Stability and change in the field of residential care for children. On ownership structure, treatment ideas and institutional logics

ABSTRACT The field of residential care for children and youth in Sweden is often termed unstable and turbulent. During recent decades the field has been subject to many changes. In this study, the development and changes in the field of residential care for children and youth in terms of ownership structure and treatment ideas will be analysed. The study is particularly focused on the changes in ownership structure that have taken place during the 2010s. It also analyses changes in treatment ideas, and discusses how these may relate to transformations of ownership structures as well as to dimensions of institutional logics, such as legislation and other types of normative pressure from the environment. The result reveals that of the approximately 450 treatment oriented residential care units (excluding homes for refugee children), close to 80 % are today run by private companies and to a growing extent by large for-profit corporations. Parallel – and possibly related – to the changes in ownership structure, the dominant treatment ideas have changed over time. The changes in the field can be summarised as a transformation from small-scale establishments with a family logic, to large-scale establishments with a professional logic, or more specifically from a domination of small family run units with milieu therapy to big business and a focus on evidence based interventions.


Introduction
The field of residential care for children and youth in Sweden is often referred to as unstable and turbulent. This sector has indeed been subjected to numerous changes both in a long term and in a short-term perspective. In line with New Public Management ideals (NPM), there has been a high influx of for-profit providers and the composition of the field in terms of ownership structures has thus undergone substantial change (Meagher et al. 2016). Parallel to the reshaping of the field, changes with regard to treatment ideas and methods linked to social work in general and residential care for children in particular, have taken place.
In Sweden there are currently around 2000 residential care units (RCU, in Swedish HVB, Hem för Vård eller Boende) of which the great majority target asylum-seeking children. 1 However, in this article our interest concerns the 'treatment oriented RCU's', meaning the RCU's that target children who are placed for reasons such as maltreatment in the home environment or behavioural problems. Whereas the RCU's that target asylum-seeking children function more like housing for minors without extensive treatment needs, the treatment oriented RCU's have an explicit treatment focus. The treatment oriented RCU's are currently around 450 in number.
Among them are those that, despite a changing environment, have been in operation for several decades, but also a great number of newcomers.
The aim of this study is to analyse development and changes in the field of treatment oriented residential care for children and youth in terms of ownership structure, as well as to present a picture of the tendencies with regard to treatment ideas on the field. The study targets transformations in ownership structure with special focus on the establishment and position of 'care corporations'. Through an analysis of the RCU's presentations of themselves today and in light of previous research regarding the field during the 80s and 90s, the study discusses changes in treatment ideas, and explore possible relations between the two parallel changes (ownership structure and treatment ideas) and their relations to dimensions of institutional logics, legislation and other types of normative pressure from the environment. Finally, the implications for social work practice are discussed.

Research background
To date, there is a small body of existing research on the privatisation of child welfare and RCU's in Sweden. This research shows a substantial increase in private provision of child welfare services since the 1990s (Trydegård 2001;Wiklund 2011;Lundström, Sallnäs, and Wiklund 2018). At the beginning of the current decade, a number of researchers presented data on privatisation in different welfare sectors, including education, health care and elderly care. Compared to other sectors, child welfare stood out as having the largest proportion of personnel employed by private providers (Hartman 2011).
The international research on this topic is scarce. However, a strong expansion and a large proportion of private care providers has been identified in for example Finland (Toikko 2017), Norway (Nordstoga and Støkken 2009;Ekhaugen and Rasmussen 2016) and the UK (Kirkpatrick, Kitchener, and Whipp 2001;Narey 2016). The rise to dominance of the for-profit private sector in residential care represents a fundamental shift in how services for vulnerable children are provided in Sweden and elsewhere.
There are a number of studies in the field of child welfare that focus predominantly on the impact of privatisation and NPM within public child welfare agencies, exploring the professional experiences of social workers (Dellgran and Höjer 2005;Liljegren, Dellgran, and Höjer 2008) and supervisors (Shanks 2016), as well as on how privatisation has affected the labour market for social workers (Shanks et al. 2018). Other studies have analysed social workers' views on purchaser/ provider splits (Wiklund 2005) and municipal procurement strategies (Höjer and Forkby 2011;Forkby and Höjer 2018) in the wake of NPM. Overall, these studies identify several unintended consequences when agencies within child welfare adopt business-like arrangements. In relation to residential care, high transaction costs and difficulties in the procurement process are examples of complications that have been highlighted. Also very high levels of company-profits have been identified and called into question .
In a recent article, Meagher et al. (2016) analysed how the market in the field of residential care for children has emerged and evolved. Institutional factors, changes in regulation and the economic logic of the market were discussed. While the focus in the study by Meagher et al. was on changes of ownership structure in the field, the emphasis in this article is rather on the relation between treatment ideas, institutional logics and ownership structure. In relation to Meagher et al., we also deepen the analysis of the entrance and growth of large companies and care corporations in the field and how their presence has affected the composition of categories of RCU's.

Institutional logics, a theoretical framework
To aid the understanding of the changes on the field of residential care, we draw on concepts from institutional analysis. Social institutions (market, profession, religion, etc.) are connected to specific logics that frame and shape social relations and behaviour within them. Institutional changes affect the mix of logics that organise a field (Thornton, Ocasio, and Lounsbury 2012). To different degrees, organisations such as RCU's are products of their time, and changes in the organisational field can partly be understood as an effect of the expiry of out of date organisations parallel to the rise of new types (with better adaption to the environment). As will be shown, it is also evident that 'history matters' and that remains of previously dominating logics can be found in the presentations of many RCU's.
The field of residential care contains a mix of newcomers and old RCU's, the latter with roots in logics that differ from those of the homes that have more recently entered the field (Hannan and Caroll 1992;Sallnäs 2000). Actors can to varying degrees allude to different institutional logics, related to for example profession, religion, state, market and social movement (see Thornton, Ocasio, and Lounsbury 2012) when starting new RCU's or renewing old ones. In line with this, when RCU's disappear from the field, it may be because they build on logics with a weakened position in relation to residential care for children. As Thornton, Ocasio and Lounsbury (2012, 171) states: 'an array of institutional logics may be available to actors, but only certain institutional logics or categorical elements of logics are readily accessible to actors and therefore likely to be used'.
To illuminate different aspects of the changes in the field of residential care, caused by the influence of different institutional logics, we will in the following sections use Scott's (2008) concepts regulative, normative and cultural/cognitive. The regulative and normative dimensions have to do with legislation, formal control and different forms of normative pressure, whereas cultural/cognitive as we use it concerns dominating institutionally embedded ideas surrounding the field. Cultural/cognitive dimensions capture surrounding ideas within the public sector in general (for example NPM), as well as those closer to the field and the profession within it (for example treatment as a solution to social problems; Evidence based practice). Theoretically, the cultural-cognitive dimension relates to taken-for-granted ideas, whereas the normative and regulative dimensions concern conformation not only to moral norms but also to legislation. Empirically however, it is not easy to separate these different dimensions (Thornton, Ocasio, and Lounsbury 2012; see also Scott 2008).
Changes in the field of residential care during the 80s and 90s Since the 80s, the field of residential care has seen major changes in the form of privatisation. From being almost exclusively dominated by the public sector in the first years of the 1980s, the later part of the 80s represents a turning point for the private-/public sector mix. By the mid-90s, for profit companies dominated the field with around two thirds of the market share (Sallnäs 2000). This change in owner structure can be understood as the result of an institutional change, which in turn altered the dominance of logics. In the 70s and 80s, the public sector was subjected to massive criticism for being ineffective and unaffordable. The recipe for success was implementation of market-oriented strategies and reforms, building on principles like competition and economic efficiencya market logic (Hood 1991;Montin 2007;Meagher et al. 2016;Shanks 2016). The increased influence of the market logic was connected to several regulative changes that contributed to the above mentioned reshaping the field of residential care. The new social services act, passed in 1982, constituted a deregulation of the residential care field that enabled for-profit actors to enter (Meagher et al. 2016), and in 1994, the public procurement act came into force, which meant that the public sector had to comply with EU competition policy (Montin 2007). Additionally, a decentralisation of financial responsibility from central state and regional level to the municipalities took place during the 90s, meaning that the municipality took over financial responsibility for residential care (Meagher et al. 2016).
When the new Social Service Act (from 1982) came into power, it not only enabled for-profit actors to operate RCU's, it also widened the definition of residential care. This resulted in a variety of care facilities with very different modes of operation being defined as RCU's. For example, the act entailed that foster homes with many placed children should by law be defined as RCU's. Due to this, a number of foster homes were redefined as privately run RCU's. In parallel, new smallscale and privately run RCU's described as 'family like' were established. This blend of residential care and foster care (but formally defined as RCU's and parts of the RCU market), had its hay day during 1980s and the beginning of 1990s (Sallnäs 2000). The basic organisation of the RCU's and the day-to-day life in them were designed to emulate the family in important aspects. These RCU's often lacked a specific theoretical model guiding their work and had a strikingly low level of formal education among the staff (even if education generally is low in this sector). This state of affairs demonstrates that 'family' was a key institutional logic for treatment at that time. The family logic has today largely given way to other logics that are perceived to better fit the current structure of the field of residential care.
Professional logic is an important concept for understanding the structure of the field and the treatment models of today. It implies the obligation and discretion to practice social work in accordance to what is sometimes labelled 'science and proven experience', as well as an ethical commitment to the client's best interest, i.e. ensuring that the client is given the best possible care, and that the costor profitof such care is of secondary importance (Arman, Liff, and Wikström 2014). Since the 1990s, the evidence based practice movement has influenced, and to some extent changed, the meaning of professional logic in Swedish social work . In its widest meaning, evidence based practice (EBP) is about including evidence from research into professional practice (i.e. making a move from proven experience towards science). In practice, EBP has meant an increased focused on more or less distinct standardized interventions. From a professional perspective, the use of standardized interventions is somewhat of a double-edged sword; on one hand, they can threaten professional autonomy and on the other, they can increase both professional status and accountability at least if they are grounded in some sort of scientific package (Björk 2016;. From the perspective of a market logic, the use of evidence based interventions, or interventions presented as evidence based, can serve as effective marketing, if the purchasers (often professional social workers) perceive them as trustworthy.

Method
Data about providers of RCU's and presentations of treatment ideas have been collected from different sources. The description of the field of residential care in the 1980-1990s, which plays an important role as a backdrop for the understanding of the later development, is based on previous studies. For the analysis of the recent developments and current situation we have used empirical data collected from a) the national register of RCU's and b) websites of privately run RCU's. In our analysis, we have excluded the secure care units governed by the central state (n = 24-25). These types of units constitutes a special kind of RCU's, which are empowered to use far-reaching disciplinary measures against the residents. By law, these units cannot be privatised.
Since the beginning of the 2010s, national reporting of RCU's is operated by the Health and Social Care Inspectorate (IVO). We have used data from their register for the years of 2014 and 2017. The register covers all units with permission to run RCU's and includes information about owner, number of beds and target groups. It also includes a few paragraphs in which the RCU's in their own words can describe their treatment approach, content of care, etc. The register consists of a basic presentation of the RCU's that may be used by social workers and others, but it also includes details regarding permissions to run RCU's, as well as IVOs inspection reports.
Since ownership structure is not apparent in the register, we have manually searched for this information on the websites of privately run RCU's and paid special notice to those owned by care corporations. In doing so, we have been able to group all RCU's that were active during 2014 and 2017 into five categories; large private companies, small private companies, municipal companies, municipal organisations and non-profit organisations (see also Meagher et al. 2016).
Companies operating three or more units have been categorised as 'large companies' and those with fewer units as 'small companies'. The rational for this cut-off is that it is reasonable to assume that smallsometimes family ownedcompanies generally start with one RCU. Some of these expand into two separate units (e.g. with two different target groups), but may still be considered as small scale businesses. There is an overall pattern showing that having few units also means having relatively few beds, indicating that these companies are small in a double sense.
When the number of units amount to three or more, we have categorised the companies as 'large'. However, it should be noted that since our interest concerns actors in the field of treatment oriented residential care, we define the size (small/large) in accordance with the companies' number of homes within the residential care market only. There may be companies that have been defined as 'small' (due to them owning less than three treatment oriented RCU's), although they may be 'large' in other fields and also in terms of e.g. economic turnover. This means that if anything, our categorisation may risk an underestimation of the market position of large companies.
We use the term 'care corporations' to distinguish a subset of large companies. Care corporations often have a complicated organisational structure built on subsidiaries, and in addition to RCU's for children, they offer a range of other care services such as elderly care, disability services etc. Residential care for children is in other words one, but often not the main, service offering.
As for the analysis of treatment ideas, RCU's websites in combination with the register have been used to gather information regarding services provided, treatment ideas and target groups. The web sites of the care corporations have been of special importance, but also other web sites have been analysed. It should be underlined that our purpose is not to analyse in detail what actually happens in the RCU's, but rather to present an overview of the treatment ideas that the companies accentuate in their presentations of themselves. We have analysed this with special focus on changes over time, and how these changes relate to different logics (family, profession etc.) and to the different types of providers (small companies, large companies and care corporations).

Results
We will start this section by describing changes in the composition of the RCU field in terms of ownership structure, with specific focus on the diminishing proportion of small companies and the establishment of large ones. Thereafter we will discuss how the RCU's present their treatment offers and how this may relate to changing ownership structure and other factors.

Corporatisation of the residential care market
As mentioned above, for-profit companies dominated the field already in the 90s. In the middle of that decade, the for-profit companies accounted for two thirds of the market, while non-profit organisations and municipalities accounted for one in ten and one in four respectively. During the following decades, the market share of the for-profit companies has increased somewhat before stabilising at around 75 % during the last decade. Parallel to this, the non-profit organisations' market share has decreased, and amounts to about 4 % today (see Table 1).
RCU's operated by the public sector accounts for about 20 % of the field today, including about 6 % that are owned by the municipalities but that are run as private limited companies. As shown in Table 1, the position of large companies was considerably strengthened between 2014 and 2017, increasing from around one fourth (23.6 %) of the field, to nearly a third (31.6 %). Today, the 31.6 % (n = 142) RCU's are owned by in total fifteen large companies. Parallel to the growth of large companies, the small companies' market share has been reduced correspondingly. Looking at the share of beds, we can see that the distribution between the categories of RCU's is similar to the proportion of units, as is the development over time.
The increasing market share of large companies is the most apparent change in the 2010s, and although it may be viewed as marginal compared to the major privatisation of the residential care field that took place in the 80s and 90s, it has had apparent qualitative implications. One of these is that the growth to a large extent is concentrated to care corporations, i.e. large companies that own subsidiaries and offer a range of care services. The presence of these corporations in the field has led to a development that may be termed corporatisation or owner concentration. By 2017, a total of five companies that may be characterised as care corporations can be found on the market of residential care (Humana, Attendo, Frösunda, Ambea/Nytida and Team Olivia). Together these five corporations control over 20 % of the treatment oriented residential care market for children, although there is a large variation in how many RCU's that each corporation ownranging from four to fifty four. All of the care corporations are owned by private equity firms or investment companies, and three are listed on the stock market. Also, four of the corporations are multinational, which in this case means activity in several Nordic countries. The care corporations started their business within adjacent fields, such as eldercare or personal assistance. In general, the care corporations expand through acquisition, and it is in this way that they have come to enter the field of residential care for children (see also Scourfield 2007). The first care corporation entered the field of residential care in 2006 by acquiring a company owning several RCU's targeting e.g. children, families and refugee minors. It was not until 2011 that (three) other care corporations followed this example. The fifth and latest care corporation to enter the field acquired a number of residential care companies in 2015.
It is difficult to establish exactly why the care corporations entered the field of residential care at the time that they did. One hypothesis is that the RCU market was discovered by the care corporations through their establishment in neighbouring areas such as elderly care. Another is that regulations on procurement, implemented at the beginning of the 90s, in the long run made it harder for small actors both to get into the field and above all to stay there. What appears clear however, is that care corporations continue to grow in this field. The corporation (Humana) that today owns the greatest share of RCU's has grown from 30 treatment oriented RCU's in 2014, to over 50 in 2017, meaning that this corporation today owns more than one in ten of the RCU's on the market. Also the other care corporations have increased their share of the residential care market in recent years.
Although the establishment of the care corporations has made a significant impression on the residential care market, there are other types of large companies present in the field. These (ten) companies own fewer units each, and also differ from the care corporations by offering a smaller range of services. Some of these organisations focus broadly on personal social servicesaimed at both children and adults and involving services such as RCU's, special education units and outpatient treatment. Others are mainly focused on out-of-home care, offering services like RCU's and enhanced foster care administered through private companies. These companies control a much smaller share of the market (10 % in total) compared to the care corporations.

New treatment ideasfrom a family logic and milieu therapy to a professional logic and evidence based interventions
Parallel to the growth of large companies there has been extensive changes with regard to treatment ideas in the field of residential care, i.e. changes in the cultural/cognitive dimension. As indicated above, the family logic had a strong position in the field in the 90s, but has lately to a great extent been replaced by a focus on so called evidence based methods. Despite the fact that family logic had a great hold of the residential care in the 90s, RCU's were (and are) in principal presented as centres for treatment. This aspect may be viewed as historically rooted, but the professional logic has become increasingly important over time. In the 90s, theories and models governing the work were often unspecified or non-existent and descriptions of target groups were made in very general terms. At the end of the 1990s only about every second head of a RCU could specify a 'known' theory or model (i.e. psychodynamic approach, cognitive therapy etc.) that was guiding their work with placed children (even fewer among the small scale units) (Sallnäs 2000). As for the RCU's that could identify an actual model or theory guiding their work, psychodynamics and milieu therapy (using the environment and day-to-day interaction as treatment agents with inspiration from for example A.S. Neill's Summer Hill) appear to have been the models of preference. Thus, together with the logic of the family, these models stood out as the cornerstones of residential care during the 1980s and 1990s (Sallnäs 2000(Sallnäs , 2009Kvaran and Holm 2012). At this point in time treatment based on cognitive behavioural models had a marginal or non-existing role in RCU's presentation of their work.
In the new millennium, and parallel to the corporatization of the residential care market, an increased optimism about what could be achieved in a residential setting emerged. This development, related to the evidence based movement and connected to a professional logic, brought the conviction that new and scientifically proven interventions could lead to more effective care (Sallnäs 2009). Parallel to this, and probably related, the faith in family as a sufficient principal for RCU's weakened. An additional factor for the increasing influence of the allegedly scientifically proven methods was probably that these models for care were perceived as more competitive and better adjusted to large companies than was the model of family. Today therefore, many of the small-scale family related RCU's appear to have given way to other types of establishments in which the allusion to family logic is far less pronounced. One simple but revealing sign of this is that RCU's with names such as 'foster home [surname]' or just the first and surnames of the family were quite common in the 90s, while there are very few establishments with such names today (National Board of Health and Welfare 1997; Health and Social Care Inspectorate 2014, 2017).
The focus on evidence based, more or less standardized, interventions as a part of a market strategy appealing to a professional logic, is today highly noticeable in many RCU's' descriptions of themselves. This is especially evident for the care corporations, but also among other large companies and municipal companies. On webpages, statements such as 'We always work with established and evidence based methods for example MI' (Ambea/Nytida 20170407) are common. The care corporations also offer a wide range of what appearsat least on the surfaceto be welldemarcated programs/models. For example, several of the care corporations name programs such as: Acceptance and Commitment Therapy (ACT), Aggression Replacement Training (ART), Community Reinforcement Approach (CRA), Dialectical Behavioural Therapy (DBT), Cognitive Behavioural Therapy (CBT), and Motivational Interviewing (MI). Many of the available programs are connected to cognitive behavioural theories, although some psychodynamic inspired ways of working are also mentioned on the homepages.
By owning a large number of units, care corporations are able to offer a diverse range of treatment alternatives. It is however difficult to distinguish a coherent rationale behind the alternatives, except perhaps that they are perceived as marketable and allude to a professional logic. Instead, it is likely that the wide range of alternatives partly is a result of acquisitions of RCU's that already had certain treatment methods in place, and partly a strategic product differentiation. As well as offering a diverse range of treatment models, the care corporations are able to target a correspondingly wide range of populations with different problem areas and diagnoses. Judging by many establishments' descriptions, an increased focus on differentiation based on diagnoses and types of problems appear to have re-emerged in the field of residential care (this idea was strong from the 1920s to the 1960s, see Lundström 1993). This differentiation can be related to the evidence based movement with its focus on specific interventions/programs for particular populations, and to a trend towards medicalisation of social problems (Lundström 2016) but it may also be understood as a way for the corporations to differentiate the 'products'.
To what extent the models offered by the RCU's actually are effective is most definitely up for debate, and the same goes for the models used in the 1980s and 1990s. In any case, the changes in treatment options is not the result of strategic planning by the state or the municipalities on basis of the needs of the children, but rather a result of handing these issues over to the market (see also Meagher et al. 2016). It is clear however, that the institutional logic of family as well as treatment ideas related to milieu therapy in the 1980s and 1990s have been out-competed by the so-called evidence based interventions. Several of the methods that the RCU's today present as part of their treatment are well known and some of them are also manual based. This is quite different from the 90s when many RCU's could not specify a known model/treatment. However, considering the fact that many manual based programs are received as one small part of 24 h care, the term only covers a minor proportion of everyday activities at a residential care unit (see Ahonen 2012). One should also remember that we have studied how RCU's present their treatment, and we know little about these presentations relate to the actual use of methods, let alone the day-to-day life in the RCU's. As Bergmark and Oscarsson concluded as early as 1994 regarding an adjacent field, alcohol and drug treatment in the social services sector, professionalization is partly imaginary 'and can largely be said to be 'de jure rather than de facto' (Bergmark and Oscarsson 1994, 45), i.e. professionalization is present in legislation, official reports and RCU's presentations but not in practice. To what degree this is valid today in residential care for children and youth is largely an empirical question. Judging from how the companies present themselves, the treatment ideals have changed from alluding to a family logic to alluding to a professional logic based on specific treatment methods. However, the educational level of the personnel appear to have changed very little. We know from statistics regarding the workforce in RCU's that half of the personnel lack a formal education that is relevant for the service, and one third lack post-secondary education altogether (IVO 2013). This is much similar to the state of the field in the 90s (Sallnäs 2000). In the light of this, it is certainly relevant to pose the question if the competence in the RCU's is sufficient for providing an environment that is permeated by the 'scientific' treatment models that the companies declare that they are using.

Summary and discussion
In this article, we have analysed changes over time in regard to ownership structure as well as prevailing treatment ideas in the field of residential care for children. We have paid interest to possible connections between the two parallel developments as well as their relation to the institutional logics that characterise the field of residential care. When identifying changes over time, we have pointed to the rise of large companies in general and care corporations in particular, as well as on the increased focus on specific treatment models in the residential care companies' presentations of themselves. We suggest that these changes are intertwined and that they can be summarised as a transformation from small family run firms and milieu therapy to big business and a focus on evidence-based interventions. In other words, the changes of the field can be described as going from small-scale establishments with a family logic, to large-scale establishments with a professional logic. This development must be understood against the background of changes in the regulatory dimension of the field; most important are deregulation and legislation on procurement. As for changes in the cultural/cognitive dimension, the increasing importance of treatment (rather than family care) and the growing significance of professional logicat least understood as methods with a base in researchare of importance.
The most noticeable change in terms of ownership structure, is the establishment of large companies and among them the arrival of the care corporations around 2010, a process that is parallel to a weakened position for the small companies. The overall picture is one of an increasing concentration of large companies and an expansion of a few care corporations. As shown, the corporations have a powerful position in a quantitative sense (currently they operate every fifth bed among treatment oriented RCUs in Sweden) and it is safe to say that the market logic they bring impact the field and the treatment options offered.
In terms of treatment ideas and methods presented by the residential care companies, it is obvious that today it is hardly competitive to allude to family as the main organising principle. The professional logic has instead become increasingly powerful, entailing that companies have to present specific treatment methodspreferably claimed as evidence based. Even though the evidence based movement is questioned both for an actual lack of evidence and on more principal grounds, it can still be regarded as a strategy for professionalization of social work . The evidence based movement with its foundation in ideas of scientific knowledge as guiding principles for interventions has a powerful position as a cultural/cognitive dimension in society as a whole, but nowadays also appears to be a strong in social work, with no exception for the field of residential care.
Evidence based practices are clearly mirrored in the way that many of the residential care companies describe themselves. The number of methods claimed to be evidence-based presented by the residential care companies is sizable and especially so among the large companies and the care corporations. The presentation of evidence based treatment models seem to have a central role in communication with the environment, with professionals in social services being the main recipients. Since the mid-1990s, the residential care companies' way of adjusting to the cultural and cognitive environment and to compete in the market is primarily by alluding to the professional logic and the evidence based movement linked to that.
As for regulative dimensions, legislative changes and deregulation opened up the field for private providers in the 1980s and 1990s. Ideals of decentralisation and the rise of NPM were important phenomena surrounding residential care that acted as driving forces behind the process of marketization. Parallel to the increased marketization and the growth of for profit providers, demands for a stronger regulatory power for the state were raised. Criticism against high levels of profit, uncertainty about the quality of care, and even so called 'care scandals', have probably reinforced this process (Sallnäs and Wiklund 2011).
The most visible expression of these regulatory demands in the 2000s is the establishment of a national state inspectorate (IVO), which was intended to facilitate increased monitoring (Lindqvist 2014;Meagher et al. 2016). The sharper audit, together with strengthened prerequisites for starting RCU's, can be seen as a regulative initiative to gain control over a field that had become very turbulent. However, the audit system has limited power to exclude RCU's or to (re) organise the field in a more fundamental way and IVOs duties do not include regulation of the market per se (Pålsson 2018).
Theoretically, one might imagine regulations that take as a starting point the different problems of the children and thus contribute to a need related structure when it comes to treatment methods, specialisation with regard to client group, etc. However, such considerations are today to a large extent handed over to the market. In this sense the regulative power has been rather weak, and it can be discussed to what degree the idea of balancing the logic of the market by setting up an enforced control apparatus have been realised. Pålsson concludes that: 'it is difficult to make certain core aspects of care auditable (e.g. children-staff relationships, care aspects that children attach importance to, and research knowledge). This implies that it is questionable whether the audits actually improve residential care quality' (Pålsson 2018, 54).
One of the motives for letting for-profit providers enter the field of residential care was to encourage innovation in terms of treatment settings and to create variation on the supply side. The growing concentration and the rise of the care corporations has largely competed out small scale RCUs (more or less based on the logic of family), so in that sense variation on the supply side has decreased. On the other hand the services that the large companies and especially the corporations offer on the market are broad and diversified in terms of presented treatment models. They allude to a professional logic high on the agenda in the 2010s and they present the variation in treatment models as such, as an important asset. However, as indicated above, the changes do not seem to reflect an increase in formal education for the staff compared the education level of staff in earlier periods. For social workers, the increased influence of market logicsincluding changes with regard to treatment options, high profits, and difficulties to overview the field of RCU'srepresent a professional challenge. This challenge includes choosing a residential care unit together with the client, as well as taking part in procurement processes. In the presentations that the RCU's provide on websites etc., there is little information regarding important aspects such as number of personnel, the personnel's educational level and training in treatment models presented etc. Furthermore, it is an open question to what degree the relatively broad palette of treatment models presented by the RCU's correspond to the actual needs of the children. From a practice perspective, one way forward could be establishing a national registerpreferably managed by IVOthat makes it possible to compare different residential care settings based on adequate quality characteristics. Note 1. During the last years, the field of residential care for children has expanded dramatically, a process that almost exclusively can be related to the high number of unaccompanied asylum-seeking children coming to Sweden.