Häusermann S. ( 2014) ‘ Social Policy Preferences and Party Choice in the 2011 Swiss Elections’, Swiss Political Science Review’ , 20, 590– 611. DOI: 10.1111/spsr.12115. Google Scholar CrossRef Search ADS   Geering D., Häusermann S. ( 2013) ‘Changing Party Electorates and Economic Realignment. Explaining Party Positions on Labor Market Policy in Western Europe’, paper presented at the ECPR General Conference, Bordeaux, France, September 4–7. Gerfin M., Lechner M. ( 2002) ‘ A Microeconometric Evaluation of the Active Labour Market Policy in Switzerland’, Economic Journal , 112, 854– 893. DOI: 10.1111/1468-0297.00072. Google Scholar CrossRef Search ADS   Gilbert N. ( 2002) Transformation of the Welfare State. The Silent Surrender of Public Responsibility , Oxford, Oxford University Press. Google Scholar CrossRef Search ADS   Gordon J. C. ( 2015) ‘ Protecting the Unemployed: Varieties of Unionism and the Evolution of Unemployment Benefits and Active Labor Market Policy in the Rich Democracies’, Socio-Economic Review , 13, 79– 99. DOI: 10.1093/ser/mwu029. Google Scholar CrossRef Search ADS   Hall P. A., Soskice D. (eds) ( 2001) Varieties of Capitalism , Oxford, Oxford University Press. Google Scholar CrossRef Search ADS   Häusermann S. ( 2010) The Politics of Welfare State Reform in Continental Europe. Modernization in Hard Times , Cambridge, Cambridge University Press. Google Scholar CrossRef Search ADS   Huber E., Stephens J. ( 2001) Development and Crisis of the Welfare State. Parties and Policies in Global Markets , Chicago, University of Chicago Press. Google Scholar CrossRef Search ADS   Immergut E. M. ( 1998) ‘ The Theoretical Core of the New Institutionalism’, Politics and Society , 26, 5– 34. DOI: 10.1177/0032329298026001002. Google Scholar CrossRef Search ADS   Kemmerling A., Bruttel O. ( 2006) ‘“ New Politics” in German Labour Market Policy? The Implications of the Recent Hartz Reforms for the German Welfare State’, West European Politics , 29, 90– 112. DOI: 10.1080/01402380500389273. Google Scholar CrossRef Search ADS   Kitschelt H. ( 1994) The Transformation of European Social Democracy , Cambridge, Cambridge University Press. Google Scholar CrossRef Search ADS   Knoke D., Pappi F. U., Broadbent J., Tsujinaka Y. ( 1996) Comparing Policy Networks. Labor Politics in the U.S., Germany, and Japan , New York, Cambridge University Press. Google Scholar CrossRef Search ADS   Knotz C. ( 2015) The Politics of Unemployment Benefit Conditionality , Lund, Lunds Universitet. Korpi W. ( 1983) The Democratic Class Struggle , London, Routledge. Kriesi H., Grande E., Lachat R., Dolezal M., Bornschier S., Frey T. ( 2008) West European Politics in the Age of Globalization , Cambridge, Cambridge University Press. Google Scholar CrossRef Search ADS   Larsen C. A. ( 2007) ‘ The Institutional Logic of Welfare Attitudes. How Welfare Regimes Influence Public Support’, Comparative Political Studies , 41, 145– 168. Google Scholar CrossRef Search ADS   Lindbeck A., Snower D. J. ( 1988) The Insider-Outsider Theory of Employment and Unemployment , Cambridge, MIT Press. Lowi T. J. ( 1972) ‘ Four Systems of Policy, Politics, and Choice’, Public Administration Review , 32, 298– 310. DOI: 10.2307/974990. Google Scholar CrossRef Search ADS   Madsen P. ( 2002) ‘The Danish Model of Flexicurity: A Paradise with Some Snakes’. In Sarfati H., Bonoli G. (eds) Labour Market and Social Protection Reforms in International Perspective , Aldershot, Ashgate, pp. 243– 265. Myles J., Quadagno J. ( 2002) ‘ Political Theory of the Welfare State’, Politics & Society , 76, 34– 57. Nelson M. ( 2013) ‘ Making Markets with Active Labor Market Policies: the Influence of Political Parties, Welfare State Regimes, and Economic Change on Spending on Different Types of Policies’, European Political Science Review , 5, 255– 277. DOI: 10.1017/S1755773912000148. Google Scholar CrossRef Search ADS   Oesch D. ( 2006) Redrawing the Class Map. Stratification and Institutions in Britain , Germany, Sweden and Switzerland, Houndsmill/NY, Palgrave Macmillan. Google Scholar CrossRef Search ADS   Organisation for Economic Co-operation and Development (OECD) ( 2013) ‘Employment Protection Indicator, OECD Employment and Labour Market Statistics (Database)’, accesed at http://www.oecd.org/employment/emp on February 2, 2015. Organisation for Economic Co-operation and Development (OECD) ( 2016a) ‘Public Expenditures and Participation Stocks in LMP’, accessed at stats.oecd.org on February 24, 2016. Organisation for Economic Co-operation and Development Organisation for Economic Co-operation and Development (OECD) ( 2016b) ‘The Incidence of Involuntary Part-Time Workers’, accessed at https://stats.oecd.org/Index.aspx?DataSetCode=TEMP_I# on February 24, 2016. Palier B. (ed.) ( 2010) A Long Goodbye to Bismarck? The Politics of Welfare State Reform in Continental Europe , Amsterdam, Amsterdam University Press. Google Scholar CrossRef Search ADS   Pierson P. ( 1998) ‘ Irresistible Forces, Immovable Objects: Post-Industrial Welfare States Confront Permanent Austerity’, Journal of European Public Policy , 5, 539– 560. DOI: 10.1080/13501769880000011. Google Scholar CrossRef Search ADS   Pierson P. ( 2001) ‘Coping with Permanent Austerity. Welfare State Restructuring in Affluent Democracies’. In Pierson P. (ed.) The New Politics of the Welfare State , Oxford, Oxford University Press, pp. 410– 459. Google Scholar CrossRef Search ADS   Rueda D. ( 2005) ‘ Insider-Outsider Politics in Industrialised Democracies: The Challenge to Social Democratic Parties’, American Political Science Review , 99, 61– 74. Google Scholar CrossRef Search ADS   Rueda D. ( 2007) Social Democracy Inside Out. Partisanship and Labor Market Policy in Industrialized Democracies , Oxford, Oxford University Press. Google Scholar CrossRef Search ADS   Sabatier P. A., Jenkins-Smith H. C. ( 1999) ‘The Advocacy Coalition Framework: An Assessment’. In Sabatier P. A. (ed.) Theories of the Policy Process , Boulder, Westview and Publishing House, pp. 117– 166. Sabatier P. A., Weible C. M. ( 2007) ‘The Advocacy Coalition Framework. Innovations and Clarifications’. In Sabatier P. A. (ed.) Theories of the Policy Process , Cambridge, Westview Press, pp. 189– 220. Sacchi S., Pancaldi F., Arisi C. ( 2011) ‘ The Economic Crisis as a Trigger of Convergence? Short-Time Work in Italy, Germany and Austria’, Social Policy & Administration , 45, 465– 487. DOI: 10.1111/j.1467-9515.2011.00785.x. Google Scholar CrossRef Search ADS   Saint-Paul G. ( 1998) ‘ A Framework for Analyzing the Political Support for Active Labor Market Policy’, Journal of Public Economics , 67, 151– 165. DOI: 10.1016/S0047-2727(97)00069-8. Google Scholar CrossRef Search ADS   Shahidi F. V., Moortel D. de, Muntaner C., Davis O., Siddiqi A. ( 2016) ‘ Do Flexicurity Policies Protect Workers from the Adverse Health Consequences of Temporary Employment? A Cross-National Comparative Analysis’, SSM - Population Health , 2, 674– 682. DOI: 10.1016/j.ssmph.2016.09.005. Google Scholar CrossRef Search ADS   Steinmo S. ( 1989) ‘ Political Institutions and Tax Policy in the United States, Sweden, and Britain’, World Politics , 41, 500– 535. DOI: 10.2307/2010528. Google Scholar CrossRef Search ADS   Stiller S., van Gerven M. ( 2012) ‘ The European Employment Strategy and National Core Executives: Impacts on Activation Reforms in the Netherlands and Germany’, Journal of European Social Policy , 22, 118– 132. DOI: 10.1177/0958928711433652. Google Scholar CrossRef Search ADS   Svalund J. ( 2013) ‘ Labor Market Institutions, Mobility, and Dualization in the Nordic Countries’, Nordic Journal of Working Life Studies , 3. DOI: 10.19154/njwls.v3i1.2523. Swank D., Martin C. J. ( 2001) ‘ Employers and the Welfare State’, Comparative Political Studies , 34, 889– 923. DOI: 10.1177/0010414001034008003. Google Scholar CrossRef Search ADS   Tepe M., Vanhuysse P. ( 2013) ‘ Parties, Unions and Activation Strategies: The Context-Dependent Politics of Active Labour Market Policy Spending’, Political Studies , 61, 480– 504. DOI: 10.1111/j.1467-9248.2012.00996.x. Google Scholar CrossRef Search ADS   Thelen K. ( 2012) ‘ Varieties of Capitalism: Trajectories of Liberalization and the New Politics of Social Solidarity’, Annual Review of Political Science , 15, 137– 159. DOI: 10.1146/annurev-polisci-070110-122959. Google Scholar CrossRef Search ADS   Torfing J. ( 1999) ‘ Towards a Schumpeterian Workfare Post National Regime: Path-Shaping and Path-Dependency in Danish Welfare State Reform’, Economy and Society , 28, 369– 402. DOI: 10.1080/03085149900000010. Google Scholar CrossRef Search ADS   Trickey H., Walker R. ( 2001) ‘Steps to Compulsion within British Labour Market Policies’. In Lødemel I., Trickey H. (eds) An Offer You Can't Refuse. Workfare in International Perspective , Bristol, The Policy Press, pp. 181– 213. van Vliet O., Caminada K. ( 2012) ‘Unemployment Replacement Rates Dataset among 34 Welfare States, 1971–2009. An Update, Extension and Modification of the Scrugg’s Welfare State Entitlement Data Set’. Neujobs Special Report, No. 2. Venn D. ( 2009) Legislation, Collective Bargaining and Enforcement: Updating the OECD Employment Protection Indicators, Employment and Migration Working Paper No. 89. Viebrock E., Clasen J. ( 2009) Flexicurity – A State-Of-the-Art Review. Working Papers on the Reconciliation of Work and Welfare in Europe, No. 1, RECWOWE Publication, Edinburgh, Dissemination and Dialogue Centre. Vlandas T. ( 2013) ‘ Mixing Apples with Oranges? Partisanship and Active Labour Market Policies in Europe’, Journal of European Social Policy , 23, 3– 20. DOI: 10.1177/0958928712463161. Google Scholar CrossRef Search ADS   Whiltagen T., Tros F. ( 2004) ‘“The Concept of Flexicurity”. A New Approach to Regulating Employment and Labour Markets’, Transfer, European Review of Labour and Research , 10, S. 166– 186. Google Scholar CrossRef Search ADS   Wilthagen T. ( 2002) ‘Managing Social Risks with Transitional Labour Markets’. In Mosley H., O’Reilly J., Schönmann K. (eds) Labour Markets, Gender and Institutional Change. Essays in Honour of Guenther Schmid . Cheltenham, Edward Elgar, pp. 264– 289. Wüest B., Fossati F. ( 2014) ‘ Quantitative Discursive Institutionalism: A Comparison of Labour Market Policy Discourse across Western Europe’, Journal of European Public Policy , 22, 708– 730. Google Scholar CrossRef Search ADS   © The Author 2017. Published by Oxford University Press and the Society for the Advancement of Socio-Economics. All rights reserved. For Permissions, please email: journals.permissions@oup.com

Journal

Socio-Economic ReviewOxford University Press

Published: Nov 2, 2017

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off