A pendulum swing in child welfare policy – the case of implementing GIRFEC in Sweden

ABSTRACT

Sweden is among those countries traditionally ranked highly in international comparisons of children’s well-being and conditions for development. However, in recent years a development towards greater inequality in health has occurred. The general welfare model’s capacity to safeguard both the universal provisions for the general population’s standard of living and targeted support for those in need has also been doubted. System-related deficiencies such as collaborative breakdown, inadequate effectiveness and lack of clarity concerning how to uphold the best interests of the child are cited as examples, and several calls for policy reformations have been raised. The Scottish model for supporting child well-being, Getting It Right for Every Child (GIRFEC) has garnered keen interest in Sweden and is an example of how ideas for policy reformation flow both between and within countries, and thereby undergo more or less radical transformations. This article analyses the first-phase implementation of GIRFEC in a Swedish county. It emerges that although there is a great deal of enthusiasm for the original model, the intention is to implement an adapted version. What similarities and differences would be realized is not clarified from the start, but is left for the implementation process. The positive reception is understood as arising from a perceived familiarity of the model, based on current practice and discourse. GIRFEC can therefore be regarded as part of a pendulum swing in which ideas are borrowed and lent between countries and contexts.

KEYWORDS:

• Implementation
• child welfare
• collaboration
• GIRFEC

Introduction

Strengthening the well-being and positive development of children and young people is a global concern and one of the 17 target areas in the UN’s sustainable development strategy. Such initiatives have been prioritized differently in various welfare systems, and the concrete policy initiatives within them are shifting (Ben-Arieh and George 2006). The Scandinavian systems have traditionally been distinguished by proactive reforms and general provisions to create favourable conditions for child and youth in a preventative spirit. These differ from conservative models in which interventions have more often been provided by civil society and where family self-determination has been prioritized under principles of subsidiarity, as well as from the Anglo-Saxon systems in which welfare institutions have been comparatively more reactively designed and aimed at protective measures (Esping-Andersen 1990; Hetherington 2006).

However, boundaries between different systems have been blurred, apart from being affected by economic harmonization and globalization because of policy initiatives and exchanges. In the area of children’s well-being, international institutions, such as the UN Convention on the Rights of the Child and EU white papers, are creating pressure towards alignment of national policies. In addition, various policy models and methods are spread through a variety of channels between countries and contexts. This process of dissemination through the media, political networks, exchanges between professional groups and so forth also involves translation processes through which the original idea is to be understood, transformed, and realized in a new context. In this way, reforms during the implementation process are subjected to repeated change or ‘iterative refraction’, when they are ‘filtered’ through the various system levels that may be involved (Weinbaum and Supovitz 2010). The original policy ‘mutates’ in an unpredictable way when new decisions are taken at every level during the process, based on various contextual deliberations, personal values and interpretations. The new policy may, for example, compete with local traditions and procedures or be incompatible with local policy (cf. Priestley 2011).

As a result, the process is open to political manipulation biased in favour of those who generate information, control access to policy venues, and synchronize or exploit group, national and institutional time-tables. (Ackrill et al. 2013, 872)

The ambition behind the Scottish model GIRFEC (Getting it right for every child) is to make Scotland the best place for children to grow up in (an ambition also declared in Swedish national policy). It constitutes an example of how policy ideas navigate through international policy currents, as it is also the strongest candidate in Sweden among the models aimed at improving child well-being. Both the Swedish Association of Local Authorities and Regions (SALAR 2018) and state bodies have come out in favour of GIRFEC, and implementation work has begun at local and regional levels in several places in the country (Ministry of Health and Social Affairs 2017). This interest has emerged even though Sweden is usually ranked highly for child well-being, while the UK has fared less well (OECD 2009; UNICEF 2013). Changes have been noted in recent years, however, where the trend in Sweden is clearly negative in many respects concerning potential inequality in child well-being and leading to a fragmentation of society in which disadvantaged groups fall further behind the rest (UNICEF 2016).

This article analyses how the GIRFEC model is understood and welcomed in a Swedish context, the ‘Children’s Best! in Kronoberg County’ initiative (CBiK). The perceived novelty of the model is studied: What is expected to be added to the current work in support of children, and how this is supposed to be done? The general knowledge contribution thus relates to the ways policy ideas migrate between different national contexts, with focus on the local translation that takes place in the implementation process.

1. What perceived problems in the local context is the model intended to solve?

2. What is expected to be the main contribution, and what are the considerations for implementation?

3. How can the interest in GIRFEC be explained in relation to current practice, and how does this relationship affect its implementation?

GIRFEC – a holistic model for child well-being

GIRFEC was introduced as an ambitious reform aiming at a holistic, integrated policy framework, bringing together practices and knowledge areas traditionally held apart and working in isolation. The model is based on the UN Convention on the Rights of the Child and builds on Scandinavian ideas related to child welfare and social rights (Stradling et al. 2009). It was adopted in 2009 as a national approach to improving the well-being of all Scottish children and young people. The model has been assessed as evidence-based and the theoretical basis as robust (Coles et al. 2016). Successes in several areas in Scotland have been associated with the model, such as halving the percentage of children in care (‘looked after children’), reduction of school exclusion and significant decreases in youth offending (Murray et al. 2015; Peterson 2016; Scottish Government 2015).

GIRFEC is more than just a policy framework; it represents a distinct way of thinking and a transformational shift in the form of an agenda for change, and further, may represent a new and emerging child welfare model. (Coles et al. 2016, 355)

Early intervention, giving the right support at the right time and achieving effective coordination are principles of the way social support should be organized. This is expressed by the appointment of a named person for every child/young person. The named person monitors the child’s development and acts as a first point of contact when there is a need for support. In more complex cases in which several support services/agencies are involved at the same time, a lead professional must be appointed, who then assumes responsibility for coordinating and monitoring the child’s progress. In these cases, support is coordinated using a joint child’s plan that documents strengths and difficulties in relation to the child, identifies relevant support actions and specifies which services/agencies are responsible for what. The framework includes that provisions should be guided by established risk and protective factors summarized in eight key areas in the Wellbeing Wheel.

In spite of legislative support and clear directives from leading public agencies, tensions have arisen during the implementation of GIRFEC in Scotland. A deeper understanding of what child well-being entails in relation to child welfare and to more fully incorporating child protection services in the common effort has been hard to achieve. The named-person function has also encountered resistance, as it has been perceived as net widening and intrusive (Coles et al. 2016), and is now offered on a voluntary basis.

Policy formation and translations

Kingdon (2011) has developed a concept of policy formation that attempts to explain why certain solutions to social problems garner attention at any given time, how this happens and which actors participate in the process. The theory covers three streams: 1) the problem stream, where current problems are cause for concern and are articulated by government agencies, media, decision-makers or special interest groups; 2) the policy stream, where policy ideas seen as potential answers to these problems are launched; and 3) the politics stream, where decision-makers are influenced to implement new reforms in line with the proposed policy. When there is a confluence of the streams, a window of opportunity opens with the support of policy entrepreneurs, who connect the three streams. Policy entrepreneurs employ various strategies to bring problems and policy together in attractive policy solutions that are then ‘sold’ to receptive decision-makers. A policy window is, however, open for a limited time, and policy entrepreneurs and decision-makers thus often work under time pressure, which can make it difficult for them to address all problems. They may therefore be forced to accept poorer outcomes when the new policy is implemented.

Kingdon’s theory has been further developed in areas including education research, with the concept of ‘borrowing and lending’ (Steiner-Khamsi 2012). It describes how, due to increased globalization, education reform, policy or research-based methods often gain attention and migrate between countries or organizations where they are ‘borrowed’ and ‘lent’ (see e.g. Czarniawska 2004). As with Kingdon, borrowing a policy from one context to another depends on a window of opportunity being opened, with greater receptiveness to innovative ideas for reform. One such opportunity arises, for example, when financial support for implementing new political reforms is suddenly made available (Steiner-Khamsi 2006).

Implementation and local adaptation

Findings in implementation research can contribute to better understanding of the difficulties that can surround the implementation of new models and methods in the area of child and youth welfare. Implementation studies of initiatives in the public schools and child and youth social services, for example, have repeatedly found that cultural and organizational conditions influence the degree of implementation, translation and use (Albers and Pattuwage 2017; Lundström and Shanks 2013; Willging et al. 2015). New methods and procedures often encounter resistance from managers and affected practitioners (Willging et al. 2015).

Durlak and DuPre (2008) argue in relation to evidence-based methods that these have been implemented without adequate adaptation to local conditions. Implementations of specific methods are complex processes surrounded with challenges that often lead to changes, backlash and unpredictable and varied outcomes (Albers and Pattuwage 2017; Wallin and Limén 2013; Willging et al. 2015). If the differences between the two versions become too great, it is difficult to determine whether the model in practice has retained its scientific support (Aarons and Palinkas 2007; Nutley, Walter, and Davies 2007; Stirman, Crits‐Christoph, and DeRubeis 2004). The question of what is the same and what is different between the current practice and the model to be implemented thus becomes central to determining what should be changed and how.

Therefore, it is essential to monitor the types of adaptions that occur instead of treating them as failures of implementation. (Durlak and DuPre 2008, 342).

The complexity of a specific method and the special resources it may demand affect its application. The simpler and more useful professionals and the organization perceive the method to be, the more it will be used (Greenhalgh et al. 2004; Rogers 2003). Contextual and organizational factors at the group and individual levels are also highly significant. Alignment with the prevailing culture and values as well as the knowledge, willingness and capacity of professionals and managers facilitate the implementation of new methods (Alexandersson 2006; Nutley, Walter, and Davies 2007).

Method

A mixed-method design was chosen to facilitate a rich understanding of the phenomenon by triangulation of data and validation of results (Creswell and Tashakkori 2007). The empirical material consists of a survey, interviews with key individuals in the implementation process, observations at meetings and workshops and analysis of articles in local media and administrative data. Administrative data refers to material including newsletters and documentation from steering group and reference group meetings. With these points of departure, the study is also in the nature of a case study (Yin 2009).

A survey study was carried out with the people who were involved in the relevant developmental work in the county during the initial stage of the process. The survey was distributed by email to a total of 130 people, of whom 76 responded, resulting in a response frequency of 58%. Four group interviews were conducted in parallel with a total of 15 key individuals from the participating local authorities. The respondents had various functions and belonged to different organizations, such as process management, social services, schools and the healthcare system. In addition, 15 articles in local media and 18 administrative documents were reviewed. Finally, a thematic analysis was performed to search for fundamental patterns, themes and categories in the combined empirical material (Bryman 2012).

The analysis began with close reading of the empirical materials with focus on how the method was received and translated and forming an overall picture of the process. Sub-themes and prominent key themes were identified, coded and categorized. Finally, the extracted themes and categories and a selection of illustrative quotations were compiled into a finished results text. For ethical reasons, certain names and designations of people, cities or places have been omitted from the results.

The local implementation – of what?

The analysis and results section show how GIRFEC is perceived by central actors involved in the implementation of the model in the Kronoberg Region. The group consists of decision-makers: politicians who have discussed the initiative in local media, managers at various levels in the cooperating organizations, local opinion leaders responsible for running the implementation and practitioners in the various agencies. After providing a picture of the respondents’ attitudes towards the initiative, the analysis is presented under three overall themes concerning 1) how the model is associated with perceived problems and solutions, 2) the transfer points and time frame for the implementation and 3) the adaptation and translation to meet local needs.

Motivation for the implementation

Specific questions about the CBiK initiative were included in the survey sent to key individuals involved in the developmental work. The aim here was to ascertain general attitudes towards CBiK among stakeholders affected by the initiative, and to ask what implementation support they needed. The result reveals a considerable and enthusiastic interest in the initiative. Virtually all (96%) agreed that there was a need to work according to CBiK, 73% said that the model was supported by their organizations and 89% said that the objectives of CBiK were clear. Further, around two-thirds (68%) believed that there was already a holistic view of children and young people in their organizations, and they reported that collaborative partners met regularly (84% agreed). Regarding difficulties, while a small group (23%) could see pitfalls in transferring the model to their own organizations, 56% believed that the main obstacles to collaboration were laws and regulations (including privacy laws) and 45% believed that the allocation of responsibility among organizations was not entirely clear. So, based on the survey overall, strong support for CBiK can be determined, but fundamental questions remain: what do they agree upon, and what is new about it?

When problems and solutions meet

An array of problems

Multiple problems related to children and young people that the new policy is intended to solve emerged in the texts and interviews analysed. Many of the problems addressed align well with the problems identified also at the national level (National Agency for Education & National Board of Health and Welfare 2018; Ministry of Health and Social Affairs 2017). As established in a document proposing to initiate the implementation, ‘The challenges in today’s society are many and serious’ (AD7). The text connects issues including worsening child and youth mental health, increasing disparities in school performance and ‘the health inequality and exclusion that come at a huge human and economic cost, to name but a few’. An opinion piece in the local press, written by a few leading politicians, talks about ‘exclusion that is passed down’ from unemployed parents to their children, to which the model is touted as a solution (LM1).

Several newspaper articles also mention problems linked to relevant organizations in the form of increased specialization in the welfare sector, which has contributed to each agency performing its services in isolation instead of working together on behalf of children and families. There is talk of a ‘silo’ organization structure and inadequate interagency cooperation as significant obstacles: ‘It is primarily the schools and social services that should abandon the inefficient and sealed silo model according to which they currently work’ (LM2). In another newspaper article, one head of municipal social services says: ‘The silo organization has to go. We do much that is good in our various professions, but we are too specialized.’ He argues that this type of organization also causes hardships for children and families who need support and are forced to ‘deal with multiple agencies, organizations and associations’ (LM8). In yet another, a couple of local politicians declare: ‘It is unacceptable to be sent from one agency to another or to have to chase them down’ (LM7). Confidentiality between agencies is also mentioned as a major obstacle (LM11).

Formulation of goals and solutions

Similar to the broad span of problems connected to the introduction, the array of goals the model is assumed to target also contains a wide spectrum reaching from universal to selective indicative prevention. An issue mentioned occasionally is reducing social marginalization by creating a safe and secure childhood for all children, or reducing the significance of socioeconomic background for school performance (universal prevention). For example, a politician responsible for the initiative clarifies in a newspaper interview that ‘[t]his is about creating a safe and secure childhood for all children. Not only disadvantaged children, but all children. This will be accomplished by promoting early intervention and coordination’ (LM11). It emerges in other articles that the primary goal is to protect at-risk children from exclusion and reduce youth offending (selective prevention). But there are also several texts that state goals connected to a reduced number of children placed in out-of-home care (indicative prevention). In a newspaper report on a municipal council meeting, it is predicted that the outcomes of the model would be ‘[l]ess social exclusion, reduced youth offending and fewer looked after children’ (LM10). In an interview with the three process managers, one relates how the project began and their strategic considerations:

The basic premise for this, when it started, was that they were placing children in residential care at great cost with poor outcomes, and how that could be managed. So, that was the gateway, and that is very often where the gateway is. You take something that is an urgent problem and at a very high intervention level. You don’t start with promotional and preventative efforts.

Improved inter-agency collaboration is recurrently expressed as a primary strategy needed to deal with the stated problems. Selling points such as ‘early and coordinated interventions’ and a ‘one-stop’ solution for children and families are used to explain the advantages. The model will provide incentives for services and agencies to combine their resources around the child and work in close collaboration. Minutes from a steering group meeting note that ‘[t]he social challenges we are facing cannot be overcome by an isolated actor or in prevailing structures and we thus need a shift of perspective that leads to improved collaboration and co-management’ (AD3). Several texts make it clear that the silo organization structure will be scrapped to improve effectiveness and achieve better outcomes (LM2, LM7, LM9): ‘Instead of working vertically, as in a drainpipe, we need to work laterally, as in a rain gutter,’ says the head of schools in one municipality (LM8). Some politicians declare in a newspaper opinion piece (LM3) that if collaboration among various actors can be established, ‘the successful Scottish model can be implemented and protect more children from inheriting exclusion from their parents’.

Watchwords like ‘effectiveness’ and ‘improved outcomes’ are used to signal the potential of the new model: ‘The collaborative model will both simplify and improve effectiveness. A more effective method will make it possible to identify at-risk families at an early stage and provide relevant support to prevent children and young people from suffering mental health issues later in life,’ says a head of social services in one of the municipalities (LM2). ‘It is not about more money; it is about working smarter,’ says a head of school health services in another (LM8). In a third, the model is considered as a solution to consequences of understaffing and large preschool classes: ‘a collaboration between social services and the schools to provide early and effective support to pupils with problems’ (LM6).

The model migrates and is borrowed

Several points of direct contact between representatives of Scotland and Swedish national and local organs have been important for awakening an interest for GIRFEC in Sweden. In 2016 the Government Offices of Sweden organized a seminar for local and regional actors in Sweden to take part in a presentation from Scottish representatives (Ministry of Health and Social Affairs 2017). Furthermore, several fact-finding trips to Scotland, mainly by groups of politicians and managers, have preceded their decisions to implement the model in various localities in Sweden (observation notes; see also Hansson 2018; SALAR 2017; Trumberg 2018; Municipality of Ystad 2018). This ‘bilateral trade excursion’ (Steiner-Khamsi 2012) has also occurred in the region covered by this study: ‘About 15 school and healthcare managers went to Scotland to learn more about a model that has been successful there’ (LM 2). This import of ideas has been repeated and organized in different ways: ‘We have been there and listened to them and they have been here and lectured. We will be visiting Scotland again soon, when we will meet them profession by profession’ (LM8). On one such occasion, a process manager described a presentation by Scottish representatives to several hundred practitioners in one city as a ‘hallelujah moment’.

A time-limited policy window is opened

Depending on how much is intended to be changed and how complicated the object to be implemented is, the time required for implementation will vary. Perhaps as a consequence of the lack of thorough analysis and decisions about this, the time issue is debated. A lengthy period for implementation would mean a loss of support and overall legitimacy for the model, but how much time for what specific changes did not have any straightforward answers. A newspaper article about a participating local authority’s initial work notes that ‘[t]he next step is a study that will take a couple of years’ (LM10). In connection with a steering group meeting, the process managers present a timetable that extends for 5 years, which triggers a heated discussion. Several municipal and county council representatives contend that the time frame is too wide and needs to be shortened. A healthcare manager says ‘if nothing happens before 2023’ his board is going to get restless, and a manager with one of the participating municipalities adds that ‘a long-term approach is hard for politicians to accept’. One process manager says during the meeting that they have been hired only until the end of the year 2019. The minutes of this meeting record that the timetable was not approved (AD4).

In an opinion piece in the local paper, some politicians impatiently declare that it is ‘time to realize the Scottish model’ in the municipalities ‘that are ready’ and that the preventative interventions cannot wait any longer (LM13). They are also critical of process management for having thus far focused on the entire county getting started simultaneously with a uniform model supported by their dissemination, instead of getting started with what is now ready in some municipalities. The time pressure surrounding the project is discussed in an interview with the process managers:

Because the politicians are much more impatient; they want it to happen preferably during their term in office, and especially if they are in the majority party. So, they have to find out that things are happening so that they settle down (Interview 1).

Consequently, the process managers felt pressure and a strong need to deliver results, so that the opened policy window would not be prematurely closed (Kingdon 2011).

Translation and adaptation of the Scottish model

A transfer of models of GIRFEC’s kind always needs some kind of translation to fit into the new environment. A report of the ongoing implementation in several Swedish cities and counties explains in different ways how the initiative ‘has taken inspiration from Scotland and is testing similar solutions in local developmental work’ (National Agency for Education & National Board of Health and Welfare 2018, 9), or that the intentions refer to ‘an adapted version of the Scottish model’ (p. 28). These writings align with the attitude towards the model in the region covered in this study.

The word ‘implementation’ occurs frequently in the analysed texts and interviews, but a lack of clarity also emerges as to whether the entire model is to be implemented or only certain parts of it, or whether various forms of adaptations will be made. It is stated in an early opinion piece that ‘[t]he Scottish model is going to be introduced in the county. Our municipality is taking the leadership role in implementing the Scottish model (“GIRFEC”)’ (LM1). Similar phrasing is found in one local authority’s budget statement (AD17). The writer of another opinion piece says that if collaboration among various actors is established, ‘the successful Scottish model can be implemented’ (LM3). Another newspaper article is more equivocal: ‘The project involves implementing the Scottish model or a version of it’ (LM9).

Several examples of interpretations of the model from the initial phase of the project are provided in the studied region. Adaptation, version and modification are frequently occurring terms. In an opinion piece, some politicians state: ‘The method is based on the “Scottish model”’ (LM7). An article in the local paper notes that ‘they are now looking at implementing a Swedish-toned version of the Scottish model’ as a potential solution to the strained finances of the Social Services Committee (LM 4; see also LM12). A newsletter distributed by project management says that ‘the work to translate the Scottish model for the county’ has commenced (AD8). In a later newsletter, under the heading ‘Adaptation of the Wellbeing Wheel’, the work in progress is described, and it is said that this part of the Scottish model ‘needs to be adapted to the reality in the county’ (AD10). In one of the participating municipalities, they will be ‘starting up a modified Scottish model in municipal services and agencies shortly before the new year’ (LM 2). As can be seen from the statements, the original model undergoes interpretations at several steps and on different levels. This process mirrors what has previously been shown in research, namely, that the ideologies, functions and interests of various sets of actors, such as politicians, steering committee members and process leaders, will be of importance when the elements of a policy stream are sorted out and combined in a potentially new way (Weinbaum and Supovitz 2010).

Reframing of current methods

A variance in the translation process that may be seen as a step farther from a mere adaptation is also argued for in the county, namely, a renaming of existing methods and ideas to fill the overall CBiK frame. The translation is in this way double-sided, including both current and imported practice.

One of the municipality’s budget and operating plans states that the Scottish model ‘can be based on the municipality’s already successful work with early intervention in a previous project’ (AD18). A representative of the healthcare system argues at a steering group meeting for a duty to ‘scale up local initiatives’ consistent with the CBiK. Similarly, a social services manager expresses that the new method ‘should be included in the current work’. One of the process managers at the same meeting relates that they have compared the Swedish BBIC (Barns behov i centrum) concept, modelled on the English Integrated Children’s System (ICS), with the Scottish praxis model and argues that it is possible to build further on the commonalities of these investigations and coordination tools. An opinion piece states that it ‘will be a logical extension to incorporate the new way of working into the city’s existing method and implement a local Scottish model’ (LM13). In several other municipalities, current services such as family centres, drop-in preschools, previous working groups and existing collaborative teams were attached to the model (AD2; AD4; AD14). A newspaper article describes an attendance team created to support pupils with problematic absences from school in one of the participating local authorities. It goes on to say that the team was started in connection with implementation of the CBiK. ‘One could say that this is our definition of how we can work to promote the best interests of the child,’ says one supervisor (LM15).

There are also clear connections to existing methods in some cases. The ‘child’s plan’, for example, is said to replace the already existing coordinated individual plan (‘SIP’), and the work should be able to harmonize with BBIC (AD 19). Also, in a group interview of key stakeholders, a head teacher observes that the joint planning process to be developed including family and children’s participation and coordinated efforts is actually what is already intended with the SIP process – even if not working as well as thought. The professionals can end up in conflict with each other and exclude the family, and the child psychiatry service does not always attend the meetings. In general, the thinking is that GIRFEC does not entail any entirely novel ideas, but nevertheless, provides a better frame in which to merge earlier intentions while it has momentum (Interview 4). This also exemplifies how a new model can be translated by local politicians and managers with the intention to boost pre-existing methods and workflows and thereby avoid additional work and extra financial expenses (Kingdon 2011).

Creation of a new model

Several texts and discussions spoke more of innovative work, rather than merely a translation or reframing of GIRFEC and current methods (LM 2, LM13, AD3, AD7, AD12, AD18). An earlier proposed decision states that the county should ‘[j]ointly develop a county-adapted model based on a shared vision and shared values inspired by the Scottish GIRFEC’ (AD7). Similar phrasing can be found in the minutes of a meeting of a healthcare committee (AD12). A newsletter from process management states that ‘[t]he parties are to outline joint models/methods related to children and young people’ (AD13). The steering group later decides to ‘in a first step, begin the developmental work with the design of the Wellbeing Wheel’ (AD3). A few local politicians write in an opinion piece that ‘we want to take the lead in Sweden and implement our own version of the model’ (LM5).

One participating local authority establishes that ‘[w]e should not simply duplicate a model, but create one based on the conditions in our county’. A head of social services says in the same article: ‘Then we will ultimately do things in somewhat different ways because we are different local authorities, but we will have a joint model …. Exactly how it will be designed remains to be seen’ (LM8). A process manager at a steering group meeting argues that: ‘Of course we should invent our own model in the county!’ One of the local management representatives, however, expressed some concern about this and wondered how ‘we are going to make sure that we teach and spread the model in a similar way so that we do not end up with nine different models’.

The minutes of a steering group meeting provide an explanation of the creation of the county’s own joint model: ‘The definition of the Wellbeing Wheel is on the agenda this February, aimed at coming to early consensus and ensuring that everyone will own the terminology’ (AD4). At a meeting of process managers, one of them thinks that ‘we should not use the term implementation’, and that ‘the model has to exist first; its implementation comes later’.

All of this with implementation is very important indeed, but it is also important to bring all the parts together so that we do not end up with a few experts sitting down and deciding on the words, on the shared language. It has to be built from the ground up. So, there should not be a gaggle of experts in a few fields who lay the foundation. You have to check with everyone because words can mean many different things (Interview 1).

In this discussion one can also find ambiguity in the understanding of what is going to be implemented: there is the GIRFEC model, on the one hand, but on the other, there should be processes for organized deviations through which the model mutates (Weinbaum and Supovitz 2010).

Discussion

The results show that the locally formulated problem and policy flows largely coincide with the main national flows, in that several of the stated problems and solutions correspond to the government and public agency agenda. The Scottish model is an example of an ‘idea right for the times’ that has gained widespread attention and dissemination (Kingdon 2011; Steiner-Khamsi 2012). The question is, what is it about the model that appeals to both Swedish government agencies and welfare workers in the child and youth sector?

A ‘selling-point’ for the model is the capacity to connect current problems, solutions and actors. Another attraction of GIRFEC is its suggestion of how to merge the UN Convention on the Rights of the Child with a comprehensive system for inter-agency collaboration for early prevention based on a proactive, holistic and inclusive approach. Concerning social work policy/practice, this articulation could function as a counterimage to some of the problematic features of contemporary practice, such as a reactive stance, long-driven specialization leading to fragmentation and individualization (Carey 2015; Dahlstedt and Lalander 2018). In addition, good results from a resource-efficient, outcome-focused and evidence-based practice could be expected, judging from the way it has been ”marketed”. This means that the model would be attractive to a large range of actors, including politicians, managers and practitioners from various fields (Coles et al. 2016).

A window of opportunity opens when the national coordinator for child and youth social services and SALAR recommend the model in work with children and young people and when the National Board of Health and Welfare and the National Agency for Education subsequently allocate project funding for local authorities to implement evidence-based, early and coordinated interventions for children (cf. Forkby 2005; Steiner-Khamsi 2006). The support of state agencies for the model and presentations by its originators in various contexts legitimize the implementation of the model. When the Scottish model is characterized as research-based and also when researchers are attached to the implementation efforts, trust in the process could be further reinforced. That a policy window is opened for a limited time becomes clear in the reported results, where politicians in particular and other decision-makers express impatience and the first seeds of worry. The ‘persistence’ and long-term commitment that leading politicians talked about in the start-up phase of the project thus seem to have waned.

The results also show that there were several interpretations of the Scottish model even before its implementation began in several of the studied municipalities and that pre-existing policies, services and activities have been coupled with the new model to a great extent.

Policy may be decided by the coupling of the politics and policy streams, which subsequently drives a search for elements in the problem stream that might serve to legitimize the already-decided policy. (Ackrill et al. 2013, 881).

Politicians, process managers and professionals are involved in this negotiation and revamping process at various levels, in line with the theory of iterative refraction (cf. Weinbaum and Supovitz 2010). As shown, the terms ‘adaptation’, ‘modification’ and ‘inspiration’ recur in several texts and utterances, which gives the impression of tremendous freedom for local authorities to depart from the original model. In some cases, there is also talk of integration with existing methods and incorporation of these in the new model. In several cases, this also seems to involve innovation and design of a new ‘version’, where implementation of the original model seems to be a minor consideration.

One of the study’s limitations is that the data used mostly derive from politicians and managers at different levels, and to a lesser extent from professionals. This skew is also found in the survey, where two-thirds of respondents consist of managers while one-third is made up of frontline personnel. Also, the respondents mainly represent the social services or schools, with a disproportionate share from the health services. However, this could be explained by the focus on the first implementation phase, which mostly engaged policy decision-makers and executive functions, and the municipal organization more than the health services. Practitioners are invited to the process in a later stage, not least in the tests of the model. Another problematic shortcoming is that neither children, young people nor their parents have been part of the implementation process yet, and consequently are not included in the research.

The pendulum effect – or the process of reimplementing the old as the new

We have shown above that there was keen interest in the implementation of a GIRFEC-inspired model in the region. As to the issue of similarities and differences in relation to current practice, there has been no analysis of what the relationship between the old and the new should be. Instead, the support thus far relies on general argumentation on the advantages of GIRFEC and how these could potentially resolve several different problems (at various levels, among various target groups and various services/organizations). This is a rather typical example of what Cohen, March, and Olsen (1972) have discussed: that solutions, problems and actors can be tossed in a ‘garbage can’ to be linked in a relatively loose manner to work in a particular decision situation. We would like to take this discussion one step further by asking about the novelty that is attributed to the model. Are there existing solutions that align with GIRFEC? We believe this question is important not only to determine the direction in which ambitions to change should be aimed but also to identify the need to analyse why earlier solutions, in such case, did not entirely succeed.

On the overall policy level, the Scandinavian (or social democratic) approach to welfare was expressed as an inspirational model in the development of GIRFEC in Scotland. The inclusive ideas that all citizens should benefit from welfare improvements and that particular care should be offered to people in difficult life situations are central pillars in both of these formulas. So is also the emphasis on a pro-active approach in which the state intervenes early to prevent later disadvantages.

Nor can the idea of broad collaboration and a holistic approach to solving problems surrounding children and young people in difficult situations be put in the GIRFEC innovation account in relation to earlier policies in Sweden. On the contrary, the collaborative philosophy is deeply rooted in the Swedish welfare system, although not as effective and uniform in practice as intended in policy.

A holistic approach is also something that has returned time and again in the context of working with children and young people in general, and specifically with those in difficult life situations. The BBIC child welfare investigation instrument that is used generally by Swedish local authorities is also a model that is expressly intended to describe children and young people from an overall perspective, incorporates multi-professional assessments and offers opportunities for follow-up. The matter of empowering children, young people and parents and making sure their voices are heard is also a central element of this strategy and of general national policy. Special educational needs assessment models used in the schools also have the intention of seeing the child in a greater, holistic context aimed at preparing coordinated action plans that correspond to the child’s needs.

Many local authorities and national actors have in recent years shown interest in ‘social investments’. The idea here is that social measures should be regarded as investments and that they must be started at an early stage in order to generate returns over time. The ambition to employ evidence-based methods is also central here – after all, if an investment is to succeed, it must be able to generate positive effects. In the striving towards more fully developed early intervention, we also find, in line with GIRFEC, strengthened home visits from child health services, but in Sweden, these are more limited to specific target groups and socially disadvantaged areas.

If one posits that what is new within GIRFEC primarily involves more advanced coordination surrounding children who have been identified as having more pronounced needs for support, one can see great similarities between the Scottish child’s plan and the existing intentions of the Swedish SIP. This is suggested in CBiK to be replaced by the child’s plan, and the intention is also to adapt to BBIC. When a SIP is prepared, a special coordinator plays a key role – similar to the concept of a lead professional in GIRFEC.

Regarding the place and accountability in the form of a named person, there is, formally speaking, already one such in the Swedish system – at least, for children enrolled in school. The head teacher is responsible for the school health services, and is thus ultimately responsible for monitoring, identifying support needs and making sure that these are realized. But there is also a central difference. The function that the named person is assumed to fill, as the first point of contact for queries and support generally for children, providing the right support at the right time, has not really been developed in Sweden. Sweden has instead incorporated such ‘one-stop’ advisory functions into other forums (family centres, citizens’ advice bureaus and the like), but these vary according to geography and the services offered by various local authorities and do not usually reach all children.

Although there may be differences in various respects that were not pointed out in the foregoing report, one can still ask what is actually to be implemented; why the existing work, generally based on the same intentions, is believed ineffective; and whether the new model is a matter of system optimization or radical system change.

Conclusions

Our conclusion is that similarities with existing intentions can have substantial impact on the positive interest in the Scottish model. That there has been an established discourse in which the GIRFEC-inspired work could be incorporated and where there are several points of intersection has probably contributed to the positive reception of the model. There has been an initial advantage in the form of a commonality of values between the new and the old that has meant the launch has been able to proceed with little conflict, even though there is a lack of clarity surrounding what is actually the same and what is different. That positive outcomes, such as fewer children in out-of-home care and a lower rate of youth offending, have been associated with GIRFEC has probably further increased interest in the model. In that the model has been able to synthesize the problems and goals of multiple actors relatively smoothly, CBiK has become a kind of ideal solution to which all parties can, at least initially, attach their specific hopes (cf. Sahlin‐Andersson 1992). By using, reinforcing and to a certain extent modifying established thinking and methods, goals are promised that the organizations already have been pursuing. In such case, the implementation of CBiK can be understood more as a system optimization whose form and content the actors seem able to predict, compared to a less controllable radical system change.

Considering the validity of this assumption, the implementation work could be concentrated on making the various components of the initiative dependent upon previously known elements and more clearly differentiating the points at which the change process should be directed. In so doing, attention could be drawn to issues related to the actual realization of the intentions expressed in GIRFEC (and other initiatives of a similar nature). In that case, it is crucial to consider a more thorough analysis of past shortcomings and how the new model can deal with concrete goal conflicts and dilemmas. These could be expected to concern how disputes between various professional claims should be settled, how early interventions can be prioritized within the limited budget of a local authority and whether antagonisms between the parties involved may arise in relation to the best interests of the child. Hence, that which is in the best interests of the child is not an essential value, but something that will in many cases need to be negotiated, and each negotiation will involve the amalgamation of differences – utterly regardless of whether various stakeholders can reach agreement on a general level.

Finally, we ask whether the positive response to GIRFEC is also connected to the major changes that have occurred in Sweden in a short time – changes that have resulted in a more unequal and stratified society where differences in living conditions have become increasingly apparent (UNICEF 2016). In this way, GIRFEC could be perceived as a hope for a return to the Nordic model and echoes of nostalgia for a bygone and more cohesive society.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the The Kamprad Family Foundation [20190094].