Social services , welfare states and places : an overview

Social services – publicly funded social services – have come a long way in Europe. From a family concern, charitable endeavour or employmentrelated benefit, in the span of a century they have been to a large extent taken over by the state and have become – at least in principle – a right extended to all citizens. This remarkable progress was part and parcel of the construction of national welfare states, although it occurred with different timings and intensities across countries and service fields. Compared to other forms of welfare provision (such as social insurance schemes) and other well established social services (such as education and health), however, many care services and services for the social inclusion of vulnerable groups have not been fully developed and still have a weak social status. Moreover, many of these services have experienced relevant changes over the last three decades and are threatened by the austerity measures implemented throughout Europe as a consequence of national fiscal difficulties and the financial crisis of 2008. In this chapter I review the theoretical and conceptual discussion that informed the COST Action IS1102 SO.S. COHESION – Social services, welfare states and places and I put forward the building blocks of our analytical approach. In the first section I position our topic – social services and their restructuring – within the contemporary debates on the service economy, the welfare state, social and territorial cohesion, as well as the post-Keynesian restructuring, highlighting their specificities and key social and economic implications. In the second section I stress the importance of a timeand space-sensitive approach to analysing changes and I review the concepts of welfare ‘regimes’ and ‘models’. In the third section I ‘unpack’ the main restructuring trends discussed in the literature, highlighting their features and implications as well as a number of key conceptual tools. In the light of this overview, I conclude posing a number of questions as to what we can learn from the national and regional trajectories and case


INTRODUCTION
Social services -publicly funded social services -have come a long way in Europe. From a family concern, charitable endeavour or employmentrelated benefit, in the span of a century they have been to a large extent taken over by the state and have become -at least in principle -a right extended to all citizens. This remarkable progress was part and parcel of the construction of national welfare states, although it occurred with different timings and intensities across countries and service fields. Compared to other forms of welfare provision (such as social insurance schemes) and other well established social services (such as education and health), however, many care services and services for the social inclusion of vulnerable groups have not been fully developed and still have a weak social status. Moreover, many of these services have experienced relevant changes over the last three decades and are threatened by the austerity measures implemented throughout Europe as a consequence of national fiscal difficulties and the financial crisis of 2008.
In this chapter I review the theoretical and conceptual discussion that informed the COST Action IS1102 SO.S. COHESION -Social services, welfare states and places and I put forward the building blocks of our analytical approach. In the first section I position our topic -social services and their restructuring -within the contemporary debates on the service economy, the welfare state, social and territorial cohesion, as well as the post-Keynesian restructuring, highlighting their specificities and key social and economic implications. In the second section I stress the importance of a time-and space-sensitive approach to analysing changes and I review the concepts of welfare 'regimes' and 'models'. In the third section I 'unpack' the main restructuring trends discussed in the literature, highlighting their features and implications as well as a number of key conceptual tools. In the light of this overview, I conclude posing a number of questions as to what we can learn from the national and regional trajectories and case Flavia Martinelli -9781786432117 Downloaded from Elgar Online at 03/29/2019 01:41:58PM via free access studies examined in the course of our COST Action and critically summarised in this book and what the possible policy lessons could be, questions which I will then resume in the concluding chapter of the book.

THE FOCUS: POSITIONING SOCIAL SERVICES IN THE CURRENT DEBATES
Social services lie at the crossroads of several scholarly debates: the debate on the 'service economy' and the tension between 'producer' and 'reproductive' services; the debate on the welfare state, its components and 'models'; the debate about universalism and social and territorial cohesion.

Social Services Within the Service Sector
In the second half of the twentieth century, employment in the service sector surpassed that in the industrial sector in all advanced economies of the Western world, heralding the advent of the so-called 'post-industrial' society. However, the service sector includes a very heterogeneous set of activities (Martinelli, 1991;Martinelli and Gadrey, 2000), comprising both services supporting production (called 'producer' or 'business' services) and services supporting reproduction (of the labour force, of citizenship, of institutions and culture, and of the general conditions of accumulation). The latter include services such as health, education and other social services, but also activities related to government and public administration, membership organisations, culture and leisure institutions. In the scholarly debate that developed in the late 1970s and 1980s around services, a sort of disciplinary division of labour occurred between economists, who focused on producer services and their role in fostering innovation and competitiveness, and sociologists and political scientists, who focused on reproductive services -social services especially -and their role in ensuring welfare, citizenship and social justice, a dichotomy that somehow endorsed the hypothesis that social services are not productive. As I will stress later, this assumption is unfounded and social services do perform a key economic role. Defining/delimiting social services is tricky. There are broad and narrow definitions (see EC, 2010;Eurostat, 2011;BEPA, 2011;Sirovátka and Greve, 2014). In the broad definition, social services encompass not only the great variety of services provided to support the welfare and the social inclusion of people -care for small children, older people and people with disabilities, social assistance services, employment and training services, services for the inclusion of disadvantaged groups -but also education and health services, social housing and even social activities such as sport and leisure. The narrow definition includes only services for the care, protection and inclusion of children and minors, older people, people with mental or physical disabilities, substance abusers, and other vulnerable groups (minorities, immigrants, etc.). But even with services in the narrow definition, there is still an issue of boundaries, as in many instances there are overlaps among different services (such as with education in the case of childcare or with health services in the case of care for older people). In this book, we adopt the narrow definition of social services and we especially focus on the social component of care services (for small children and older people) and services for the social inclusion of vulnerable groups (such as recipients of social assistance, immigrants and minorities), with the addition of housing.
A key dimension of social services is their public character. Public or publicly supported social services have registered a spectacular growth in the second half of the twentieth century in all European countries. By taking charge of many such services, the state contributed to the defamilisation and de-commodification (Esping-Andersen, 1990) of activities previously provided within the family or purchased in the market and to the transformation of charitable activities in social rights. All these processes represented a major vector of the social citizenship principle and were characterising features of the so-called 'Keynesian' welfare state (see next section in this chapter).
However, in contrast to the widespread belief that social services are a mere redistributive tool to ensure social justice, hence a cost that has now become unsustainable, I contend that publicly funded social services perform a key economic role and actively support development. First, as stressed by Swyngedouw and Jessop (2005; see also Jessop, 1999), public social services contribute to lower the cost of reproduction of the labour force, thereby sustaining the accumulation process. By providing free or affordable education and training, health care, childcare, housing, etc. the state increases the purchasing power of workers, indirectly subsidising employers. Secondly, public social services perform a Keynesian role, since by providing jobs and salaries, they support domestic demand for goods and services (see also EC, 2010). This countercyclical role and its potential to stabilise or relaunch the economy in a time of crisis have recently been stressed by the European Economic and Social Committee (EESC, 2014). Third, public social services represent a 'social investment'. The view that social services should be considered 'productive' was first put forward by Alva and Gunnar Myrdal in the 1930s and was re-launched at the end of the 1990s with the EU 'social investment strategy' (Hemerijck, 2012). It argues that social policy should not be considered a cost but an investment Flavia Martinelli -9781786432117 Downloaded from Elgar Online at 03/29/2019 01:41:58PM via free access which contributes to lower future costs by making people better skilled and more productive. This double role of social services -contributing to social justice and to accumulation -makes it all the more important to understand change trends, their drivers and their impacts. volume). The former support the purchasing power of households in the market; the latter presume the public organisation and/or co-ordination of the actual production of services.

Social Services, Universalism and Territorial Cohesion
As a key component of the welfare state, social services are a major tool for social inclusion and territorial cohesion. Their capacity to achieve both these aims hinges on the degree of universalism their provision affords.
As stressed by , universalism is a 'deeply normative concept' (p. 1), closely aligned with both the social democracy and social liberalism traditions, that was central to the construction of the post-WW2 welfare state in many countries, most particularly in the Nordic countries, but to a certain extent also in the UK . It promoted equality and solidarity across social classes, breaking away from the paternalistic and stigmatising poverty relief measures of the time. It is the basis of the principle of social citizenship (Marshall, 1950) and the characterising feature of the Nordic welfare model (Esping-Andersen, 1990), which inspired many post-WW2 Western welfare states, albeit with different degrees of determination and institutionalisation. Besides being an ideal and a characterising attribute of the Nordic model, universalism is also an administrative principle , which significantly affects policy designs (see Mätzke et al., in this volume). But it remains a highly fuzzy and contested concept.
In the 1980s and 1990s, universalism came under criticism, from both the feminist movement (Williams, 1992), which denounced its male-biased and paternalistic stance, and other social movements, which stressed the 'difference-blind' approach of many social policies and services (Anttonen et al., 2012, p. 9), especially in what concerned minorities and other disadvantaged groups (see also Weinzierl et al., in this volume). These 'new' social movements, which replaced the traditional class-based movements (Martinelli, 2010), called for recognition of diversity, user-centred approaches, as well as users' choice and empowerment. Their claims contributed to a re-discussion of the notion of universalism (Häikiö and Hvinden, 2012), lessening some of its initial rigidities and further articulating its meaning.
In defining access, for example, a major discussion involved the tension between universalism and selectivism ; see also Thompson and Hogget, 1996), but it has become accepted that there can still be universalism in providing services to groups defined by their position in the human life cycle, such as for instance all children aged 0-3 or all older people. It is also understood that, within these categories, there can Flavia Martinelli -9781786432117 Downloaded from Elgar Online at 03/29/2019 01:41:58PM via free access be positive discrimination or selectivism, i.e. in favour of users in greater need via for instance means-tested programmes, as well as particularism, i.e. tailoring services to particular needs or cultural frameworks, and that these specifications not only can overcome the apparently irreconcilable tension between universalism and diversity, but can also be a way to make universalism more effective (Vabø and Szebehely, 2012). Taking into account the discussion above, the notion of 'universalistic social services' I adopt here implies equal access for all persons characterised by a given need, to the same level of good quality (and affordable) service, independently of gender, ethnicity, income, or place. 2 The last two dimensions are especially crucial for ensuring social inclusion and territorial cohesion. But while the socially inclusive impact of more or less universalistic service systems across social classes or income groups has always been at the core of social policy debates, the territorial aspect has been somewhat neglected or left implicit. And yet, territorial disparities in the supply of social services -either among different regions, or between urban and rural areas -exist in every country, even Nordic ones (Trydegård and Thorslund, 2001), albeit with different intensities. They are a major vector of social exclusion, which profoundly undermines the principle of universal social citizenship, however defined within each national boundary. Why such territorial disparities exist and persist depends on many structural and institutional factors. As I will stress in the next sections, they especially depend on national economic development and welfare state trajectories, but also on the 'vertical division of authority' within the state in what concerns the responsibility for social services.

The Restructuring of Social Services: Modernisation and Social Innovation
Within the welfare state and among social services in the broad sense, the services we address in this book -care services and services for the social inclusion of vulnerable groups -have remained somewhat the 'Cinderella', compared to more established social insurance provisions or to education and health services. In fact, with the possible exception of some Nordic countries, their recognition as a social right and their development as universalistic public services occurred much later and to a much lesser extent. This already 'weak' status has been in many instances further undermined by the restructuring of the last thirty years.
This restructuring -often referred to as 'modernisation' (Huber et al., 2006) -started in the 1980s, with the end of the 'golden age' of growth and welfare state expansion, as a consequence of several transformative pressures (see Starke, 2006 for a review;Ferrera, 2008;Bode, 2009 (OECD, 1981), with consequent attempts at curbing public expenditures and increasing the efficiency of welfare spending. These two types of pressures found an odd convergence in the post-Keynesian (and EU-sponsored) discourses and strategies of subsidiarity and liberalisation, heralding a season of multifaceted processes of change, often referred to as the 'neo-liberal turn ' (Crouch et al., 2001). Although bottom-up claims undoubtedly played a role, it is my contention that it is mostly top-down strategies -i.e. engineered by national governments and the EU -that have driven the restructuring of social services in the last three decades, especially after the completion of the Single European Market and the establishment of the European Union in 1992. Alongside the introduction of new market-inspired management principles in the organisation of public services and a re-scaling of government authority, generally away from central state responsibility and towards both the EU and the local level (Sellers and Lidstrom, 2007;Keating, 2009), a liberalisation of the 'market' of social services has taken place, with the entry of new suppliers -both profit and non-profit -alongside or in substitution of state-provided services (Ascoli and Ranci, 2002;Schmid, 2003;Carmel and Harlock, 2008). The onset of the financial crisis in 2008 has then exacerbated social needs, while placing further pressures on public spending and determining, in many countries, dramatic cuts in the public support of social services, through various mechanisms. (On restructuring trends, see the third section in this chapter.) Parallel to the above top-down restructuring processes, the last thirty years have also witnessed a blossoming of bottom-up initiatives at the local level, which aimed either at improving the way social services were supplied or at providing services in instances where they were not supplied or had been curtailed. These local collective initiatives -characterised by practices of users' participation, co-production and social economy -have been studied in the context of social innovation theory (Moulaert et al., 2005;Nicholls et al., 2015;Brandsen et al., 2016). The subject has become quite fashionable in the last ten years, even more so after the financial crisis of 2008, and significant expectations are being attached to 'socially innovative' initiatives as a means to reduce public outlays and involve communities in the provision of social services (Murray et al., 2010;Mulgan, 2012). However, several scholars have also highlighted how, while socially innovative local initiatives can contribute to give voice to, Flavia Martinelli -9781786432117 Downloaded from Elgar Online at 03/29/2019 01:41:58PM via free access and empower, social groups that are excluded from certain services and decision-making processes, they cannot and should not compensate for a retrenching welfare state Martinelli, 2012a).

A TIME-AND SPACE-SENSITIVE APPROACH
Having positioned social services within the current debates, I can now start unfolding the analytical approach proposed in this chapter. The restructuring of social services has taken place with different timings, intensity and features across countries and service fields, yielding different results depending on how context-specific structural, institutional and sociocultural factors have conditioned change trajectories. Therefore, to investigate changes in the organisation and supply of social services in Europe in the last thirty years, it is necessary to deploy a time-and spacesensitive approach, which involves taking on board the notions of welfare regimes (over time) and welfare models (across space), as well as the notion of path-dependency. Two main welfare 'regimes' or periods and five main welfare 'families' or models are considered. These are not new or original. I here summarise the main features of a number of ideal types from the literature for purely heuristic purposes, i.e. as tools to help us understand patterns of change over time and differences across space. In fact, being a relevant component of the welfare state, social services were established in given periods and in given forms, depending on places; and these specificities have conditioned subsequent processes of change. In other words, restructuring trajectories are path-dependent, whereby apparently similar processes can yield different outcomes depending on the starting configuration.

The Time Dimension and Welfare Regimes
In the second half of the twentieth century two main periods are generally identified in what concerns the role of the state and the forms of state intervention: (1) the period of the Keynesian (welfare) state, which for some countries started with the great depression, but is generally positioned between 1945 and 1980; (2) the period of the post-Keynesian or neo-liberal (welfare) state, which started in the early 1980s and still persists. These periods are also referred to as 'eras', 'phases' or 'waves' (Hemerijck, 2012). I prefer to call them regimes, 3 since they are characterised by distinct and interrelated sets of ideologies, discourses and principles about the role of the state, different policy objectives, strategies and tools, as well as different actors (see also Jenson, 2012). The Keynesian regime  was characterised by the dominance of the nation state on all internal and external affairs, a state which heavily intervened in both the economy -featuring the so-called 'developmental' state (Dickens, 1998), with robust policies in support of accumulation (commercial, industrial, regional policies), as well as direct ownership of productive activities and infrastructure -and in society -featuring the modern 'welfare' state, with more or less extended social protection measures and publicly provided social services.
The Keynesian state intervention was couched in a 'nation-strengthening' strategy, aimed at supporting (a balanced) national growth and at reducing internal social and geographical 4 inequalities. The role of the state was to support but also steer the market, through investment in strategic sectors. Intervention in the social domain had both redistributive and developmental aims, since social security and social services contributed to lowering the cost of the reproduction of labour, hence indirectly supporting accumulation (Jessop, 1999;Swyngedouw and Jessop, 2005). This was the 'golden' period of welfare state expansion, when social protection, from a private or charity affair, became a 'right' (Jenson, 2012). Although heavily oriented towards supporting the 'male breadwinner' model, the Keynesian welfare state held -in principle -a universalistic approach, i.e. aimed at providing free public services for all, albeit to different extents and with different temporalities depending on countries. This approach was based on a class compromise and a social pact between the two main collective actors of that time: capital (represented by national corporate champions) and labour (represented by national unions), with the state acting as mediator. The state itself was typically Weberian, i.e. a hierarchical bureaucracy, with formalised accountability processes (Jenson, 2012).
The neo-liberal regime (1980s onward) set in after the first major post-WW2 economic crisis of the mid-1970s, the slowdown of growth rates and the incipient fiscal difficulties of Western states. With the Reagan and Thatcher administrations in the 1980s, neoclassical economics resurfaced and supply-side policies gained consensus, finding a concrete political outlet in what has come to be called the 'neo-liberal' paradigm (Hemerijck, 2012). In this new phase, the market was viewed as the best allocation mechanism and the state as an encumbrance to the free working of the market. Liberalisation, deregulation and privatisation became the key principles guiding the restructuring and retrenchment of state intervention, in all domains.
By the end of the 1980s, the new paradigm had become hegemonic throughout the Western capitalist world. A major driver of its diffusion in Europe was the acceleration of European integration following the Single European Act of 1986 and the completion of the Single European market further strengthened the neo-liberal principles, undermining the sovereignty of the national state in several economic policy areas and enforcing a number of corollaries of the market liberalisation strategy, such as the privatisation of many state-owned activities and infrastructures, now viewed as unfair competition (Schiek, 2013).
In what concerns the welfare state, the 1990s witnessed major changes in social policy and in the public supply of social services, as documented in a now quite large body of literature (see, among many others, Crouch et al., 2001;Huber and Stephens, 2001;Ascoli and Ranci, 2002;Taylor-Gooby, 2004). Ensuring social protection to (almost) all was deemed no longer feasible. Universal entitlements were considered an unsustainable burden on the public budget and the need to increase the responsibility of individuals and families was asserted. In the domain of unemployment and poverty an 'activation' strategy was deployed, attempting to steer social assistance towards employability. In the domain of in-kind social services, the imperatives of liberalisation and competition gained ground, reinforced by the growing financial difficulties of many a state: new public management practices were introduced as well as outsourcing to private providers; the service 'market' was liberalised, allowing the entry of private providers -both profit and non-profit -in services that in many nations had been an exclusive state monopoly; competition among providers and free choice principles were adopted. In some countries, greater responsibilities were given to local administrations, deemed the best level for efficient governance.
The periodisation above is obviously not as clear-cut as described. The shift from one regime to the other was differently timed, depending on countries and regions. In some 'latecomer' Southern European countries and regions, for example, a number of universalistic welfare state measures were actually introduced in the 1980s and 1990s, i.e. during the neo-liberal phase, albeit many of these measures were not adequately funded (León and Pavolini, 2014).
Starting in the early 2000s, a possible third phase, inspired by the 'social investment' approach -i.e. a new set of principles made explicit in the Lisbon Council of 2000 -has been identified by a number of scholars (Morel et al., 2012;Hemerijck, 2012), although it would only apply to the welfare domain, rather than to state intervention in general. The new approach, which found political legitimacy in a series of European documents and directives (see Goméz-Barroso et al., in this volume), allegedly marks a setback of the neo-liberal paradigm and the return of social policy  (Hemerijck, 2012). It reverses the neoliberal idea that social policy is unproductive, arguing that it represents an investment, and hence a productive factor. It also recovers a key role for the state in lessening the ('new') social risks brought about by structural and social changes, addressing needs from a 'life course' perspective and stressing the necessity to start investing very early in the life course of citizens, in order to create a more competitive work force. At the same time, it maintains the goal of changing the welfare state from a passive benefit system to an 'activating' mechanism geared to 'capacity building' and it confirms the need for mobilising individual and family responsibilities. Whether the social investment approach can be considered a fullyfledged regime remains questionable. As stressed by Morel et al. (2012), the social investment is a 'perspective', still in search of political consensus, rather than a settled paradigm. Jenson (2012) notes that the strengthsand weaknesses -of the approach lie precisely in its ambiguity, which makes it acceptable to many parties, but also makes it rather indeterminate. Moreover, it should be noted that, despite its rhetoric, the social investment strategy does not really apply to groups such as older and other vulnerable people, except, perhaps, in terms of delaying care costs.
The financial crisis of 2008 and its aftershocks certainly represent a major (turning) moment in the evolution of Western welfare, although it may not necessarily mark the beginning of a new regime. Its impact in Europe has been significant, but diversified. Almost everywhere it has dampened the social investment approach and legitimised a relapse into, and an aggravation of, neo-liberal recipes. As such it has affected existing trajectories, sometimes by accelerating ongoing trends, other times interrupting set courses. In some countries, its effects have been dramatic, notably in a number of Southern European nations where drastic cuts in public spending and social services have curtailed even long-established basic services such as health care (Petmesidou and Guillén, 2014). But important cuts in the public support of social services have been observed also in other countries such as the UK or some Nordic countries . Two main effects seem to emerge, as also witnessed in our COST Action: first, a generalised slowdown and more selective application of the social investment strategy, privileging social services that have more potential for capacity building and growth (e.g. early childcare or 'active' labour market services); secondly, a new cleavage between the North and the South of Europe, after a period of relative convergence in the form and extent of public social services (see Martinelli,Chapter 19, in this volume).

The Space Dimension: Welfare Models and the Multiscalarity of Governance
The chapters in this book investigate changes across several European countries and, in some cases, different regions of the same country. The spatial is thus a very relevant analytical dimension, calling for a spacesensitive approach, which in turn involves taking into account both place and scale. In fact, national and regional trajectories are strongly conditioned by place-specific factors -i.e. the socio-economic and institutional context of any given place -and the geometry of the state -i.e. the articulation of authority among different government scales.
In what concerns place, the first systematic attempt to pin down and classify the diversity of welfare state forms across Western countries was carried out by Esping-Andersen in his Three Worlds of Welfare Capitalism (1990). In that seminal work, he identified three basic models: 5 (a) the Social democratic or Scandinavian model; (b) the Liberal or Anglo-Saxon model; and (c) the Conservative/Corporatist or Continental model. Since this first taxonomy, the debate has evolved, 6 different groupings have been proposed, and two additional models have been identified: (d) the Familistic or Mediterranean model; and (e) the Transition or Central-Eastern European model. I do not intend here to address the methodenstreit surrounding the nature and validity of such taxonomies. 7 The grouping of countries is strongly contingent upon which components of the welfare state (social insurance, social policies) and forms of support (cash transfers, in-kind benefits) are actually considered; moreover, these groups/models are generally well represented by one exemplary country, whereas most of the others do not so perfectly fit the one single model and are often borderline or 'shape-shifters'. Nonetheless, it is useful to sketch these models as 'ideal types' and loosely categorise countries in such welfare 'families' 8 as a heuristic device to better contextualise changes and understand differences among nations and regions.
The description of the five 'families' that follows is a synthetic blend of several representations and discussions (Ferrera, 1996;Anttonen and Sipilä, 1996;Arts and Gelissen, 2002;Bambra, 2005;Kazepov, 2008;Jenson, 2012;Schieck, 2013). 9 The main taxonomic parameters retained here are the extent of public support, the proportion between cash benefits and in-kind services, the universalistic vs. contributory blend, the providers mix (state, market, community, family), the ideology about women and their position in the labour market, and the type of industrial relations.
In the Liberal or Anglo-Saxon family (represented first and foremost by the United States, but also by the UK and Ireland in Europe), 10 the state had generally a weak role, relying significantly on the market, whereas social risks were individualised and relieved especially through meanstested cash transfers. The inclusion of women in the labour market was not actively encouraged (e.g. through the public provision of childcare). The model was also characterised by rather adversarial industrial relations.
In the Social democratic or Nordic family (represented by Sweden especially, but generally extended to Iceland, Norway, Finland and Denmark) social risks were afforded through the (quasi universalistic) provision of in-kind public services, and women were strongly encouraged to participate in the labour market. There was also significant cooperation among social partners (trade unions and employers' organisations) in the organisation of welfare.
The Corporatist or Continental family (Germany in primis, but also Austria, Belgium, the Netherlands and sometimes France 11 ), was strongly reliant on contribution-financed and state-regulated social insurance schemes (the so-called 'Bismarckian' model), i.e. strongly anchored to the employment status (of 'male breadwinners' and their families), whereas women's employment was not encouraged. It was also characterised by cooperative industrial relations, with sector-and/or nation-wide collective bargaining. The role of community organisations -mutualistic, philanthropic, religious -was also very strong in providing in-kind social services as a 'collective solidarity actor' (Jenson, 2012), often with public support.
The Familistic or Mediterranean family (Italy, 12 Spain, Portugal, Greece, with the possible addition of Malta, Cyprus and Israel; see Gal, 2010 for a discussion on this) was added later to the taxonomy. It was first conceptualised by Ferrera (1996) and further elaborated by others (see Gal, 2010 for a review). It was characterised by a distinct mix of features: 13 relatively recent democratisation and industrialisation, weak state traditions (with relatively inefficient -often clientelistic -bureaucracies), a 'Bismarckian' model of social insurance, a historically residual role of the state in social assistance and only recent public engagement in developing modern welfare provisions (Ferrera, 1996). In fact, countries in this family are 'latecomers' in what concerns the public provision of universalistic social services, many of which were introduced during the neo-liberal phase in the 1980s and 1990s, albeit often underfunded (Da Roit and Sabatinelli, 2013;León and Pavolini, 2014;Petmesidou and Guillén, 2014). But the most distinctive feature of the model was the strong reliance on the family as main provider of care services (Leitner, 2003).
The Transition or Central and Eastern European family (Poland, Slovakia, Czech Republic, Hungary, Romania), was also conceptualised later. This is the least homogeneous family 14 and it groups countries that, beyond their diversity, have common roots in their former socialist and authoritarian welfare institutions. These institutions have undergone a rapid transformation in the wake of the transition to a market economy and accession to the EU, which has put strong pressures on the decentralisation, de-institutionalisation and marketisation of social services (Koldinská and Tomeš, 2004;Fenger, 2007;Hacker, 2009).
As will become clear throughout the book, these models -especially the first three, which were defined with reference to the 'Keynesian' welfare regime -have profoundly changed over the last thirty years. But it remains useful to keep them in mind as 'starting' configurations, to better appreciate changes and differences.
The spatial dimension of our analysis is not limited to place; it also addresses scale. A key feature of welfare systems, which strongly conditioned their trajectories, is the geometry of the state, i.e. the role of, and the relations among, different scales of government, notably, between the central state and the local administrations. On this aspect too, there is a rich debate (see Bennett, 1993;Goldsmith, 1996;Kazepov, 2008;Rauch, 2008). As I shall stress in the next section the 'vertical' division of authority among different levels of the state -central, meso-and local -in what concerns the regulation, the funding, the planning and the provision of social services, as well as the degree and tradition of local governments' autonomy, significantly affects both the features and the impacts of the restructuring of social services.

Path-dependency and Restructuring Trajectories
The time and space analytical dimensions just described merge in the notion of path-dependency. This concept is particularly useful when addressing structural and institutional changes (Martinelli and Novy, 2013;CAP Martinelli and Sarlo, 2014), even more so in the case of social services. As many chapters in this book will show, national, regional and local restructuring trajectories in the last three decades have been strongly conditioned by the timing of reforms and the pre-existing institutional contexts. In other words, the way national, regional and/or local service systems have been transformed has been dependent on when the neo-liberal restructuring started and which geometry of the state, welfare institutions, socioeconomic structure and actors, as well as sociocultural norms characterised each context. As stressed by Jenson (2012, p. 62), changes are promoted by specific actors (organised interest groups, political forces, international think tanks, etc.), but they are 'grafted' onto specific welfare systems, which are, in turn, 'historically located and rooted'.
Path-dependency also explains why apparently similar strategies may have yielded very different results, depending on the context they were implemented in. For example, the introduction of outsourcing, de-institutionalisation or cash allowances has had very different impacts in systems characterised by a strong tradition of universalism and direct public provision of services (see for instance Anttonen and Karsio, in this volume;CAP Jensen and Fersch, 2013), compared to systems characterised by a historically residual role of the state or an authoritarian state tradition (see Kröger and Bagnato; Kubalčíková et al., in this volume).
There can also be instances of 'path-breaking', although these are more rare and difficult. One such moment has certainly been the implementation of the Single European market and the enforcement of liberalisation in 1992; another has been the precipitation of the financial crisis in 2008, which in many places -including a number of Nordic countriesinterrupted set courses and/or precipitated abrupt changes. It remains to be seen, however, whether such shocks have been absorbed in force of the resilience of existing institutions or have set entirely new courses.

UNPACKING THE RESTRUCTURING OF SOCIAL SERVICES: MAIN TRENDS AND KEY ISSUES
Having underscored the importance of a time-and space-sensitive approach, I now turn to the chief focus of this book and 'unpack' the main features and key analytical dimensions of the restructuring of social services, as they are discussed in the literature. Several concepts have been mobilised to address the changes that occurred in social policy -and in the organisation of social services -since the 1980s, albeit their meaning and definition often overlap. The most used term is restructuring, which does not point to any specific direction, but implies structural changes in the way the welfare state and its social services components are organised. Reform (Taylor-Gooby, 2004;Bahle, 2008) and recalibration (Ferrera et al., 2000) are also used, similarly implying a general remodelling of the welfare state as a response to structural changes and the emergence of new social risks. Another term, often used in EU documents, is modernisation, which gives a positive ring to processes aiming at making the provision of services less bureaucratic, more efficient and innovative (Newman, 2001;Cochrane, 2004;Newman et al., 2008). More normatively charged is the concept of retrenchment (Pierson, 1994;Clayton and Pontusson, 1998;Starke, 2006), which implies a reduction in the support provided by the state to certain social policies and services. Then, of course, there are more clear-cut terms -which also specifically apply to public social service -such as privatisation, which refers to the growing (re-)involvement of private providers, be they the family, the market or the community (Leibetseider et al., in this volume), and marketisation, which implies a shift towards a market logic and market providers Anttonen and Meagher, 2013;Anttonen and Karsio, in this volume).

Main Processes and Directions of Change
Out of this literature, I have identified and mapped three broad interconnected processes and directions of change since the 1980s: (1) changes in the extent and form of public support to social services, generally featuring what I call a disengagement of the state; (2) changes in the 'vertical' division of responsibility for social services, i.e. a re-scaling of authority among different levels of government, generally away from the central state; and (3) changes in what I call the 'horizontal' division of responsibility among public and private providers of social services, in the direction of a more diversified providers mix. Each of these main trends was driven by several interrelated, transformative pressures -structural, institutional and ideological (Jensen, 2011) -acting on both the demand and the supply side of social services (see Table 1.1). They, in turn, involve different, often overlapping, sub-processes and features.

Changes in the extent and form of public support towards a 'disengagement' of the state
In this first 'family' of changes three main concurrent processes can be grouped: (a) a relative contraction in the amount of publicly supported in-kind services; (b) the outsourcing of production to private suppliers; (c) a shift to, or preference for, cash transfers in lieu of in-kind services.
In what concerns the reduction in the extent of public support to in-kind services, it should be stressed that this process might have occurred without a decrease in overall public social spending. Indeed, as shown by a number of studies (Huber and Stephens, 2001;Castles, 2005;OECD, 2008), the curbing of public social expenditures sought by many a government in the 1980s and 1990s did not really succeed and in many countries the actual overall amount of social spending reached unprecedented peaks in the late 1990s and early 2000s, as a consequence of the tremendous growth of social needs. But even if overall social spending and, more specifically, the absolute amount of resources allocated to in-kind social services may not have changed, in many countries and service fields there has been a relative contraction. In fact, as a consequence of structural and socio-cultural changes such as ageing, intensified migration flows, greater female activity rates and more numerous single-parent families, increased unemployment and poverty, there has been a dramatic growth in needs and users of social  services, while public support has been unable/unwilling to keep pace with the growth of demands. The relative contraction has occurred through different, often concurrent, mechanisms: introducing greater selectivity in access, through for instance the tightening of eligibility criteria; increasing users' co-payments through the introduction of means-tested procedures and users' fees; reducing the 'intensity' of services per user by cutting the length or depth of the services provided to individual users. Some of these mechanisms had already been introduced in the 1990s, but the financial crisis of 2008 and its aftershocks have determined a further stress on the spending capacity of governments and, in some countries, even a contraction of expenditures in absolute terms (especially in Mediterranean ones, but also in the UK and some Nordic countries) (Saraceno, 2013). In these countries, many in-kind public services, and even cash-transfers, have been curtailed or altogether discontinued, as is well illustrated in a number of chapters of this volume (Kröger and Bagnato; Deusdad, Javornik et al.; Häikiö et al.; Mas Giralt and Sarlo).

Proliferation of community and bottom-up (socially innovative) initiatives from users and civil society
Note: * The financial crisis of 2008 is not a driver on its own, as it was the abrupt manifestation of existing structural trends and contradictions, but it has been separated here to stress its specific impacts in accelerating or changing social service reform trajectories.
Source: Author's compilation. There have been changes also in the form of public support, with a generalised disengagement of the state from the direct public provision of in-kind services. This type of disengagement includes two main processes, both of which have been justified with the aim of introducing competition among providers and granting greater choice to users. In many cases, however, it was a pragmatic way to shed operational responsibilities and reduce direct costs. The first process involves outsourcing (also referred to as subcontracting or externalisation), whereby the state entrusts the actual production of public social services to private suppliers (for-profit or non-profit) but still finances them (entirely or partly). In this case, private providers are more or less regulated and accredited by the state and might be selected on the basis of bidding procedures (competitive tenders). The second process involves a shift to, or a preference for, cash benefits, instead of in-kind services (Ungerson, 2004), in the form of allowances or vouchers more or less targeted or earmarked to purchasing specific services on the private market. In both instances, the extent of public support, at least in principle, may not change, but new providers step in. As such, these two forms of state disengagement also fall under the rubric of the 'horizontal' reorganisation of responsibility for social services, addressed in the third family of changes below.

Changes in the 'vertical' division of authority towards a 'hollowing' of the national state
Until the 1970s, most European countries -with the notable exceptions of Germany and Switzerland -were characterised by a strong centralisation of authority at the national level. Since then, there has been a generalised re-scaling of authority, first downward towards the regional or local scales, and later upward towards the EU (Hooghe and Marks, 2001), a process that has 'hollowed' the central state sovereignty in a number of policy domains and has strongly affected social services (see Table 1

.1).
In what concerns the downward re-scaling, also called 'vertical subsidiarity' (Kazepov, 2010), in many countries the last quarter of the twentieth century has witnessed -with different timings and intensity -an administrative decentralisation process (later called devolution), whereby a number of policy responsibilities, including social policy and/or services, were transferred from the central to lower scales of government (Sellers and Lidstrom, 2007), although not always accompanied by a parallel devolution of resources and/or tax levying authority. In some countries, it was the 'meso-' tier that was privileged, such as with the establishment of regional governments in Italy or autonomous communities in Spain, but the lower level of government authority, the municipality, township or commune (which in most countries already had responsibilities for social services) Flavia Martinelli -9781786432117 Downloaded from Elgar Online at 03/29/2019 01:41:58PM via free access was also concerned. This process was partly the result of bottom-up 'regionalist' or 'localist' claims for greater autonomy, but was also fuelled by the EU discourses about 'subsidiarity' and a 'Europe of regions', which sponsored the normative assumption that administrative decentralisation meant greater administrative efficiency and democracy (Andreotti et al., 2012).
In the vertical division of authority, the local/municipal level is indeed particularly important, since in most countries this is the level directly in charge of organising and delivering social services (Wollmann et al., 2010). The process by which it came to hold this responsibility, however, varies greatly. As is the case for welfare models, also local governments are involved in a 'typology' debate, with different attempts at grouping nations in function of one parameter or other (Goldsmith, 1996;Bennett, 1993). In this debate, the 'Northern' or 'Scandinavian' group (but also the UK), characterised by a longstanding 'functional' autonomy of local governments, albeit strongly regulated by the central state, is contrasted with the 'Franco' or 'Napoleonic' group (France, but also Belgium and sometimes Italy), historically characterised by a strong centralisation of authority, or with the 'Continental/Central European' group (Germany, the Netherlands, Switzerland), characterised by a more or less complex/ efficient co-governance among different government tiers, or yet with the 'Southern' group (Italy again, Spain, Portugal) where, after the decentralisation process of the 1980s, regions and municipalities now enjoy a significant legal and political autonomy, which does not necessarily correspond to a real 'functional' autonomy (Goldsmith, 1996, p. 189). This is to say that here too path-dependency counts and affects the way local governments operate. And, in fact, understanding the 'vertical' division of responsibility among government levels in static terms is not sufficient, as administrative capabilities are not established by decree. The current -quite complex -geometries of authority observed across European states are the result of processes that were embodied in, and/or grafted onto, pre-existing institutional configurations and administrative traditions and have, hence, produced different results. A case in point is Italy, where a common administrative structure -very centralised until the 1970s and later strongly decentralised -has yielded very different social service systems, depending on local government traditions and cultures (Costa, 2009;Pavolini, 2015).
This said, in contrast to the assumption that decentralised authority means more democratic and efficient governance per se, many authors argue that decentralising responsibility for social services in a context of economic crisis has been a strategy implemented by national governments to 'avoid blame' and shift the unpleasant task of managing austerity to Flavia Martinelli -9781786432117 Downloaded from Elgar Online at 03/29/2019 01:41:58PM via free access the local level (Keating, 1998;Rodríguez-Pose and Ezcurra, 2010;Bonoli, 2012;León et al., 2015;see Sabatinelli and Semprebon, in this volume). It should also be noted that, although decentralisation has been the prevailing trend, in some policy domains and functions, the state has maintained centralised control and there are even cases of re-centralisation of authority. Many cash transfer provisions linked to national social insurance systems, for instance, have remained firmly managed at the central level, albeit often through regional and local offices, in the same way as in many countries 'core' regulatory aspects concerning social services (standards, entitlements, access or accreditation) have also remained centrally defined. In some countries and services there are also instances of (re)centralisation (Wollmann et al., 2010), either via the (re)introduction of central control over some services, such as is the case of child care in Germany (CAP Mätzke, 2012) or via the enforcement of centrally decreed austerity measures in the funding of social services, such as in the UK (Lowndes and Pratchett, 2012;CAP Yeandle, 2014).
In what concerns upward re-scaling, with the completion of the Single European market and the signing of the Treaty of Maastricht in 1992, the EU has come to hold supra-national power in many economic policy domains that were once the prerogative of national governments: commercial policy, industrial policy, regional policy and, for the Eurozone, even monetary policy. As stressed by Scharpf (2002), European integration has 'succeeded beyond expectation' in economic terms, but not so well in social policy terms (p. 648), creating 'a fundamental asymmetry between policies promoting market efficiencies and those promoting social protection and equality' (p. 665). In fact, while member states have been progressively legally constrained -through 'hard' law' -by EU directives that enforced market integration, liberalisation and competition, the harmonisation of welfare systems and the establishment of a common social policy framework were prevented by the great diversity that existed, not only in terms of national capacity to pay for social transfers and services, but also 'in normative aspirations and institutional structures' (p. 645), not to mention issues of domestic political consensus. Hence the choice to regulate social policies through 'soft' law mechanisms (i.e. the 'open method of coordination') and 'voluntary' tools (i.e. performance indicators, benchmarking, peer reviewing, etc.), the effectiveness of which is undermined by their very voluntary nature.
And yet, the 'hard' rules imposed by the economic agenda of the EU have had a much more far-reaching impact on social services than the 'soft' tools in the domain of social policy. Many highly regulated and/ or formerly state-operated services (from banking, to transportation and telecommunication, including social services) were severely affected by EU rules on market liberalisation, which viewed them as unfair competition and supported processes of privatisation and/or opening to private suppliers. Although in the mid-2000s, following the White paper on services of general interest of 2004, an important discussion developed about the specificities of 'social' services within services 'of general economic interest' (EC, 2006;, which acknowledged the welfare dimension and inclusive role of social services (EC, 2007, pp. 5-7) and provided some margins to national and local governments for preserving the public provision of social services (see Gómez-Barroso et al., in this volume), many member states have moved from direct provision of in-kind services to outsourcing and/or cash transfers (see next section below).
This said, what is missing in the debate on government decentralisation and vertical subsidiarity in the domain of social services is a clear identification of 'who' does 'what'. As I shall argue later, the way the different functions involved in the production and delivery of social services -regulation, financing, planning, implementation -are attributed to the different tiers of the state profoundly affects their capacity to respond to needs.

Changes in the 'horizontal' division of responsibility towards a diversified service providers' 'mix'
A final set of changes concerns what is generally referred to as the 'welfare mix' (Evers and Wintersberger, 1990;Evers, 2005;Jenson, 2012), i.e. the relative contribution of the four main parties involved in the production of social services: the state, the family, the market (for-profit providers), and the community or third sector (non-profit organisations), which constitute the so-called 'providers diamond' (Jenson, 2015). The re-shuffling of responsibilities among these providers also goes under the name of 'horizontal subsidiarity' (Kazepov, 2008;, 'welfare pluralism' (Abrahamson, 1995) or 'new welfare governance' (Klenk and Pavolini, 2015), and has been labelled 're-mix' in the chapter by Leibetseder et al. (in this volume).
The division of labour between the state and other providers was always considered a key factor in the literature about -and classifications ofboth welfare regimes and welfare models (see section 2 in this chapter). In Esping-Andersen's analysis (1990), the main distinction was between the state, on the one hand, and market and family on the other, and the main trends observed in the period of the construction of the Keynesian welfare state were 'de-familisation' and 'de-commodification', meaning that the state was taking over responsibilities traditionally carried out within the family (especially care) or purchased in the market (by the richer classes). These processes were most prominent in the Nordic countries, which became the 'ideal-type' of the universalistic, social democratic welfare state model, i.e. a model that ensured access by all its citizens to the same (high) quality public services, independently of origin, gender, income or place. The other welfare models were somewhat described 'against' this ideal type.
Other authors, however, considered also a fourth provider, alternatively referred to as the third sector, the non-profit sector, the non-governmental sector, the civil society, or the community (Evers, 1995;Bode and Brandsen 2014;Jenson, 2015). This is certainly not a 'new' provider, as philanthropic or charitable organisations and mutual aid associations or cooperatives provided social assistance and services to the poor and/or to specific communities already in the nineteenth century (Martinelli, 2010). In some countries (e.g. Germany) they have remained a pillar of (public) social services, whereas in other countries (e.g. Nordic countries, but also France and Italy) they were progressively replaced by the state. They are now (re) gaining attention in conjunction with the development of the debate on the social economy (Nyssens, 2006) and social innovation Evers and Ewert, 2015;Jenson, 2015).
The specificity of this fourth type of provider is that it includes private organisations, hence suppliers with some degree of formal organisational structure, that are non-profit, hence having a 'social' mission and generally working for a 'community' of users, be it a neighbourhood (e.g. residents' associations), a category of workers or users (e.g. mutual aid associations or targeted support association), or a membership group (e.g. cooperatives). They thus generally have a local nature, but they may also belong to quite large 'umbrella' organisations or networks, on a national and even international level. In the current debates, these different organisations are generally lumped together, but there are profound differences within the category that need to be stressed, the main one being the difference between voluntary and non-profit organisations. Voluntary organisations do not pay their workers and generally do not ask fees or payments from their users. Non-profit organisations, even if they do not distribute profits, do hire and pay their workforce and hence need a steady flow of resources, either from the state or from users, to function. They also operate in the market, albeit not for a profit, and must often compete with for-profit providers (e.g. in competitive biddings for outsourced public services). A few chapters in this book will highlight these ambiguities and weaknesses of the third sector as provider of services (see Bode; Leibetseider et al.; Anttonen and Karsio).
All this said, the providers' mix that was in place at the height of the Keynesian regime in every country or region -whatever its configurationbegan to change in the 1980s and 1990s, when the neo-liberal paradigm set in. The key trend in this 're-shuffling' of the welfare mix has been a disengagement of the state, as described earlier. All three forms of such disengagement -contraction of public support, outsourcing to private providers, and shift to cash transfers -have involved an increased role of the other providers. Whether it was the family, the market or the third sector depended on the specific configuration of the welfare system and the restructuring strategies implemented. In fact, some of these changes were explicitly engineered by the state, such as with the outsourcing of public services or the shift to cash transfers; other changes came about 'by default' (Saraceno and Keck, 2010), i.e. as a spontaneous adjustment to the contraction or insufficient growth of public services (Da Roit and Sabatinelli, 2013; see also the chapters by Kröger and Bagnato; Deusdad, Lev et al., in this volume).
In the case of outsourcing, both market and third sector providers have benefited, depending on central and local government strategies. As shown by Anttonen and Karsio (in this volume), even among Nordic countries there are differences, depending on whether private companies were allowed -or found it convenient -to enter the public service market. In Italy, legislation on outsourcing especially favoured the third sector (social cooperatives) (Bifulco and Vitale, 2006). In the case of cash transfers, in the form of allowances or vouchers to purchase services in the private market, the effects on the providers' mix depend again on context, regulation and generosity of monetary transfers. In principle, cash transfers are supposed to enhance users' choice and, especially in the case of care activities, to give women the freedom to choose between caring in person for children or other non-self-sufficient persons (while receiving an allowance) and purchasing services in the market. But, in reality, this choice is constrained by many factors. In contexts characterised by unemployment and poverty and/or where fees charged by private providers are too high, women will tend to care in person for needy family members. In places where an informal market of privately hired caregivers has developed, in the absence of employment regulation enforcement (such as in some Mediterranean countries), cash transfers can contribute to reproducing the black market of immigrant care workers (Van Hooren, 2012;Williams and Brennan, 2012). In some places where the economic crisis has worsened unemployment and poverty, cash allowances often become a source of income for the whole household. In any case, cash-transfers generally involve either a 're-commodification' or a 're-familisation' process (Da Roit and Sabatinelli, 2013).
A major consequence of the horizontal re-shuffling of responsibilities among service providers has been an increase in complexity and, in some cases, a fragmentation of responsibility. Both involve higher transaction costs, while requiring greater coordination mechanisms, which are not always ensured. Moreover, complexity may become 'disorganisation' (Bode, 2006; see also Chapter 4,in this volume), hence reducing the quality and the reach of services. And it often reduces transparency, as information is not readily available to the most vulnerable groups, thereby limiting accessibility. But, as was the case for the vertical reorganisation of authority, in order to better understand the consequences of the new horizontal division of responsibility, one must look at 'who does what', as I will explain in the coming section.

The Main Functions Involved in Public Social Service Systems
The features and impacts of the restructuring of social services depend very much on 'who does what'. In the organisation, production and delivery of public social services, I identify four key functions: (a) regulation; (b) funding; (c) coordination, planning and monitoring; (d) production and delivery. A first division of responsibility for these functions occurs in 'vertical' terms, i.e. among the different tiers of the state. The diagram proposed in Table 1.2 -with an example from Italy -can be a useful tool to summarise and chart the role of the different government levels in the different functions, for any given social service, at any one time or place. But clearly identifying 'who does what' in 'horizontal' terms, i.e. among the different parties of the providers' mix, is also relevant when it comes to the actual production and delivery function.
The regulation function generally pertains to the state. It defines rights and duties, service standards, entitlement and access criteria, procedures for the production, accreditation, delivery and monitoring of social services, etc. Most importantly, it defines which level of government does what and which suppliers are authorised to provide services. We have already addressed the indirect regulatory role of the EU in social policythrough its agendas and 'soft' law instruments -and its rather stronger influence in what concerns social services as economic activities. But since social policy is still by and large a national prerogative, it is at this level that regulation still has the greatest impact: how much is defined at the national level -especially in what concerns service standards and eligibility -has fundamental consequences on the universalistic dimension of social services. The more nationally regulated are social services, the more uniform the supply and coverage is likely to be across places; in contrast, when the regulatory function is decentralised, territorial differences in quality and access are inevitable, especially when there are differences in economic development levels and administrative traditions. In Italy, for example, the regulatory role of the central state in the domain of social services was weak to start with, but the 2001 constitutional reform has granted full regulatory autonomy to the regional governments, thereby enhancing the already quite high regional differentiation in service systems (Costa, 2009;Pavolini, 2015). Strong (national) regulation is even more necessary when private providers are involved -e.g. in the case of outsourcing -in order to guarantee minimum requirements in terms of service standards, access, training and contractual conditions for workers. This said, the lower levels of government can -and do -concur in the regulation function, especially in what concerns regulating the actual production of services.
Funding is another key function, since the amount and provenance of financial resources affect both the quantity and the quality of services. This is still an eminently public function, although co-payments and coproduction (from users, the family and the community) are increasingly called upon. A great variety of configurations is possible, though. At one extreme of an ideal-typical spectrum can be found a hypothetically fully centralised system where resources are levied at the national level (through fully centralised taxation) and territorially allocated according to given parameters (population, needs, etc.). At the other extreme can be found a hypothetically fully decentralised system where regions and/or municipalities have complete fiscal autonomy and hence levy their own resources and fund their own services. In the former case, the central state can perform an important redistributive role, i.e. ensure that places have the resources necessary to provide social services independently of their tax base. In the latter case, the amount and quality of services would depend exclusively on the wealth of places. In reality, most countries have mixed systems and the central state retains some redistributive power. But regardless of the system, the degree of the central state's redistributive power -together with the degree of national regulatory power -strongly affect the universal character of social services (see the next section in this chapter).
The coordinating, planning and monitoring function involves the actual organisation of social services, i.e. making choices and implementing procedures to assess needs, allocate funding to specific services and territories, attribute tasks to specific actors, and monitor the supply. This function too is still eminently a state prerogative and is generally shared among different government levels, although the bulk is generally carried out at the lower echelons (the regional and/or the local levels). In some cases, the meso-level acts as the chief planning tier (for instance the Regional governments in Italy); in many other cases, the planning of services is directly entrusted to the lower level (e.g. municipalities or districts).
The actual production and delivery of social services is the final task in the system. This function is necessarily carried out at the local level, i.e. at the municipal or districts level, since social services require proximity to their users. It is in this function that the state -in those countries where it had stepped in 15 -has most 'disengaged': from direct -and often soleprovider, it has evolved into a 'commissioner' of social services (Diamond and Liddle, 2012), in the case of outsourcing, or an 'enabler' of market mechanisms, in the case of cash-transfers. In the first instance, the state retains its role of coordinator, in the second it abdicates this role to the market. There is of course also the instance of insufficient development or actual retrenchment of public social services, in which case production is carried out by the family, the market or the third sector 'by default'. Either way, it is in this function that the 'horizontal' division of responsibility is most relevant, as the family, third sector and for-profit providers are increasingly (re)entering the production of services, alongside, on behalf, or in substitution of the state.
In conclusion, key issues in the vertical division of authority -which bear on the universalistic dimension of social services -are the degree of national control over regulation and funding (i.e. redistribution) and whether the division of authority and responsibility is clearly defined. As some of the chapters of this volume will show, the restructuring of social services has seen in many cases a devolution of authority without a parallel devolution of funding capabilities (Sabatinelli and Semprebon) and a complexification -if not a blurring -of both the vertical division of authority between the different tiers of the state and the horizontal division of responsibility among providers (Bode; Leibetseder et al.), which has determined fragmentation, disorganisation, lack of transparency and reduction of accountability in the organisation of the service supply.

Key Issues in the Vertical and Horizontal Division of Responsibilities
'Who does what' -and which functions are retained by the state, at which level and to what extent -is not just an idle academic concern, since it directly affects a number of key aspects involved in the public support of social services. A first key aspect is that of universalism. As stressed in section 1 of this chapter, this was the pillar of the Keynesian welfare state and the basis of the notion of social citizenship. Even in its 'softer' version, i.e. allowing for positive selectivism and particularism, it implies the basic right for people to have access to good quality and affordable services, independently of origin, religion, gender, income or place. From the discussion above, it clearly appears that the capability to ensure both social inclusion and territorial cohesion hinges directly on how much the central state retains control over the regulation and funding of social services, ensuring a redistribution of resources in favour of the poorest and/or neediest social groups and territories. Irrespective of systems and traditions in the vertical division of authority among different levels of the state, it is still the national government -at present -that can ensure the necessary redistributive mechanisms. The challenge is then how to conjugate universalism with choice.
Another key aspect that depends on central state regulation is how much the new division of responsibility among the family, the market and the third sector is formalised, i.e. 'above board', transparent and accountable. When outsourcing and cash transfers -but also privately produced social services -are not regulated, i.e. are not subject to formal rules and monitoring, informal, 'grey' or outright 'black' market activities can develop, involving exploitative labour relations, lack of professional training and/ or neglect or abuse of vulnerable users. For example, when cash transfers are provided it makes a great difference whether they are earmarked, i.e. they can only be spent to purchase services from accredited -formalsuppliers (as is the case, e.g. in Denmark; CAP Jensen and Fersch, 2013), or they are not monitored at all, allowing for the purchase of 'informal' services from hired caregivers, often immigrant women (as is the case, e.g. in Italy; see Van Hooren, 2010;CAP Martinelli, 2012b). The same applies to third sector service providers, which can be highly regulated and integrated into the public system (as for instance in Germany) or informally set up to meet social needs not covered or regulated by the state (as for instance in Greece; see Häikiö et al., in this volume). It also makes a difference for family carers, who in some countries are formally recognised and financially supported. In other words, the more regulated and monitored are the different policy tools, the more formalised and accountable they are, also in terms of quality of service for the user and working conditions for the providers.
A final issue concerns the role of local initiatives in triggering and upscaling social innovation in the production and delivery of social services. In the diversified panorama of national, regional and local processes of change examined in the course of the Action, several innovative local initiatives have been analysed, some integrated in publicly funded (experimental) social policy programmes, but most of them of a spontaneous nature trying to address ignored social needs, new needs or needs determined by cuts in public services. They have generally involved multiple local actors, such as the civil society, non-governmental organisations, public authorities, and sometimes also private companies. Four chapters in this volume are dedicated to the analysis of some of these local initiatives (Häikiö et al.; Weinzierl et al.; Mas Giralt and Sarlo; Brokking et al.). They raise major analytical and normative questions, related to the actual socially innovative impact of such initiatives. According to the more emancipatory conception of social innovation (Moulaert et al., 2005;; see also Mätzke et al., in this volume), socially innovative initiatives must empower users, transform social relations among actors involved and bring about some form of durable social progress, beyond the individual or the community. In many of the observed initiatives, however, these outcomes were not attained. Local action merely substituted for an absent public provision of (satisfactory) services, did not involve/challenged the overall system, and remained confined at the local level, with limited societal impact and sustainability over time.

CONCLUDING QUESTIONS
Against the above background, the chapters that follow in this book have all explored the directions, features and specificities of changes in the public provision of social services in different places and service fields. They have all investigated the impacts of changes from one or more among the five 'perspectives' mobilised in the course of our COST Action (see Memorandum of Understanding, 2011): (i) cost efficiency and user satisfaction; (ii) democratic governance; (iii) social and territorial cohesion; (iv) labour market of care and social work; (v) gender. They have all addressed one or more of the following questions: • What were the prevailing changes at work in any given place or service in the last decades and in which direction? Were there any detectable new trends compared to those identified here?
Flavia Martinelli -9781786432117 Downloaded from Elgar Online at 03/29/2019 01:41:58PM via free access • How has the financial crisis of 2008 and its aftershocks affected restructuring trajectories? Are there continuities or discontinuities with regard to the pre-existing neo-liberal restructuring processes? • Has the 'social investment strategy' affected in any way the ongoing restructuring? • Are there differences in processes and directions among service fields? • How have these processes been influenced by pre-existing welfare systems and other context-specific factors? In other words, how has path-dependency played out in national and regional trajectories? • Is there convergence or divergence among national and regional trajectories in Europe? With respect to which features? • What are the effects of such changes, especially in what concerns the satisfaction of needs, users' choice and empowerment, democratic governance, social and territorial cohesion, the conditions of social work, gender? • What are the implications and challenges of these trends in terms of social policy?
These are rather ambitious questions, which this book does not claim to answer in any definite and exhaustive way, especially given the great variety of contexts and services. But despite this variety, as the editors will stress in the last part of the volume, all chapters have contributed some tesserae towards a better understanding of what is happening to social services in Europe, calling attention to a number of social consequences and policy challenges.