The horizontal ‘ re-mix ’ in social care : trends and implications for service provision

This chapter re-conceptualises the notion of ‘welfare mixes’ in social care and addresses the main changes taking place in European countries. First, we provide a typology of care provision modes that refines the so-called ‘welfare diamond’ (Jenson, 2013) and we discuss the theoretical implications of welfare mixes, which originally set out to promote equality and partnership among different providers, striving for user-centred and universal social services. However, reviewing the changes that have been introduced over the past few decades, we detect a tendency whereby re-mixes accentuate social inequalities between lower and higher income groups due to increased fragmentation and marketisation, which threatens equality among service providers and universal provision. We take a look at the main trajectories that frame and shape welfare mixes in social care. The latter are described in terms of the relative share of public, for-profit, non-for-profit or family-based providers. Our aim is to construct a conceptual map of the way responsibilities are shifting among care providers and producing re-mixes. First, we outline a framework for conceptualising welfare mixes. We then contrast this with changes that lead to re-mixes, and we develop a new typology of care provision modes. Next, we present a few illustrative examples drawn from the empirical material shared within the COST Action IS1102 SO.S. COHESION – Social services, welfare states and places to highlight such shifts in re-mixes. Finally, we attempt to discern the main social impacts that may result from different types of shifts, paying particular attention to the consequences for equality among users and partnership among providers.


INTRODUCTION
This chapter re-conceptualises the notion of 'welfare mixes' in social care and addresses the main changes taking place in European countries. First, we provide a typology of care provision modes that refines the so-called 'welfare diamond' (Jenson, 2013) and we discuss the theoretical implications of welfare mixes, which originally set out to promote equality and partnership among different providers, striving for user-centred and universal social services. However, reviewing the changes that have been introduced over the past few decades, we detect a tendency whereby re-mixes accentuate social inequalities between lower and higher income groups due to increased fragmentation and marketisation, which threatens equality among service providers and universal provision.
We take a look at the main trajectories that frame and shape welfare mixes in social care. The latter are described in terms of the relative share of public, for-profit, non-for-profit or family-based providers. Our aim is to construct a conceptual map of the way responsibilities are shifting among care providers and producing re-mixes. First, we outline a framework for conceptualising welfare mixes. We then contrast this with changes that lead to re-mixes, and we develop a new typology of care provision modes. Next, we present a few illustrative examples drawn from the empirical material shared within the COST Action IS1102 SO.S. COHESION -Social services, welfare states and places to highlight such shifts in re-mixes. Finally, we attempt to discern the main social impacts that may result from different types of shifts, paying particular attention to the consequences for equality among users and partnership among providers.

THE WELFARE RE-MIX IN SOCIAL CARE PROVISION
The 'welfare diamond' -with the state, family, community and market featuring at each of its four corners -has aroused much interest in social policy research and furthered our understanding of the complexity of care production (Evers et al., 1994;Jenson, 2013). Mixes are shaped and framed by many changes. The recent restructuring of social services in the wake of the financial crisis has, in many countries, taken the form of a retreat by the state from its role in welfare policies. The state's role has been, to a greater or lesser extent, (re)taken over by the market, the community and the family. Permanent austerity has also forced the state to seek ways to raise the productivity of services and care work. The result is a complex, differentiated -and at times contradictory -set of social care services that spans across countries and regions, across fields of social care. The concept of 're-mix' relates to Martinelli's notion of 'horizontal' division of responsibility (Chapter 1, in this volume) and is used here to describe and analyse shifts in the providers' mix, i.e. among the four main actors involved in social care: the state, the family, the for-profit sector, and the non-profit or community sector.
Since the late 1980s, 'welfare pluralism' (Johnson, 1987), also known as the 'mixed economy of welfare' (Abrahamson, 1995), was seen as an emerging principle that re-shuffled the rights and responsibilities among the market, welfare organisations and households and redefined the role of the state (Evers, 1990). It provided an alternative model for social policy at a time when the privatisation of welfare was becoming a very powerful doctrine and policy trend. 'Welfare mix' theory was based on the idea that parallel providers, as well as close collaboration among them, were needed in social service provision in order to secure the best possible outcomes in terms of the satisfaction of needs, the participation of users, and the prevention of any kind of monopoly. It was assumed that multiple providers can offer a better variety of services, generate more diversity, and constitute equal players in the field. An important aim was to make the voluntary and for-profit sectors visible in welfare production (Evers, 1995). The 're-mix' idea also connects to current discussions which stress 'horizontal subsidiarity' (Kazepov, 2010), 'new welfare governance' (Bode, 2006) and 'hybridisation' of providers (Evers, 2005). The concept of horizontal 'remix' is used to further develop approaches and concepts pertaining to the empirical dimensions of welfare mixes. Besides changing the nature of collaboration, re-mixes involve tasks being re-shuffled among actors involved in care in complex and contradictory ways. This can challenge parity among providers. In this chapter, we focus on social care services. All people need social care, most particularly at the beginning and end of their life course. Social care (Anttonen and Zechner, 2011) covers a wide range of activities that help people to cope with their daily life. Compared to health and education, though, social care is a less specialised and professionalised service, as there is a less clear demarcation between informal and formal care, unpaid and paid care (see Kröger and Bagnato, in this volume). Therefore, it serves as a suitable illustration of re-mix trends in the European context.
Our empirical material consists of case studies presented in the context of our COST Action. As these cases do not systematically represent all social care services or all European countries, we also integrated findings from other sources. However, although these cases serve to illustrate our conceptual taxonomy and re-mix shifts, they do not enable us to generalise about countries or welfare regime trajectories. To foster accuracy, we also circulated the chapter among the authors of the COST Action case studies addressed here.

FROM WELFARE MIX TO RE-MIX IN SOCIAL CARE
Before we outline the current trajectories leading to welfare re-mixes, we must first establish what welfare mixes mean for social care. The concept of 'mix' distinguishes analytically between four main actors that provide social services (Evers and Svetlik, 1993). The four providers are not perceived as mutually exclusive but as complementing each other (despite some rivalry and competition) in covering informal and formal provision, private and public sources of care, and voluntary and professional engagement (Evers and Olk, 1996;Evers and Guillemard, 2013). The notion of 'mix' underscores the connectedness of informal and formal as well as public and private care provision. Table 6.1 summarises modes of provision, relating them to the four suppliers featuring in the diamond, and further refining the taxonomy.
First, we find the family or immediate community, which constitutes the main provider of care nearly everywhere, with different degrees of involvement depending on the user group, be it small children, frail older people or newly arrived immigrants. Altruism and/or reciprocity characterise the exchange, and users are mostly relatives or close community members. Consequently, the family or immediate community forms an exclusionary system of support and cannot guarantee equality in access to care. It may lead to neglect when moral obligations cannot be fulfilled, for instance due to lack of resources or mutual respect (Evers and Olk, 1996). In the non-profit sphere, community, voluntary and other civil society-based providers are private actors that deliver services to approved categories of users needing help, sometimes members of the associations in question. They are divided into two main subcategories, the first based on voluntary, unpaid work, and the second based on more or less professionalised paid work. Users are cared for according to the organisation's targets and their own needs -an aspect that may hinder the equal distribution of services, while the voluntary nature of some of this work might hinder professionalisation. Although these private, non-profit organisations generally endorse solidarity and aim to ensure users' political and social participation and empowerment, over the past 20 years many had to transform into market actors due to increased participation in competitive tendering.
In the market sphere, we also find two main categories. On the one hand, there are for-profit commercial enterprises, which sell services directly to customers -who pay for them and enjoy the freedom to choose -or receive payments from public authorities, in the case of outsourcing. It is widely believed that private for-profit providers enhance competition, innovation and choice. On the other hand, we find self-employed caregivers who are privately hired by users or their families. Their work is more or less formalised and regulated, depending on the country.
Finally, there is the public sector, which carries out multiple tasks and plays different roles in social services and social care: it may regulate, finance and/or provide services depending on the welfare state context. The state may regulate the provision of care work, by setting for example minimum standards for public and private providers. The state can also directly provide services, which implies that the facilities are owned and financed by either the national, the regional or the local government; that those who work in those facilities are public employees; and that entitled users are citizens or residents. The state may also finance social services indirectly, via tax breaks or benefits for users, or via subcontracting or outsourcing services to private providers. In general, the state is regarded as an actor that aims to achieve social equality and the universal provision of services, but it may also inhibit private initiatives, neglect the needs of minorities, and act in a rather hierarchical and overly bureaucratic manner.
The welfare mix approach -in its optimistic version -is based on the idea that these diverse actors tend to complement each other. The various care providers do not typically possess the same strengths, and by combining efforts in different ways they can also cover care needs in different ways. This diversity can, in fact, offer a menu of choices for users and caregivers alike (Anttonen and Sipilä, 2005;Evers and Guillemard, 2013;Powell, 2007). On the other hand, boundaries between providers have also become blurred and a new hybridity has evolved which might endanger equality, universality and political accountability (Evers, 2005). Bode (2006) also underlines that all providers, with the exception of households, have experienced intensified competition and state regulation, abandoning their initial values as a result of the marketisation drive.
Seen from that perspective, the previous debate on how best to offer social provision in welfare mixes (Powell, 2007) might get lost in the re-mix process. Early welfare mix theorists criticised both the dominance of the public sector and marketisation, seeking instead a truly pluralist model that recognised the third sector as an important actor (Evers and Svetlik, ina Leibetseder, Anneli Anttonen, Einar Øverbye, Charles Pace and Signy Irene Vabo -9781786432117 Downloaded from Elgar Online at 12/11/2021 11:12:46AM via free access 1993; Evers and Laville, 2004). How do recent trends fit into the theory and practice of welfare mixes? Are re-mixes leading towards increased marketisation of care production?

SHIFTS AND TRENDS: WELFARE RE-MIXES IN THE MAKING
Besides political-normative perspectives on how the distribution of social services and care ought to be organised, an analytical framework is needed in order to describe the rationalities and roles of different actors and the shifts occurring among them. We will therefore now propose some finegrained distinctions within the most common trends in the reorganisation of tasks among the actors involved in welfare re-mixes, as defined above. A lot of attention has been paid to the shift away from public sector delivery of welfare and social services in current market economies (Meagher and Szebehely, 2013). This pertains to at least three processes: Marketisation is a major component of the current transformation of care services. It refers to a growing role of the market provider. It is, however, a multi-faceted concept and phenomenon. Here we specifically address the process whereby an increasing share of formerly public services are contracted out to for-profit providers, usually through competitive tenders or customer choice models. In this way, market actors assume a central position not only in countries that previously featured substantial public provision, but also in contexts where this was not the case.
Marketisation can take place without public sector retrenchment. Care services can be generated by motivating market providers to produce services and consumers to purchase them with the help of public money, such as vouchers, or other financial incentives such as tax rebates. It can also occur by outsourcing the provision to for-profit providers: the national, regional or local governments issue a call for tenders and diverse providers compete for a government contract to provide the service. Very often, thus, marketisation implies that for-profit providers deliver care services which are still fully or partly funded through public sources -be these taxes or mandatory contributions.
It is important to note that while marketisation can be a state-led, active policy measure, it can also just occur through inaction or drift. Active marketisation takes place through tenders and the issuing of vouchers, leading towards a 'managed' social care market. Passive marketisation occurs instead when public funding and provisions do not keep pace with increased demands and for-profit providers develop to answer needs. Familialisation implies an increase in the role of the family. The concept of 'de-familialisation' was extensively used to indicate social developments whereby functions previously handled by family members were removed from the domestic sphere and transferred to the state, market providers, or the voluntary and community sector. 'Re-familialisation' correspondingly refers to transitions whereby functions that were once removed from the family's sphere of responsibility are returned to the family (Leitner, 2003). Both forms of change can entail active or passive processes. Passive (or de facto) re-familialisation is a result of public withdrawal, market withdrawal (for example due to decreased ability to pay formal or informal market providers), or community withdrawal (if non-governmental or community organisations reduce their services). Conversely, active re-familialisation occurs when government authorities encourage the family to take on more care obligations, through, for example, the granting of allowances to parents caring for children or family members looking after frail relatives at home (Saraceno and Keck, 2011).
Whether family-based social care provision is actively supported by the state or taken for granted as a family's obligation, the provision of family care can, to some extent, be regulated by law. Regulated family care occurs, for example, when the state intervenes, through legal measures or economic and moral incentives, in how much and what kind of social care the family provides.
Communitarisation. Non-governmental organisations (NGOs) and community-based organisations (CBOs) are sometimes labelled 'community' organisations, or 'third sector' organisations. We suggest the concept of 'communitarisation' to describe increases in the role of the nonprofit sector that have occurred as a consequence of shifts within the horizontal division of labour. 'De-communitarisation' implies a declining role for NGOs and CBOs as service providers, because more care is provided either by the state, the market or family. Conversely, 're-communitarisation' can take place when the state or other actors withdraw, or do not expand fast enough to meet increased demands and the community -whether in the form of voluntary associations or non-profit organisations -takes over again.
Because communitarisation may apply to diverse organisations, we distinguish between formal communitarisation, which involves large NGOs, and informal communitarisation, which involves less formalised CBOs, based on voluntary initiatives. Recently, however, the role of third sector organisations has changed, due to the overall marketisation of care production: third sector actors are increasingly expected to compete with for-profit providers and ask for user fees. There still are, however, genuine In conclusion, marketisation, familialisation and communitarisation are processes that alter the welfare mix in horizontal terms: there are different providers in each society, and their role and status vary, as do the relations among them, also in function of governments' policies. Since the 1990s, a major market 'turn' has taken place in a number of European countries. Marketisation, in particular, is often accompanied by a reduction of government expenditure on care and, in this respect, can be distinguished from outsourcing, in which private organisations deliver social services while public authorities bear the costs.
On the other hand, the public financing of social services can also expand, as has happened in both childcare and care for older people throughout Europe in the 1990s and early 2000s (Bouget et al., 2015). New social security schemes can be established, or access to free, tax-financed social services can be introduced. It may also provide larger direct and indirect subsidies to users or private providers, for example through tax deductions for users of private services or insurance or direct subsidies for private investments in services (e.g. for the construction of private kindergartens or nursing homes).
To sum up, the overall system of service provision in contemporary Europe is based on a variety of rationalities in care production, a complex set of relations among different providers, diversity in funding and steering mechanisms, and diversification in user behaviours.

WELFARE RE-MIXES IN THE CONTEXT OF EUROPEAN SOCIAL CARE
Our empirical evidence includes examples drawn from our COST Action, which focus on specific social services in a number of European countries, regions and municipalities. We especially looked at those that addressed re-mixes and changes in terms of marketisation, familialisation or communitarisation in provision, funding and regulation. Based on our previous theoretical discussion, we have identified and grouped the observed shifts according to the taxonomy presented in Table  6.2. The table was devised as a tool to facilitate the description of how these changes played out in various cases, taking into account possible differences within countries in diverse service fields, and to help discern the social impacts of these changes. Our focus is thus on re-mix trends as they have manifested themselves through the mechanisms of marketisation, familialisation and communitarisation. The deep and extensive changes in state and local government administration and management are touched upon only in relation to these processes. This means that the focus is not on changes within the public sector -or vertical governance changes (see Sabatinelli and Semprebon, in this volume) -but on transfers of responsibility for services away from governments and towards diversified social care mixes. Since shifts between providers (state, family, market, community) are more nuanced than is generally acknowledged in the debate, marketisation, familialisation and communitarisation have been further specified, for instance through the distinction between active and passive marketisation, as proposed in Table 6.2. In what follows, we will illustrate these more specific trends.

Active and Passive Marketisation
Marketisation is a widespread ideology, doctrine and policy strategy in most European countries (Bode et al., 2011;Brennan et al., 2012;Salamon, 1993;Yeandle et al., 2012). It is, however, important to underline differences among prevailing forms of marketisation. Active marketisation is used here to refer to the active or proactive role of the state in ensuring the shift to the market. Passive marketisation denotes a growing role for market provision as a consequence of weak or retrenching state intervention in social care, often amid growing demand.
In care for older people in Denmark, for example, services have undergone a significant marketisation shift, with users regarded as customers who make informed choices, a shift justified by population ageing, fiscal stress and neo-liberal principles. The emphasis on users as customers was ushered in by national legislative reforms enacted by a centre-right government in the early 2000s, which reduced municipal autonomy. Until 2002, in fact, municipalities could choose whether they preferred direct provision or outsourcing to private providers, but in 2003, national legislation granted users the right to choose between municipal and private services. In practice, this has introduced 'tender-marketisation', where private providers have to meet specified quality standards in order to be authorised by municipalities. Private and public providers then receive the same fixed payment (CAP Jensen and Fersch, 2013).
The system is supposed to encourage competition among providers, but has led to new issues in service provision. Jensen and Fersch (CAP 2013) talk about a 'tyranny of the clock', where strict schedules prevent home help providers from responding to unexpected urgent needs. The case of Danish care for older people well illustrates how governments can be active in their marketisation efforts, particularly in promoting different choice models and outsourcing public services through tenders to for-profit providers.
In Finland, marketisation has taken place primarily through outsourcing via competitive tendering, which is now in use in most municipalities. The share of for-profit service provision in Finland has risen sharply since the early 1990s, most notably in child welfare services and intensive 24-hour care for older and disabled persons. For-profit provision in 2013 accounted for roughly half of all publicly funded services (CAP Leinonen et al., 2012;CAP Anttonen and Karsio, 2013; see also Anttonen and Karsio, in this volume) compared to the early 1990s, when the share of for-profits was close to zero.
In Austria and Germany, social service provision has always been strongly tied to public as well as non-profit providers (besides the family) (Evers and Laville, 2004). This is the case both in care for older people and childcare. In terms of care for older people, however, a shift has been observed from non-profit towards for-profit providers, as well as towards informal care-giving, albeit with some regulation (CAP Bode, 2013;CAP Leibetseder, 2016a). Austria and Germany have also introduced cash allowances to motivate citizens to purchase private services. The introduction of a universal care allowance in Austria and of a social care insurance option in Germany represent a further expansion of the public financing  CAP Leibetseder, 2016b). Thus, marketisation does not necessarily mean the withdrawal of state funding. Both countries have also stepped up the financing and provision of childcare services, as has also occurred in many other European countries (Ferragina and Seeleib-Kaiser, 2015). Both the Czech Republic and Slovakia, which have a strong tradition of state provision of residential care for older people, have embraced a marketisation approach when they joined the EU. The Czech Republic has moved away from exclusive provision by government agencies towards the participation by non-profit and for-profit providers in long-term residential care for older people. There has also been an increase in state support for informal care-giving. In Slovakia, care service provision for older people is still mainly a government responsibility, but this is changing in the direction of a plurality of providers (Kubalčíková et al., in this volume).
The United Kingdom has undergone a number of changes in its service design. Among others, it has introduced a new type of economic support for family carers. It is unlikely, though, that these indirect subsidies to private providers (including family providers) will be sufficient to reverse the overall downward trend in public funding in the UK. Although new legislation in the UK, including the Localism Act of 2011 and the Welfare Reform Act of 2012, has given local authorities new autonomy, it has become rather difficult to exercise it within the financial restraints emanating from years of centrally imposed budget cuts, as shown in the case of the municipality of Leeds (CAP Yeandle, 2014). Moreover, local citizens are now asked to approve any increase in council taxes above national guideline levels.
Southern European countries historically had a comparatively small public service sector in social and health care (Stoy, 2014;Martinelli, Chapter 1, in this volume). Only in the 1980s, increased state provision began supplementing existing residual (mostly church-related) non-profit service organisations. However, rising demands have also led to an expansion of market-and community-based care. With the 2008 financial crisis, the development both of public support and private provision has ground to a halt.
In the municipality of Reggio Calabria, in Southern Italy, the very limited development of both direct public provision and outsourced public childcare services, in parallel to a significant growth in demand, has triggered a clear process of passive marketisation or marketisation by default, with the development of an unregulated system of private for-profit or non-profit childcare providers (CAP Martinelli et al., 2014).
In Malta, a significant shift has taken place from public provision and church-related NGOs in the field of care services for older people towards market-based services. The government has not outsourced any preexisting public services to private organisations, although new residential homes have been created as public-private partnerships, and the government purchases places within private residences. Meanwhile NGO provision of meals-on-wheels was transferred to for-profit providers in 2015. And yet, public or NGO-based provision of social care in Malta has not kept pace with the increased demand, and non-subsidised market-based services nowadays fill the void (Pace et al., 2016). In many of the Action's other case studies concerning Southern Europe, 'passive' marketisation is at work, as direct and indirect public support of care services is not keeping up with demand. This kind of passive marketisation leads many users who cannot afford for-profit services to exit the formal social service systems, resorting instead to the re-familialisation of care and/or to situations where care needs are not properly met.

The Informal Marketisation of Care
In elderly care, some cases bear witness to the detrimental effect of the provision of cash benefits, low public provision of services and rather mediocre regulation and enforcement. These factors tend to lead to the creation of informal markets, where mostly informal private caregivers (usually immigrants) deliver service provision.
Public support of care services for older people in Mediterranean countries such as Spain or Italy began to develop in the 1980s, spurred by increased demand (CAP Deusdad, 2013;CAP Martinelli, 2012). The financial crisis of 2008, however, interrupted this 'catch-up' process and provoked a sharp retrenchment of public funding. Therefore, direct and indirect public funding of care services for older people, as well as for people with disabilities and children, has remained residual. Doñate et al. (CAP 2013) show in their case from Spain that the very advanced legislation on 'personal autonomy' from 2007 upgraded the country from a residual (family-or charity-centred) system to a quite encompassing caring state. In 2013, however, benefits for residential care, as well as care allowances for family carers, were simply discontinued. In Italy, which also followed a 'latecomer' trajectory (Da Roit and Sabatinelli, 2013;CAP Martinelli, 2012;CAP Bagnato et al., 2014), dramatic cuts in public support since 2010 have resulted in the disappearance, or at least severe reduction, of many services, especially in Southern regions and municipalities, forcing many users to fall back on the family or turn to the informal market (private hiring of immigrant caregivers). In Malta, unregulated informal marketisation of care for older people has also increased. Families who can afford it often do not provide care services themselves, preferring to employ informal home care workers (mostly Filipinas). This solution, however, is unaffordable for those who only receive old age pensions (Pace et al., 2016). Ultimately, even the informal market option has become more limited in these countries, for instance in Greece, due to huge unemployment and loss of income among all social groups (CAP Konstantatos, 2013). The use of cash benefits for the care of older people has stimulated informal market provision also in Germany, Austria and the Czech Republic. For example, the introduction of 'care allowance' has supported informal care given either by family members or informally hired, sometimes undocumented, migrant caregivers in the Czech Republic (CAP Kubalčíková and Havlíková, 2013). In Austria, recent legislation has introduced a special status for 'informal' care workers with very low wages and less protected working conditions compared to 'native' employees, in an attempt to introduce some degree of formal regulation (CAP Leibetseder, 2016b). In Germany, the growth of informal care work has not led to any legislative changes, but non-profit care providers have been demanding reforms for years (Neuhaus et al., 2009;Lutz, 2015).

Active and Passive Re-familialisation of Care
Historically, a significant share of care work has been progressively removed from the realm of the private household and entrusted to formal services, whether public or private. However, the nature of this change differs across social services and countries. Care model researchers have shown that, even today, some countries lean much more on families for care provision than others (León, 2015). The term 're-familialisation' describes a situation where households and families have to take up again a share of care responsibilities. Here too there can be passive and active re-familialisation.
An important difference between childcare and care for older people must be stressed here. Childcare service provision has grown steadily in most European countries, most particularly pre-school services, whereas there is much more variation in the scale, scope and targeting of care services for older people and other dependent adults. Re-familialisation thus takes very different forms depending on which services and countries or regions within countries are studied.
A clear trend of passive re-familialisation is evident, for instance, in Malta with regard to care for older people. Here, low-income people have had to turn to family members for care, while better-off people can purchase (informal immigrant) services with their own money (Pace et al., 2016). In the municipality of Reggio Calabria (CAP Martinelli et al., 2014;CAP Bagnato et al., 2014), the limited provision of public services and, recently, the further reduction of public support for both childcare and care for older people have prompted many families who cannot afford private care to resume care responsibilities. Deusdad and Zafra (CAP, 2013) show how the heavy cuts in public funding of both care services and housing services for older people in Tarragona (Catalonia) have reversed previous service gains and have triggered passive re-familialisation, whereby deprivation and inequality between classes and genders have resumed.
Passive re-familialisation can also be observed in the field of care for older people in the Nordic countries (Rostgaard and Szebehely, 2012). Local case studies in Finland indicate that stricter targeting of residential and home care services has accompanied reduced government expenses. Families have had to take on more responsibility for care tasks than before and this has also become an official objective that is enshrined in many government documents (CAP Anttonen and Karsio, 2013). The newly revised Social Services Act of 2016 supports this development, turning passive re-familialisation into an active endeavour. The new law underlines the participation not only of users but also of their family members, and makes the role of family members much more explicit than before. Conversely, in the Danish case, according to the law it is not families who must take over care responsibilities but the clients themselves, who are 'activated' (enablement policy). This also strengthens the clients' decisionmaking discretion, fostering individualisation (CAP Jensen and Fersch, 2013).
A relevant exception is childcare services in the Nordic countries, the Netherlands, France, Austria and Germany. In this field, continued de-familialisation is observed, in the form of increased subsidised or publicly financed and sometimes even publicly provided kindergartens (Ferragina and Seeleib-Kaiser, 2015), which continued after the 2008 crisis. Some countries have also extended mandatory parental leave schemes. A move towards increased kindergarten coverage and longer periods of paid parental leave can be detected in other European countries as well, although not as strongly (Ferragina and Seeleib-Kaiser, 2015). Indeed, childcare and care for older people have historically had different statuses, the former traditionally enjoying stronger political and legal support. Nevertheless, there is some evidence that, to save money, Finnish municipalities prefer to grant cash benefits to parents of small children than to provide publicly funded childcare services, a policy that has widened the gap between higher-and lower-educated mothers, the latter preferring cash benefits (Mahon et al., 2012). Such a development would indicate a process of stratified, active re-familialisation.

The Communitarisation of Care
'Communitarisation' implies that the role of third sector organisationswhether large Non-governmental organisations (NGOs) or smaller Community-based organisations (CBOs), whether based on voluntary work or paid professional work -is increasing in the overall production of care services. If the state withdraws public funds for social services, these organisations may move in 'by default' and fill the gap. After the 2008 economic crisis, in many places the third sector has become more active in the provision of social services. Compared to previous phases, this features a form of passive re-communitarisation. Small, ad hoc neighbourhood-based, but often networked, CBOs are filling the gap left by the dramatic reduction of the public care sector in many contexts as a result of austerity measures. In Greece, for example, there has been an increase in local self-organising, bottom-up networking initiatives, addressing health, housing and neighbourhood support (CAP Adam and Papatheodorou, 2014;Häikiö et al., in this volume). Semprebon and Vicari (CAP 2014) examine an initiative to empower immigrants to build their own houses. Vaiou and Siatitsa (CAP 2013) reported on community-based organisations in Athens that provide services such as soup kitchens, food banks, communal cooking, medical wards, as well as counselling and support centres. In Catalonia, Escobedo and Escapa (CAP, 2014) describe the role played by parents' organisations in the management of out-of-school care in times of crisis, albeit supported by government resources.
These community initiatives sometimes take the form of political mobilisation (CAP Vaiou and Siatitsa, 2013). Persons risking eviction or court proceedings are helped by community initiatives offering legal support, as well as initiatives to organise public demonstrations and civil disobedience. García and De Weerdt (CAP 2013) analyse how such a community initiative in Spain mobilised against paying mortgages in a context of rising unemployment and the burst real estate bubble. Jolanki (CAP 2014;Jolanki and Vilkko, 2015) describe a commune of older adults that was established in order to respond to the problems of access to public care services and to help develop their own design for life within a shared residence in Finland.
There are also instances of 'active' communitarisation, when local governments actively engage with third sector organisations, albeit often with insufficient funding. In the UK, for example, the financial crisis has generated severe cuts in national funding, putting further pressure on local government budgets. As illustrated in the case of services for older people in Leeds (Yeandle, 2014), local governments are therefore increasingly involv- ing community-based -often voluntary -organisations in the delivery of social services, sometimes by awarding them contracts for publicly funded services, sometimes by encouraging grassroots and voluntary organisations to develop additional support, following the ideal of a conservative Big Society that functions without state resources. A similar active communitarisation trend is observed by Fraisse (CAP 2014) in his analysis of childcare services in the French cities of Lille and Nantes. There, left-wing local governments have implemented a centrally-directed expansion of childcare by involving mainly third sector and user associations rather than for-profit providers. The increased role of local organisations is thus 'publicly-led', with local authorities supporting the development of new services, with strong support from the central state as well. In contrast to the UK, however, the French welfare re-mix remains embedded in a national regulatory and funding framework.

THE SOCIAL IMPACTS OF HORIZONTAL RE-MIXES
The concept of 'horizontal re-mix' has been used in this chapter to describe changes that are re-shuffling welfare mixes in multiple and contradictory ways. Our main conclusion is that impacts of the current horizontal re-mix diverge substantially from those postulated by the welfare mix theory of the 1980s and 1990s. The idea of partnership and close collaboration among the sectors has become eroded as a result of increasing market dominance in many countries. If one actor gains a (near) monopoly in the mixed economy of care and influences all other actors' performance, one would hardly expect to find equal partnerships and mutual ties among service providers. Even increased competition among service providers might hinder such fruitful partnerships. It appears that market-, third sector-and family-based care provision modes have attained a stronger position than earlier; most particularly, countries that used to rely significantly on the direct public provision of services have taken steps to introduce welfare mixes. Following the early assumptions of welfare mix theory, diverse providers were thought to lead to greater variety in welfare provision and, through collaboration between equal players, tailored services were supposed to cater for individualised needs. In contrast, the more recent re-mixes feature multiple trends that do not support these claims. These trends, as shown in our taxonomy and examples, suggest problematic and sometimes even negative re-mix effects.
First, care inequality is increasing. On the one hand, higher income groups can access better and more individualised service provision, as they can afford to pay out of their own pockets or combine individual payments with cash benefits in order to secure higher quality services. On the other hand, stricter targeting and means testing exacerbates social exclusion and stratification, as publicly supported services just provide a bare minimum for the poor, and low-income populations above the minimum income threshold may be too poor to pay for care services. The latter groups must increasingly rely on family networks (passive familialisation) or local CBOs (passive communitarisation). Thus, while the richer users have been transformed into consumers enjoying ample choices, in the case of the poorer users, services are provided unevenly, across social groups, service fields and places, creating new inequalities that threaten universalism and social citizenship. In other words, welfare re-mixes tend to benefit better-off citizens, which was not what welfare mix supporters aimed at, when the argument was first brought into welfare theory. Secondly, government involvement in the governance of social services is diminishing in many contexts, although with different intensities across places and services. The retraction of public funding largely results in a greater role for profit providers and self-employed care workers. But governmental withdrawal also opens up ungoverned spaces that lower the potential for productive collaboration between diverse providers or the active state-led coordination of these. This leads us to ask: should the role of the state be stronger than is suggested in the welfare mix literature, which placed so much faith in equal partnership? Less state involvement easily leads to the dominance of market providers.
Finally, the growing complexity of care services is becoming a major problem. Not only does it result in splits among diverse providers in the same area; sometimes clients have to engage with multiple providers to cover one specific care need. Furthermore, the fragmentation and disruption of services challenge accountability and assessment, as matters of responsibility and factual service provision are rather arduous to address.
Our main conclusion is, thus, that the concept of 'horizontal welfare re-mix' is needed to assess the overall transition from state care service provision towards a mixed provision of such services, as well as towards a more market-based provision of care. The retreat of state provision and public funding in social care tends to result in a lower integration of, and greater fragmentation among, service providers. The original idea of welfare mixes -idealised as bringing together the best of the service sector actors, including families and households, while lowering expenditurehas certainly not been fulfilled. Instead, a social care landscape has emerged that operates less harmoniously than before.