THE SUBSIDIARIZATION OF SOCIAL POLICIES: ACTORS, PROCESSES AND IMPACTS: Some reflections on the Italian case from a European perspective Open Access

This paper seeks to disentangle analytically the trends towards the subsidiarization of social policies in Europe addressing how the social policy reforms changed actors’ governance arrangements and altered the government scales involved over the last few decades. Following an introductory analysis of the main causes of the subsidiarization of social policies (sections 1 and 2), attention is devoted to the differences between welfare systems in Europe (section 3) and the way they influence the directions of change. The hypothesis is that, despite the fact that the subsidiarization of social policies is a converging rhetoric in most European countries’ social policy reforms, the impact of the process varies according to the specificities of the respective regulatory frames at the national or sub-national scales. This is exemplified in the final part of the paper (section 4) which considers the Italian case and the main critical dimensions that emerge through these processes.

Since the end of the 1970s, structural changes deeply affected socio-demographic and the socio-economic spheres in all industrialized countries. These included the aging of the population, increasing instability of family arrangements, and decreasing fertility rates coupled with a deep economic restructuring process and the spread of new modes of production. These changes undermined the functioning of welfare institutions. Social policies had developed in a context of economic growth, increasing available resources and relatively stable needs. Emerging conflicts and exploding needs (from unemployment to care) contributed to the fiscal crisis of the state and necessitated reforms. These reforms – occurring in a context of strong budgetary constraints – are characterized in most cases, implicitly or explicitly, by two major processes: (a) a reorganization of regulative powers at the different territorial levels; and (b) a multiplication of actors involved in designing, managing and implementing social policies. The combined effect of these two processes can be called the subsidiarization of social policies, both in vertical (i.e., the territorial reorganization of regulatory powers) and horizontal terms (the multiplication of actors).

The concept of subsidiarity¹ is usually defined as the principle which states that matters ought to be handled by the smallest (or, the lowest) competent authority. This implies that a central authority should have a subsidiary function, performing only those tasks which cannot be performed effectively at a more immediate or local level (Waschkuhn 1995). The concept gained momentum since the end of the last century and has now become a rhetorical keystone behind most institutional reform strategies implemented in Europe² aiming at promoting complex arrangements of multi-level governance.³ Different actors at different scales interact and bargain without necessarily a clearly settled hierarchy (Le Galès 2002; Bache and Flinders 2004).

The aim of this contribution is to address these processes at a European level from the point of view of social policies, highlighting the specific features of the Italian case. Following an introductory analysis on the main causes of the subsidiarization of social policies (sections 1 and 2), attention will be devoted to the differences between welfare systems in Europe (section 3) and the way they influence, or even pre-structure, the directions of change. The hypothesis is that despite the fact that the subsidiarization of social policies is a converging rhetoric in most European countries’ social policy reforms, the impact of the process varies according to the specificities of the respective regulatory frames at national or sub-national levels. This assumption will be exemplified in the final part of the paper (section 4) which considers the Italian case and the main critical dimensions that emerge through these processes.

The territorial reorganization of social policies represents a complex process emerging due to the endogenous and exogenous changes briefly sketched above. National welfare systems started to experience difficulties in addressing an increasingly differentiated and fragmented need scenario (McEwen and Moreno 2004). In order to fully understand these complex processes and their impact on the regulative powers and the allocation of competences at the different territorial levels we need to adopt a long-term perspective, deciphering the spatial shifts which occurred in the last three decades.

Up to the second half of the 1970s most welfare policies were regulated at the national level. This was true not only for contributory based policies, but also for active labour market policies and other standardized welfare services (e.g., Job placement services) (Ferrera 2006). These policies were accompanied by specific forms of spatial Keynesism through which the state promoted the economic development in peripheral areas such as the Mezzogiorno in Italy, Andalusia in Spain and parts of Scotland in the attempt to iron out existing differences at the sub-national levels (Brenner 2004: 460). In this phase, welfare policies were characterized by a collective management of individual standardized risks, within the frame of a national territorial dimension.

Since the second half of the 1970s, the internationalization and globalization of economic relations western industrialized countries have resulted in a rapid increase of inflation rates and unemployment, accompanied by a drop in economic growth rates and the weakening of the regulative capacity of the nation state (Amin 1994; Jessop 2002). The specific and dynamic equilibrium that also created social cohesion and redistribution during the golden age of the welfare state collapsed when the virtuous circle of full employment and high growth rates was interrupted.

The processes of institutional change which accompanied these transformations in Europe followed a double direction. From the economic point of view, both supra-national and sub-national actors become relevant. On the one side, the European Union established the common market, defined competition rules and introduced the Euro. On the other side, cities and regions acquired a strategic importance, in the frame of global competitiveness (Swyngedouw 1997; Crouch et al.2001; Buck et al.2005; Salet 2006) because the economic performance of their territories increasingly becomes the basis for their revenues (see section 4).

From the point of view of social policies, the scenario is slightly different. At the supra-national level, with a few exceptions mainly related to the rhetoric of the employment strategy (Ashiagbor 2005), social policies played a rather marginal role. Formally important steps have been taken such as the common objectives in the fight against social exclusion and poverty adopted at the Nice European Council, and with the NAP's inclusion strategy. These documents, however, did not translate into concrete policy measures at the European level. What happened, on the contrary, is that sub-national actors, in particular regions and cities, increased their role also in regulatory terms. The resulting scenario is characterized by the coexistence of all policy levels, negotiating and renegotiating constantly the new political opportunity structures available (Le Galès 2002). The rhetoric legitimizing this process highlights proximity to the citizens, often underplaying the risk of lacking common rights, standards and opportunities.

Within the processes of territorial re-organization of social policies we can recognize two main trends over the last 30 years: an implicit and an explicit form of rescaling. In the case of implicit rescaling, what changes is the relative weight of specific measures regulated at different territorial levels. For instance, policies regulated at the national level loose resources, claimants and the like in favour of policies regulated at the sub-national, local level. In the case of explicit rescaling – forms of territorial re-organization of social policies – it is explicit reforms that shift regulatory power to other levels. The latter, for instance, occurred in most European countries for all public employment services, which become increasingly localized (Lundin and Skedinger 2000; Ferrera 2006).

Both processes took place in almost all European countries and it can be said that in general, since the end of the 1970s up to the mid 1980s, the implicit model of change prevailed, while explicit reforms become more relevant after the second half of the 1990s. The institutional inertia of Keynesian solutions to specific conditions of need in fact slowed down the reform process and unemployment had to become a structural phenomenon, before the fiscal pressure on the welfare state could trigger deep reform processes.⁴

Since the second half of the 1990s, in parallel with the attempts to contain the dynamics of unemployment with the spread of both active labour market and activation policies, we witness explicit forms of territorial re-organization of social policies through reforms aimed at modifying the allocation of regulative capacities at different scales of government, legitimized by the spread rhetoric of the principle of vertical subsidiarity. Also in this case the existing contextual regulatory conditions influenced the timing and the direction of the reforms.

These trends were also paralleled by changes in the financial relations between different scales. In particular, in the 1990s reforms aimed at increasing the financial autonomy of sub-national levels of government both in terms of revenues, fostering own taxation (Stegarescu 2005: 301), and in terms of expenditure, through the increase of costs to be covered by local resources. Both these trends implied a reduction of the existing differences of the degree of financial decentralization⁵ (see Table 1) (Thieben 2003).

The consequences of these trends are complex to decipher because a high share of sub-national taxation does not necessarily correspond to a high degree of sub-national autonomy. In fact, sub-national taxation might be strictly regulated by national/federal criteria, which heavily structure and restrict their autonomy (OECD 2006a). In order to understand these changes it is, therefore, necessary to embed them in the regulatory frames characterizing the different welfare systems in Europe. In fact, the reforms of the last decades produced a quite diversified landscape characterized by complex interwoven regulative powers at different scales, constantly negotiating within the frame of the subsidiarity principle. This subsidiarization did not only frame the institutional relations among levels of government, but also the role of different actors vis-à-vis the public–private lines of differentiation in the management and implementation of the policies themselves.

The second process of change that the different welfare systems are facing is represented by the multiplication of actors involved in designing, managing and implementing social policies.

We can look at this process from the perspective of horizontal subsidiarization. In fact, all these processes have been taking place since the early 1980s. The externalization of social services (Dixon and Hyde 2001), the individualization (Beck 1998) of specific measures and – in some cases – their privatization (Dixon and Hyde 2001), strongly contributed to a new role for private – both for profit and non-profit – actors, in the design, management and implementation of welfare policies (Ascoli and Ranci 2002; Powell 2007).

The reasons for the involvement of new actors are manifold. Budgetary constraints played a major role, but are not the only reason. The opinion that widening the range of actors would be functional for increasing the policy effectiveness and efficiency has been paralleled by the spread of a participatory rhetoric which viewed the involvement of civil society – strongly rooted in the full deployment of the subsidiarity principle – as a necessary step to empower the people in finding a solution to their problems (Ranci 2006).

The involvement of new actors regarded initially the implementation of social services and their management. Only in recent years their role shifted to planning and the co-definition of goals. All that determined a re-definition of the policy making process, the spread of new models of governance in the management of social policies, implying a shift from a hierarchical-vertical regulation towards a more cooperative-horizontal one.

This re-definition of the policy-making process requires new state capacities aimed at supporting the transition from a government function of a democratically elected body towards a form of governance, in which the public actor aims at facilitating and coordinating the implementation of policies. This occurs within a frame in which the discussion about social policy design, its management and implementation takes place in an increasingly fragmented and uncertain context, characterized by a high degree of negotiation among the actors involved.

As a reaction to the crisis of the welfare state, the reform processes – in their double meaning of vertical and horizontal subsidiarization – produced a steady shift from a vertical towards a horizontal coordination of social policies, which finds in the local dimension its ideal level of implementation. Despite the fact that these tendencies are common to most European countries, the development and institutionalization of the new governance arrangements do not converge. On the contrary, the results of these processes of change seem to produce a territorially structured diversification. This diversification varies according to socio-economic context and institutional milieu⁶ with all the specificities and differentiated degrees of freedom, ranging from the high level of the Comunitad Autonomas and the Länder to the low level of intra-national differentiation in France.

Di Gaetano and Strom (2003) emphasize the fact that these different institutional contexts favour specific governance arrangements at the local level; that is to say: formal and informal rules actors build and follow, developing practices and networks, which tend to contribute – in one way or another – to the stability and dynamics of the local society and political regime (Le Galès 2002: 15). Other scholars have gone beyond the local dimension, witnessing how the emerging governance arrangements are highly coherent matching the respective national welfare systems they belong to, following a path-dependent logic which tends to pre-structure the outcome of the current processes of change (Jessop 2002; Kazepov 2005). In this sense, reform processes do not occur in an institutional vacuum, but are deeply rooted in the institutional settings of the individual countries which define – through specific constraints and opportunities – the directions of change. Welfare systems and their functioning are therefore a privileged perspective in understanding the processes of subsidiarization of social policies and of their impact.

Understanding the processes of subsidiarization of social policies cannot avoid considering what the main features of European welfare systems are. It is these very features and the underlying regulative principles that translate the common transformation processes into context specific outcomes and practices which influence the allocation of resources and responsibilities. Following the legacy of Karl Polanyi's seminal work, The Economy as Instituted Process (1957), in the last few decades the debate on welfare systems has been revitalized connecting regulatory principles to specific policies, in particular: (a) the principle of redistribution to universalistic policies, (b) the principle of exchange to liberal-market-oriented policies; and (c) the principle of reciprocity to policies based on active or passive subsidiarity. This coherence of the regulative principles is also reflected in the country-specific governance arrangements which get structured – in a path dependent dynamic – within different local contexts. The recent debate focused both on the number of existing models and the underlying regulatory principles, as well as on the kind of indicators to be considered in the taxonomic exercise. The literature on this issue is quite diversified (Esping-Andersen 1990, 1999; Mingione 1997; Ferrera 1998; Gallie and Paugam 2000; Art and Glissen 2002; Kazepov 2005). Most authors propose – implicitly or explicitly – a classification based on the equilibrium among different regulatory principles and the way in which the respective institutions socialize risks. In particular, the country-specific equilibrium among the state, the family (plus the third sector) and the market (Esping-Andersen 1990; Mingione 1997; Ascoli and Ranci 2002) contributes to ways in which the allocation of resources is defined, and, in doing so, shapes the constraints and opportunities actors face in defining their life strategies. Within this picture, the state (through the rule-of-law), plays a major role, because it defines also – implicitly or explicitly – the role of the other institutions. In fact, by defining access criteria and the generosity of provisions, for instance, the state not only allocates resources to specific groups, but implicitly (and functionally) also the responsibilities in the production of well-being to the other institutions, be it the market (through commodification) or the family (through familialization) (Esping Andersen 1999, Kazepov 2005).

Through these perspectives, we can identify five main welfare systems within which the European welfare systems can be placed: the liberal welfare system, the social democratic welfare system, the corporatist welfare system, the familistic welfare system and the welfare system of transition countries. A brief comparative sketch based on data reported in Table 2 might help in highlighting the existing differences which will be synthesized in Figure 1 at the end of the section.

(1) The Liberal welfare system is characterized by a relatively residual role of the state, which intervenes only when both the market and the family have failed in allocating resources. The market is the prevailing mechanism of regulation and integration in a highly individualized (see positions in Figure 1) and competitive society. The main example of this model is represented by the United States, where the importance of welfare policies is extremely limited. Among the European countries, the United Kingdom is the closest one to this model, even though some substantial differences due to the legacy of the Beveridgian welfare state and the relatively developed social services. Poverty and inequality rates are, nevertheless, among the highest in Europe (see Table 2) and social policies – often means tested –assumed, after the long period in power of the Conservative government (1979–1997), a rather negative image (Rodney 2004). Generosity and adequacy of income support levels are on an intermediate position (see ‘x’ axis in Figure 1), lower than in Scandinavian countries, but definitely higher than in South European ones. From the territorial point of view, differentiation among regions occurs more in relation to activation policies and to the related local governance arrangements (for training, stages, social insertion, …). Social assistance benefits and other income support measures per se are more or less managed homogenously throughout the country, but the joint PES and Benefits agency management of the New Deals is increasingly fragmented and depends very much on the resources available locally, which are in turn related to the degree of competitiveness regions and cities can achieve in the market (Buck et al.2005). The governance arrangements developed within this welfare system tend to be, from this point of view, pluralist and managerial, coordinating public–private actors and prioritizing efficiency over equality.

Despite successful targeting and efficient poverty reduction, the incidence of poverty is higher than in other European countries and income inequality very high (see Gini index in Table 2).

(2) The Social democratic welfare system is characterized by a pervasive role of the state which takes over responsibilities that other systems delegate to the family or the market. Measures are universalistic, addressed to all citizens according to need. A wide range of in-kind services and monetary transfers is supplied, and redistribution is the most important form of allocation, even if in the last decades some second-level insurance-based schemes have been introduced. Market dependency, poverty and inequality are the lowest in the EU (Ervik and Kuhnle 1996). Adequacy of income support measures is a fact and local differentiation tends to decrease, except for social assistance and activation measures, which depend on local political and economic contexts. The debate over the consequences of this differentiation brought about a homogenization of standards in most Scandinavian countries (e.g., Sweden in 1998). Governance is regulated by managerial but also participative arrangements in which public authorities, rather than private non- or for-profit actors. Income inequality and poverty before taxes and transfers is very high, after taxes they are the lowest (see Table 2) thanks to the overall design of their welfare system and its redistributive capacity. In fact, income transfers like child and family allowances, unemployment benefits, and in-kind services keep large parts of the population well above the poverty line. This welfare model is the most generous one and allocates the highest share of the national GDP (see Table 2) to social expenditure, maintaining a high degree of social consensus, as both social and economic results are positive. Reforms in recent years have been approved in order to keep social expenditure under control, but the overall system configuration is – viewed comparatively – still the most generous and encompassing.

(3) The Corporatist welfare system is characterized by social policies conceived in a meritocratic way: contributory schemes reproduce the socio-economic status that families achieve in the labour market through the position of the breadwinner (Gallie and Paugam 2000). The family is the prevailing social agency and, constant with this, it is strongly supported in its caring role by specific targeted in-kind and monetary state provisions (active subsidiarity). Reciprocity is an important integration and resources allocation form (see the ‘y’ axis in Figure 1). In fact, even though the state intervenes extensively, this intervention is mediated by the role of the family. Dependency from the market is higher than in the social democratic model, but lower than in the liberal one. Bismarckian Germany (with Austria, France, Belgium) is a clear example of this welfare system. In-kind measures may vary at the Länder level like activation measures, even though within a frame of guaranteed rights. Governance is regulated by corporative arrangements with a strong relevance of negotiation among intermediate (often non-profit) bodies and public authorities.

In the scholarly literature, this model is often conflated with the familistic one (Esping-Andersen 1990, 2000). Indeed, they share the double logic underlying the way in which social risks are managed. Contributory based benefits on the one side, and the subsidiarity principle regulating the relationship between the state and the family on the other. However, the role of the state is quite relevant in making the difference in terms of generosity and services.⁷

(4) The Familistic welfare system, like the corporatist model, conceives social policies in a meritocratic and fragmented way, but it is less generous and very unbalanced in the provision of monetary benefits, which prevail against in-kind services (Kazepov and Sabatinelli 2006). Fewer resources are targeted to family policies (passive subsidiarity) and to other contributory and means tested schemes. The consequence is that the family is overloaded with social caring responsibilities without or with few institutional resources provided by the state (Naldini 2003) (see Figure 1 and the high scores of the familism index).

All south European countries (I, ES, PT, GR) show a high degree of local variation. Policies are highly segmented and targeted to particular categories. Dependency from the market resembles that of the liberal model with the exception that families reduce this dependency (through low divorce rate, low single households, etc., see Table 2) more extensively. As far as social assistance schemes are concerned, the Familistic welfare system is the most problematic one. The level of benefits is much lower and local measures and formal entitlements do not always guarantee payment. Local differentiation and discretionary power of social workers are much more spread because of heavy budgetary constraints. Reforms in Portugal and Spain at the end of the 1990s attempted to modernize social assistance schemes connecting them to national framework laws, but they failed in homogenizing access criteria and provision. This is true, in particular, for activation (or insertion) policies and the respective governance arrangements involved, which are in many cases regulated by particularistic and – sometime clientelistic – principles with a strong place for private (also non for profit) interests. There is, however, no legally enforceable social right. Income inequalities and poverty rates are higher than in the other European countries. The familistic welfare system has, in fact, a low redistributive capacity and is unable to prevent poverty. The poverty rates observed before and after social transfers in Italy indeed show the lowest difference among the countries considered (see Table 2).

(5) The welfare system in transition countries is not yet a consolidated model with clear characteristics. Both the conditions producing vulnerability and the institutional frame aimed at contrasting them experienced a dramatic change since the 1989. Most Central European and East European countries belonging to this model underwent a deep structural change in the economy with sharp GDP decreases followed by high increases. The reforms implemented in the last decade to accompany these changes and to contrast its potentially negative impact have ambivalent consequences. Countries like Poland, for example, give a more important role to market regulation (Starega-Piasek et al.2006), while others like Slovenia are investing more in coordinated market and social policies. The starting basis, however, is a quite homogeneous distribution of income with – in most cases – below average inequalities. Yet, the dynamic of change and the impact of the policies adopted in the last decade will bear their consequences in the coming years. First signals come from the greater (e.g., Slovenia, Czech Republic) or lesser (e.g., Slovakia) ability of policy transfers to reduce the risk-of-poverty rates significantly (Atkinson et al. 2002). Becoming part of the European Union implied for these countries, starting even before 2005, an important input for reforming social policies towards institutional designs closer to those of other continental European countries. Governance arrangements follow these heterogeneous lines and provide examples of all forms of interaction among actors.

Figure 1 synthesizes what has been briefly sketched out in the short descriptions of each welfare system using a familism and a statism index.⁸ Cross tabulating the two indexes the European social models clearly emerge: in Europe it is either the family or the state which take on social responsibilities. This alternative is not without effects on the rescaling processes and the new governance arrangements, as the family lacks the state's redistributive capacity.

Considering the fact that different institutional contexts tend to pre-structure both the impact of the subsidiarization processes and the directions of change, in this section I would like to address the criticalities emerging from these processes in Italy as a specific case in which consequences are particularly extreme.

The pioneering literature on processes which can be connected to subsidiarization (Walker 2002; Rodríguez-Pose and Bwire 2003; Jones 2005; Rodríguez-Pose and Gill 2005; Keating 2006; Morgan 2006) identifies the following emerging criticalities:

• a.

  the territorial coordination of the actors involved in the policy: the multiplication of both private and public actors at the different territorial levels opens up – in case of inadequate institutional agreements – opportunities for discretion, of potential conflicts and of a policy implosion (reciprocal vetoes, inability of taking relevant decisions, policy stagnation, …);

• b.

  the institutionalization of sub-national disparities: the increasing localization of the decision-making processes tends to consolidate differentiated practices legitimizing the development of local models which might further differentiate along economic differentials; and

• c.

  the accountability of the decision-making process: the multiplication of actors and their territorial fragmentation tends to weaken the democratic control over the access criteria of actors into the political arena (Crouch 2003).

These criticalities are related one another and tend to be particularly problematic in the Italian case due to the structural socio-economic and socio-political divide, which consolidated over decades at the sub-national level. Table 3 shows the striking differences existing between the northern and southern regions. Unemployment rates and income inequalities display extremely high values in southern regions and this situation is worsened by the fact that vis-à-vis a low share of population (slightly more than 30 percent) living in southern regions, about 70 percent of poor families live there (2005).

These differences are reflected in the highest coefficients of territorial variation at the NUTS-2 level (Regions) in Europe (see also Table 2) for employment and unemployment rates. However, it is not only the market which contributes to this outcome but also the pre-structuring role of the Italian welfare system and the reforms which have been implemented in the last 25 years which make the three criticalities identified above particularly relevant for the Italian case.

The first critical issue relates to the synergic effect of two relevant reforms of the centre-left government which have been approved in 2000 and 2001 and implemented later by the centre-right government. The first reform was a national framework law on social policies (nr. 328/00), aiming at radically re-organizing a policy area which, after the transfer of the competence on social policies to the regions in 1977 (dpr 616/77), produced an institutional vacuum and a growing policy differentiation at the sub-national levels; one without common national guidelines the regions developed their own regional welfare system (Kazepov 1996; Fargion 1997). The new framework law provided guidelines attempting to keep the differentiation among territorial levels under control, and developing a complex and innovative system of multi-level governance involving different scales and different actors also foreseeing homogeneous access criteria and benefits (Bifulco and Vitale 2006).

The second major reform introduced the principle of vertical subsidiarity by the Constitutional law nr. 3 of 2001. Regions became the main actor overturning the previous distribution of powers and acquiring legislative power over all subjects not explicitly competence of the State including social policies. The new context, emerging from the interaction between the two reforms, foresees strong institutional relations among the different territorial levels, in particular through:

• 1.

  the widening of the matters in which the state (legislating on principles) and the regions (legislating on implementation) should coordinate their activities;

• 2.

  the definition of minimum common standards (livelli essenziali) as transversal issues requiring a coordination also on matters which are exclusive competence of the regions; and

• 3.

  the equalization of all territorial levels of government and the dismissal of the hierarchical order foreseeing the primacy of the national level versus the regional and municipal level.

Lacking forms of institutionalized coordination and conflict resolution vis-à-vis the deep changes occurred, brought about in the last few years growing conflicts among scales (Bartole et al.2003; Vandelli 2005). This paralyzed not only the implementation of the constitutional changes, but – more generally – the implementation of policies tout-court, in particular the planning and management of policies at sub-national levels. A clear indicator for this paralysis are the claims of the nation state and the regions at the Constitutional Court: between the constitutional reform 2001 and 2004 claims raised up to 160 of which more than 90 percent came from centre-left governed regions against the centre-right national government (Unioncamere 2005: 265–9).

The institutionalization of sub-national disparities is related to the reforms mentioned above, but started in Italy after the first wave of decentralization at the end of the 1970s when the Presidential Decree 616/77 decentralized competences on social policies to regions. This decentralization took place foreseeing a national framework law and financial compensation mechanisms to keep intra-national differences under control. However, the framework law only came with the 328 framework law in 2000. Up to then, most regions started to design their own framework laws on social services and policies, producing forms of citizenship segmented along territorial lines and exacerbating already existing differences (Fargion 1997, 2004; Mirabile 2005; Kazepov 2006; Ranci 2006). The new framework law was considered a path-breaking novelty aimed at re-equilibrating the tendencies towards the territorial differentiation of rights in a frame involving the different territorial levels not only in the management and implementation of social policies, but also in their design. The new framework law went even further by involving most stakeholders in the policy design. In practice, social policies are now defined through Piani di Zona, i.e., concerted Area Plans designed by the stakeholders of the specific areas (the Ambiti) into which regions are divided (Kazepov 2006). Generally, these areas include one larger municipality or a group of associated smaller municipalities and all relevant actors (including third sector and voluntary associations …) active in one territory.

This major change has been deeply undermined by the Constitutional Law 3/01 mentioned before, which allowed regions to regulate social policies autonomously. Moreover, some relatively innovative measures that the framework law introduced have never been implemented. The Reddito Minimo di Inserimento, a social assistance scheme aimed at providing a last safety net balancing activation and a guaranteed minimum income, is one of the most striking examples. Municipalities are now not even obliged to foresee minimum income measures.

The institutionalization of sub-national disparities in terms of rights and duties has been paralleled by the increasing divide – or even de-synchronization – between policy rescaling and financial rescaling. On the one side there is a relevant devolution of competences at sub-national levels of government starting from the second half of the 1990s onwards. On the other side – in the five years of the Berlusconi government – the transfer of resources from the centre aimed at covering the decentralization costs dramatically declined (see Table 4). This sharp decline of state transfers has been accompanied by increased taxation autonomy only nominally. In fact, the share of own revenues increased not so much because the flows of incoming resources increased, but because the reduction of state transfers changed the relative weighting of municipalities’ own revenues. Moreover, municipalities have been blocked in their taxation autonomy, granted by the new constitutional reform in 2001 and the formally adopted fiscal federalism (Guerra and Zanardi 2006: 12). The critical issue emerges from the joint effect of this blockade and the reduction of transfers from the state, impeding municipalities to compensate the loss of state transfers increasing their own taxes. As a consequence existing sub-national differences institutionalized, with extremely negative consequences on the existing north–south divide.

As Table 4 clearly shows, northern regions and municipalities can rely on higher fiscal capacities, while southern regions and municipalities are more dependent from state transfers. This situation is worsened by the fact that national compensation schemes foreseen by the constitutional reform of 2001 have not been yet implemented⁹ (Pitzalis and Zanardi 2006). The same is true for the definition – at national level – of minimum common standards. The perverse effect of this situation is that rights (and duties) do not depend so much from the specific conditions of need; they rather depend on the place where the person is born. This tendency spreads around in Europe but with a weaker impact than in Italy in terms of sub-national inequalities and already with some counter reforms addressing the inequality of the subsidiarization processes of social policies (for instance the reforms in Sweden in the 1990s and early 2000s).

As far as accountability is concerned, in Italy the relationship between public bodies and the third sector (Ascoli et al.2003: 180) sees an overrepresentation in the policy-making process of professionalized and well structured organizations. Smaller voluntary associations and less organized cooperatives are often marginalized and there is the concrete risk of polarization within the third sector, also in terms of a balanced representation within the public decision-making processes. Also from this point of view, therefore, we witness the concrete risk of widening the institutionalization of the existing sub-national differences. This occurs also in relation to the legitimating principles on whose basis actors interact, which are increasingly regionalized and translated into regional framework laws.

In this contribution I have tried to show that welfare reforms in Europe are characterized by a process of subsidiarization of social policies. I also maintained that in order to understand the specificities of this process, we have to understand the regulatory contexts within which it takes place: similar processes of change filtered by different regulatory frameworks might produce divergent impacts.

Within this frame, sub-national bodies (municipalities, but prominently regions) become increasingly important. The strong accent that recent reforms put on devolution, decentralization and active welfare policies has provided them with new regulatory powers which, in a framework of overall fragmentation, brought about the need for coordination of the increased number of different actors. The degrees of autonomy regions have and the resources at their disposal, however, still very much depend on the overall regulation at the national level. In fact, social policies retain a double territorial nature. They are both local and national: passive policies (e.g., unemployment benefits) are still defined mainly at the national level, while activation and in-kind policies are defined increasingly at the local level. It is for this very reason that the nation state's influence on local policies is still relevant, in particular in relation to redistribution, which still plays an important role in Europe.

In this sense, the new forms of governance may well be differentiated and partly fragmented, but as long as relevant resources are regulated and redistributed at the national level – and they are regulated in most European countries at the national level – the degrees of coherence with national welfare systems are higher than expected. Also fragmentation and polarization are lower than in other countries, like the United States, where the low level of policy intervention exposes people in need to the increased speed of change of the market. The market changes faster than political redistributive institutions, which are more resilient. This resilience, however, offers sub-national bodies the opportunity being actors of institutional innovation. Being bold we could say that sub-national levels of regulation are – once again – laboratories of how social policies construct citizenship, social inclusion and participation (Le Galès 2002; Kazepov 2005). The real challenge, in fact, plays out in the definition of who is included and who is excluded. From this point of view, the Italian case provides an interesting perspective into the potentially negative consequences of the subsidiarization of social policies process: a differentiated landscape of rights (and duties) which shape the unequal citizen.

I would like to thank Marco Arlotti for his help and for organizing and calculating the data for all tables. Stefania Sabatinelli should also be thanked for her comments. The anonymous reviewers helped focussing the contribution, forcing me to clarify some key conceptual issues. The usual disclaimers apply.

Yuri Kazepov is Professor of Urban Sociology and Compared Welfare Systems at the University of Urbino. In 1995–1996 he was Jean Monnet Fellow at the European University Institute and in 1998 Visiting Professor at the University of Bremen. He is a founding member of the Network for European Social Policy Analysis (ESPAnet) and the Vice-president of RC21 of the International Sociological Association. His fields of interest are urban poverty, social exclusion, citizenship, social policies in a compared and scalar perspective, urban governance. On these issues he has been carrying out research and evaluation activities for the European Commission and the Italian National Research Council. Among his most recent publications are, as editor, Cities of Europe. Changing Contexts, Local Arrangements and the Challenge to Social Cohesion (Blackwell, 2005) and, as co-author with Domenico Carbone, Che Cos’è il Welfare State (Carocci, 2007).