A Policy Decoupled from Practice: Children’s Participation in Swedish Social Assistance

In many countries, there are calls for children to be allowed to participate in social work decision-making. This article analyses child participation vis-a`-vis social assistance (SA), which is a municipal cash benefit representing the last safety net. In Sweden, SA is part of the professional field of social work and national policy recommends that children are consulted. The aim is to analyse local policies and practices regarding child participation, and the data are based on case studies in six social service offices. Unlike most participation studies, children’s participation is conceived as an institutional pressure and the concept of decoupling is used to examine how local authorities relate to participation. The findings show that in local policies there is a general openness towards participation, but in practice no efforts are being made to promote participation. The absence of participation is analysed as deriving from organisational barriers (practices are adult-centred and child welfare units are seen as responsible for participation) and the scepticism of social workers (participation is an infringement on parental obligations and children should be protected from involvement in financial issues). The article ends with a discussion on the decoupling and adequacy of children’s participation in settings comparable with Swedish SA. In this investigation, we ask, “Do they go to daycare?” “How are things at school?” “Are your kids happy at school?” and “How are things?” and so on. I think that’s enough [information] unless there are other signs. And if that’s the case, it’s not us who investigate how the kids are faring in the family. No, if we find something out that gives rise for concern, we have a duty to report it (Unit Manager, Municipality 4).


Introduction
At a policy level, there are increased calls for children to be allowed to participate in social work (e.g. Gallagher et al., 2012;Kennan et al., 2018). A major influence is the United Nations Convention on the Rights of the Child (CRC) and its Article 12, which accentuates the right of children to be consulted in relation to public decisions (Wyness, 2018). In Sweden, the CRC was ratified in 1990 and on 1 January 2020, it was incorporated into national legislation. The CRC has contributed to the institutionalisation of a perception of the necessity for participation, and in policy and research 'children's participation' has become a central principle according to which practice is evaluated (Wyness, 2018;Kennan et al., 2018;Toros, 2021). However, the way in which frontline workers should enact participation in practice is often less specified, and further empirical analysis of the prerequisites for participation in different social work settings is needed. This article concerns children's participation vis-à -vis a specific social service, namely social assistance (SA), which is a municipal means-tested cash benefit constituting the last safety net in Swedish society.
Contrary to many other countries, Swedish SA belongs to the professional field of social work (Stranz and Wiklund, 2015). Central state authorities advocate that the SA is guided by the 'child perspective' and principle of 'in the best interests of the child', which inter alia means that the voice of children are heard (IVO [The Inspectorate of Health and Social Care], 2015; NBHW [The National Board of Health and Welfare], 2013,2015). According to the national guidelines, local authorities should give a child '. . .the opportunity to express his/her opinions on issues that affect him/her. If the child does not convey his/her opinions, his/her views shall be mapped out as far as possible in some other way. The opinions of the child shall be given due weight in accordance with his/her age and the maturity. Children who are 15 years old have the right to speak on their own behalf' (NBHW, 2013, p. 24). However, no more detailed information is provided regarding the practical execution of participation.
The article is part of a research project exploring how the Swedish policy principle of the 'child perspective' is applied in an SA context. The project consists of case studies conducted in six local social services and data in the form of documents and interviews with SA representatives. In another publication (På lsson and Wiklund, 2021), we show that the ambiguity of the concept of the child perspective results in multiple local translations. Our analysis shows that it can refer to dissimilar work aspects such as securing a child's basic necessities, promoting social inclusion, identifying social problems among children, emphasising the gainful employment of parents and possessing documentation about children. A noticeable finding is the weak presence of what is conventionally regarded as a central ingredient of a children's rights perspective, namely that children should be given the opportunity to impact decisions. The difficulty concerning the realisation of children's participation is consistent with studies from other domains of child and family social work (e.g. Biljeveld et al., 2015;Ten Brummelaar et al., 2018;Toros, 2021;Vis et al., 2011). However, few studies have examined services comparable to SA, not only in Sweden but also internationally. Also, unlike most studies on participation, this article draws on institutional theory to analyse the absence of participation (cf. Boxenbaum and Jonsson, 2017). From this perspective, children's participation is viewed as an institutional pressure and the analysis focuses on how participation aligns with other prevailing goals and values structuring the service.
It should be noted that in the literature that there are a number of definitions of the term 'participation' (e.g. Wyness, 2018), but in the article, we refer to situations where the opinions and experiences of children are taken into account in decision-making. The aim is to describe and analyse policies and practices regarding child participation in SA. The following research questions are posed: How is participation outlined at a local policy level in the municipalities in question? To what extent do social workers actually consult children and how do they consider the appropriateness of children's participation in SA?
The article is structured so that we first describe the characteristics of Swedish SA. We continue by reviewing research about participation. As few studies explore SA, we mainly draw on adjacent fields of social work. After outlining our theoretical position and data, we present the findings showing that participation is a rare phenomenon, and an analysis indicates that the causes are both organisational and value-based. The article ends with a concluding section where we summarise our findings, reason about the lack of child participation in SA and extract the implications for practice.

Swedish SA to households with children
Swedish SA is a benefit that functions as a public poverty measure. The subsidy is governed by national framework legislation (the Social Services Act) and a national minimum standard rate, but local social services enjoy a great deal of freedom in realising the legislation. The main goal of SA is to ensure that recipients enjoy a 'reasonable standard of living' and self-sufficient. Politically appointed executives head the work, but the decision-making is delegated to caseworkers (Stranz and Wiklund, 2015). In terms of its organisation, SA is part of what in Anglo-Saxon literature is referred to as personal social services (PSS).
In Sweden, PSS also provide services in the fields of child protection and drug abuse treatment for adults. In practice, these fields tend to be specialised and researchers have discussed the existence of a silo mentality (Stranz and Wiklund, 2015). Traditionally, SA focuses on adults and has been described as 'familistic' by concealing children within the family unit (Fernqvist, 2011;Heimer and Palme, 2016). Studies on children's participation are rare, but document studies show that the needs of children are largely invisible (Fernqvist, 2011).
Children in families on SA, in particular long-term SA, can be considered as one of the most economically vulnerable groups of children in society. In 2018, 139,000 children-or approximately 6 per cent of all children in Sweden-lived in households that at one time or another during the year had received SA (NBHW, 2019). In recent decades, knowledge has increased concerning the impact of growing up in economic hardship corresponding to SA recipiency. There are a number of studies showing that children tend to experience it negatively, and that economic deprivation results in social exclusion relative to their peers (Ridge, 2011). Furthermore, continuous poverty is a factor that elevates the risk of a child having bad health and low grades at school (Reiss, 2013), being subject to abuse and neglect (Bywaters et al., 2016) and experiencing unemployment, health problems and substance abuse as an adult (Weitoft et al., 2008;Cooper and Stewart, 2017).

Previous research about children's participation
Theoretically, the interest in children's participation has been stimulated by the sociology of childhood, which has had an impact on the contemporary understanding of childhood (cf. Corsaro, 2017). The sociology of childhood has emphasised children as social actors, i.e. capable of acting as informants concerning their own living conditions. This should be contrasted with a more traditional perspective where children are regarded as incomplete human beings and objects in need of protection. The validity of the dichotomy is not uncontested (Tisdall, 2015), but the preference for children as social actors has resulted in participation emerging as a central theme in childhood studies (Wyness, 2018).
Under the influence of the childhood sociology, a range of studies have been conducted exploring participation in social work settings. Most studies concern services, such as child welfare, that to a greater extent than SA directly target children. Several models exist that outline how participation can be improved; among the most renown are Hart's 'Ladder of participation' (Hart, 1992), Shier's 'Pathways to Participation' (Shier, 2001) and the 'Lundy Model' (Lundy, 2007). The focus of 'participation research' is, in general, factors that foster versus impede participation (cf. Kennan et al., 2018). Fostering factors include, for example, high-quality relationships between children and social workers, that children are provided with support for participation  and the use of advocates representing children's needs (Kennan et al., 2018). Impeding factors include, for example, the time constraints of caseworkers  and frequent turnover of social workers (Biljeveld et al., 2015). In particular, the attitudes of professionals have been identified as crucial for implementing participation (Biljeveld et al., 2020) and according to research, a common obstacle is that professionals view children as vulnerable and in need of protection (Biljeveld et al., 2020;Kosher and Ben-Arieh, 2020;Toros, 2021). The research base indicates that there are difficulties concerning the realisation of participation, but also that participation can improve children's well-being and the tailoring of services (Vis et al., 2011).
At the same time, participation may be regarded as normative and there is research calling into question whether it is inherently adequate in every situation (Tisdall, 2015). For instance, there may be service users who are unwilling to engage in participation , participation can turn out to be more of a box-ticking exercise than a meaningful one (Holland et al., 2010), or the forms of participation offered might not correspond to a child's requests (Vis et al., 2011). In addition, the focus on linguistic capacities risks excluding certain groups of children (Horgan et al., 2017). Furthermore, when children are provided with the opportunity to participate, their actual impact may be limited (captured in the term 'tokenism') (Tisdall, 2015).

Children's participation according to institutional theory
To understand how local authorities deal with participation, we employ thinking from institutional theory (cf. Boxenbaum and Jonsson, 2017). We assume that social services are receptive to institutional pressures, as their legitimacy depends on the fulfilment of societal expectations. Child participation constitutes a pressure that is difficult to dismiss since it is required by both national bodies (NBHW) and supranational frameworks (CRC). However, social services tend to be subject to multiple and sometimes competing pressures, which may accommodate elements that run counter to each other. Regarding SA, it is conceivable that the adult-centredness guiding the work may be difficult to reconcile with children's participation (cf. På lsson and Wiklund, 2021). It is not uncommon that certain principles are more or less embedded in the rhetoric, while having a limited bearing on the day-to-day work. A concept used to describe such a gap between policy and practice is decoupling (Meyer and Rowan, 1977;Boxenbaum and Jonsson, 2017). The reasons behind decoupling may vary and it is not necessarily adopted consciously, but it often occurs when a principle is difficult to integrate into the day-to-day work, because it collides with other principles structuring practice or because professionals lack a belief in its adequacy. To understand why decoupling occurs, it is crucial to explore how professionals appraise child participation relative to other organisational goals.

Methods and material
The study consists of case studies conducted in six municipalities and a regional ethical committee in Stockholm approved the design in advance. In order to select the municipalities, we analysed national statistics on SA. Municipalities with fewer than 15,000 inhabitants were excluded from the sample group. The statistics displayed (i) the proportion of families with children receiving SA and (ii) the sum of the benefits granted to these families. These measures were used as dependent variables in regression models (OLS/Ordinary least squares) controlling for socio-economic factors (e.g. municipal unemployment rates and the standard cost of PSS) and demographics (e.g. the proportion of single households and inhabitants with a foreign background residing in the municipalities). The ambition of the regression analyses was to create robust models with a high predictive value of municipal variations in granting SA to families with children, controlling for structural conditions. Thereafter, we sampled residuals; i.e. municipalities scoring unexpectedly high ('generous') and unexpectedly low ('restrictive') in this respect. The surmise was that a variation would increase the likelihood of heterogenic local strategies towards households with children. The final sample consisted of six municipalities (three 'generous' and three 'restrictive'). The population size ranged between 22,000 and 95,000 inhabitants per municipality.
We collected three kinds of data. First, local guidelines uncovering formal policies intended to govern SA. All municipalities had guidelines intended to support social workers in their decision-making. The guidelines delineated municipal policy and values that should permeate the decision-making as well as the advice given to social workers regarding benefits. The guidelines differed in length and detail, but all contained passages specifying administration relative to households with children. The documents were sometimes available on the website of municipalities, but since these are for internal use only they were mostly e-mailed to us. The documents were subject to qualitative content analysis (Bowen, 2009) in that we extracted and analysed sections of texts from the guidelines that related to participation.
Secondly, we analysed case files (n ¼ 344) concerning households with children receiving SA at the initiation of the data collection (September 2018). When we visited the municipalities, employees provided us with lists covering households with children. In one municipality, we conducted a total sample (n ¼ 92), while in another, we pursued randomised samples of approximately fifty case files per municipality in order to prevent the sampling from being systematically skewed. Depending on the number of households with children each municipality, we followed a principle where we selected every other, every fourth or every fifth, etc. from the lists. The case files were studied in the offices of the social services and took approximately twenty working days in total. Various anonymised pieces of information were extracted. In this article, we have analysed notes written by caseworkers over a one-year period (September 2017 to September 2018). The information studied regarded (potential) notes on the frequency and content of encounters/conversations (e.g. face-to-face, over the telephone, during home visits/meetings with parents) with children.
Thirdly, we analysed information from twenty-four interviews (including fifty-seven participants) performed in the autumn of 2018 and spring of 2019. The participants were informed about the research project and gave their consent to participate. Two interviews were performed by both authors and the remainder were conducted by one of the authors. Interviews were conducted with employees in various organisational positions in order to gain knowledge about both policies and the reasoning of case workers. The number of interviews in each municipality was decided so that they provided data that could sufficiently address our research questions. In each municipality, we conducted an equal number of interviews with the heads of the local committees (n ¼ 6), heads of the PSS (n ¼ 6), unit managers heading the operational work (n ¼ 8) and social workers processing SA (n ¼ 37). Individual interviews were conducted with the heads of local committees/PSS and unit managers (apart from two municipalities which had two managers), while group interviews were conducted with social workers (the number of participants ranged between two and nine). The interviews were semi-structured and used a protocol covering questions about the administration regarding children. One interview theme explored respondents' thoughts on participation. We deliberately asked about attitudes towards encountering, involving and interviewing children in the decision-making. We also urged the interviewees to elaborate on their opinions. The interviews lasted between one hour and one and half hour and were transcribed verbatim. Interview transcripts were read through several times and text sections dealing with respondents' opinions on child participation were identified. The texts were subject to a thematic analysis (Braun and Clarke, 2012) where we coded observations in terms of what they expressed, while simultaneously relating them to theoretical thinking on organisations as guided by multiple values. Two overall themes were established that could explain the misgivings regarding child participation (organisational and value-based), which were divided into four subthemes. The material of the article is summarised in Table 1.
Besides the separate analysis described above, analytical work was also done where we interlinked findings from the three data sources. In this phase, we related the findings to the concept of decoupling in order to draw inferences. No salient municipal differences could be discerned between 'generous' and 'restrictive' municipalities. Hence, the study is mainly based on municipal commonalities. By building on case studies, our ambition has been what is commonly referred to as analytical generalisation (Alvesson and Skö ldberg, 2009). This means that even though further studies are required to corroborate our claims, it is reasonable to believe that the analysis is more or less applicable in contexts where the organisation of SA, as well as institutional values towards children are similar to the Swedish context.

Findings
The findings begin with a description of the respondents. Then follows an analysis of the participation policy outlined in the guidelines and formulated by management (participation in local policies-a general openness). Thereafter, we analyse case file data and considerations of unit managers/case workers (participation in practice-decoupling and scepticism).

Respondents
Just over half of the unit managers and case workers have a degree in social workers. The others have Bachelor's degrees in adjacent fields (e.g. sociology, political science) or are counsellors. Over half of the participants are under the age of forty years. Only five are male.

Participation in local policies-a general openness
The guidelines differ in terms of whether they address children's participation, but the general picture is that there is a principle of openness. The guidelines from two local social services do not specifically mention taking children's views into account, but in the remaining four municipalities, the guidelines convey such information. Two refer to Article 12 of the CRC stipulating a child's right to be heard. These municipalities plus one other also refer to the fact that due weight should be given to a child's opinions in accordance with his/her age and maturity. Three municipalities state that, given parental consent, it is possible to consult children to ensure that the benefit satisfies the needs of children. One municipality has guidelines stating that social workers must respect the parents if they are unwilling to do so, but also that children who are consulted are informed about the purpose. All four municipalities that explicitly recommend children's participation state that children's needs may also be investigated by asking the parents. In this respect, two municipalities have drawn up checklists covering questions that should be put to the parents about their children (e.g. their leisure activities, health and schooling). One municipality explicitly dissuades social workers from using children as interpreters, where the parents' grasp of the Swedish language is poor. As has been stressed, Swedish SA is governed by politically appointed executives, who are responsible for implementing the participation (cf. Stranz and Wiklund, 2015). The impression gained from interviews with local committee heads is that they are not against including children in principle, but they do not impose any systematic requirements. They mostly stress a lack of detailed knowledge about the practical execution of SA and that they delegate the issue of participation to frontline workers. Heads of PSS are slightly more familiar with the administration, but claim that participation is not an issue that has been discussed in detail. The following line of reasoning by a head of PSS is illustrative: IP: I don't think that they do that [meet children] and there aren't any set rules. Naturally, sometimes children participate in meetings, but I'm pretty sure that the case workers don't have any structured interviews with children. Then perhaps you exchange a word or two with a child sometimes. You always need to think about the organisation and who should do what, but then I think that demands a slightly different structure at work (Head of PSS, Municipality 3).
A couple of heads of PSS mention the CRC and that there might be reason to revise how they deal with participation, but their general belief is that the current policy suffices.
In summary, and following national guidelines (NBHW, 2015), children's participation has left a mark on the local policy level in several municipalities. This indicates that it tends to be important to relate to the external pressure requiring that children's participation should be part of SA (cf. Boxenbaum and Jonsson, 2017). However, no formalised efforts exist to promote participation.

Participation in practice-decoupling and scepticism
Despite several municipalities having guidelines about participation, social workers scarcely encounter children and they are considerably more hesitant than the management regarding meeting or listening to children as part of their decision-making.

Children as informants-virtually non-existent
According to the study of the file notes covering 344 households with children, social workers virtually never encounter children (either faceto-face, over the telephone or digitally). Five of the municipalities have not held any individual meetings with children during the year studied. In one municipality (the municipality where the total population was studied), there were six examples of young people turning to the social services and receiving personal appointments and in a further five instances, there were children who had called social workers. In these rare instances of personal contact, the conversations concerned, for example, a young person wishing to live on his/her own or that a family needed larger accommodation, but there were also general conversations about their situation at school, health, summer job prospects and descriptions of family conflicts. On twenty-five occasions, the file notes reveal that children (usually young children) were present at meetings between parents and social workers or during home visits. However, the meetings did not specifically concern children's needs and there is no indication that they occurred due to an explicit strategy to involve children. Hence, case file data do not indicate that children are acknowledged as social actors in relation to SA decision-making.
Children's participation as an organisationally alien principle So how do social workers reflect on (the absence of) child participation? Regarding organisational barriers, time constraints highlighted in earlier studies (e.g. Gallagher et al., 2012) only appear to play a minor role. The main impression is that the hesitation stems from the fact that child participation is a principle that is alien to the main goals of SA.
The adult-centredness of SA To begin with, respondents interpret SA as geared towards the financial situation of adults-not children. The primary principle structuring work is to investigate whether the applicants are entitled to the subsidy and to help households become self-sufficient (cf. Pålsson and Wiklund, 2021). Consequently, in terms of investigation procedures, there is no discrimination between households with/without children. To make adequate decisions, it is sufficient to have information on the financial situation of a household and that the adult adheres to his/her individual planning (e.g. job seeking). Investigations about the life situation of children are considered to fall outside the main objective of SA. Information about a child that might be relevant for a decision is, for example, whether the child attends school, is ill, or has a certain hobby, and respondents argue that such basic information can be collected from the parents. Some respondents have difficulties discerning the purpose of consulting children, since their services are largely adult-oriented. This is expressed in the following extract: It's the adults who are responsible for providing for their children and we work with enabling adults to look after and fend for themselves. Therefore, quite frankly, we look at their income and expenditure and also what they've done to try to resolve things in a different way and the long-term plan is about them supporting themselves (Unit Manager, Municipality 5).
Child welfare units are the child experts Also, child participation is an issue viewed as pertaining to the work of child welfare units. As mentioned above, Swedish SA is part of the PSS (which also conducts child welfare investigations) and in the municipalities studied, as in most municipalities, these areas are carried out by different units (Stranz and Wiklund, 2015). Some claim that they are not specifically trained to interview children-again, their area of expertise is assessing eligibility and assisting adults to become self-sufficient. In this respect, respondents mean that if there is a reason to investigate children's needs, it is logical that it is the task of child welfare workers. A common view is that if they are concerned about whether parents meet a child's financial needs, they should not initiate contact with the child in question themselves, but make a child welfare referral instead.
In this investigation, we ask, "Do they go to daycare?" "How are things at school?" "Are your kids happy at school?" and "How are things?" and so on. I think that's enough [information] unless there are other signs. And if that's the case, it's not us who investigate how the kids are faring in the family. No, if we find something out that gives rise for concern, we have a duty to report it (Unit Manager, Municipality 4).
Summarising the organisational barriers, children's participation is a principle that does not seem to fit with the current division of work and goals. The main principle that is given primacy is adult-centredness, i.e. that investigations and services target adults (cf. Fernqvist, 2011;O'Brien and Salonen, 2011). This is reinforced by the fact that in Swedish PSS, social workers processing SA are in general not trained or expected to involve children. Instead, the responsibility of involving children lies with separate child welfare units devoted to investigating the needs of children Scepticism towards participation for the sake of the parents and the child The analysis also shows that social workers are sceptical towards children's participation. The scepticism boils down to values, where it is inappropriate to intrude on the lives of parents/children in the context of SA (cf. Biljeveld et al., 2015).

Children's well-being-a parental obligation
The scepticism partly stems from the well-being of children foremost being regarded as a parental obligation and that they should not take over this responsibility. Several respondents see it as an essential part of their work to ingrain in parents that it is they, and not the social services, who are the main providers for their children's well-being. Thus, refraining from involving the children can be analysed as a strategy to strengthen parental ability in order to promote self-sufficiency. According to this line of reasoning, involving children could compromise parents' sense of responsibility, which is reflected in the following group interview: Social worker 1: That we shouldn't take the responsibility away from the parents, because it's still the parents that are mainly responsible for their children.
Social worker 2: [The responsibility] for planning your finances. And it's not just a matter of granting subsidies, because that will lead to the parents not taking responsibility and not planning their expenditure in the future. So you need to bear that in mind (Social workers, Municipality 6).
Another aspect of parental responsibility reflects a sort of ethical standpoint that intruding on the family is stigmatising. There are those that believe that SA recipiency does not automatically give social services the right to investigate children. The assumption is that families on SA are entrusted with taking parental responsibility to provide their children with the necessary support. The only time intruding on the integrity of families is seen as legitimate, is if there are any signs of child maltreatment. This is exemplified in the following extract: Just because you come to us, doesn't necessarily mean that you're a bad parent or that we have to map out their situation. Instead, you might just be experiencing a temporary financial problem. Perhaps you've just arrived in Sweden; perhaps you haven't learnt to speak Swedish. You might still be a good parent and want what's best for your child. We don't want to say that everyone who comes to us is a bad parent. So there's a risk that that's the case (Unit Manager, Municipality 2).

Children should be protected from participating in financial issues
Another source of the scepticism comes from that it is regarded as potentially harmful for children to be involved. Some respondents advise against involving children for fear of it leading to children taking responsibility for the financial situation of the family. With such a view, participation is understood as potentially distressing rather than empowering. This can be viewed as an example of a protection perspective that has been discussed in prior studies (Biljeveld et al., 2015;Kosher and Ben-Arieh, 2020;Toros, 2021). For instance, several respondents claim that children should not worry about the family economy, which could be the case if children are expected to give their opinions on the benefit. Social worker 1: Or have a headache over things (Social workers, Municipality 1).
Thus, the absence of participation can also be analysed as a corollary of scepticism among social workers towards consulting children. At the core lies the fact that approaching children is regarded as stigmatising vis-à -vis the parents and there are fears that participation may lead to children being made responsible for financial issues (cf. Biljeveld et al., 2020;Kosher and Ben-Arieh, 2020;Toros, 2021).

Concluding discussion
The last few decades have seen stronger calls to enable children's participation in decision-making. The CRC is a strong influence as it underlines the rights of children to affect decision-making (Wyness, 2018). In Sweden, SA is a service in relation to which national policy recommends that children are consulted. SA is a cash benefit, which targets one of the economically most marginalised groups of children. In Sweden, SA is part of the professional field of social work. The aim of this study has been to describe and analyse local policies and practices regarding child participation in SA. The study was designed as case studies of six local social services which differed in terms of generosity regarding the approval of subsidies to families with children.

Summary and discussion
The main findings show that most of the municipalities studied have local guidelines outlining participation, and management tends to express a principle of, albeit unspecific, openness towards participation. However, there are no systematic efforts to give children a voice. A case file analysis reveals that children are scarcely encountered and from the perspective of the professionals, the idea that children should participate in SA is weak. The absence of participation can be understood as deriving from organisational barriers (SA is adult-centred, child welfare units are seen as responsible for children) and the scepticism of social workers (child participation is an infringement on parental obligations, children should be protected from participating in financial issues).
The absence of participation can be understood as an expression of decoupling (cf. Meyer and Rowan, 1977) between national and local policy on the one hand and practical administration on the other. Even in municipalities where policy encourages participation, there is no organisation fostering participation and there is a professional hesitation towards it. The fact that local policies tend to describe children's participation as something to pay attention to, mirrors that it is perceived as important to obey in order to gain legitimacy. However, the closer to the frontline the principle of child participation is analysed, the more alien its practical aptness appears. This decoupling can be understood based on the fact that SA is guided by multiple goals that can be difficult for frontline workers to concurrently abide by (cf. Boxenbaum and Jonsson, 2017). The fact that child participation is disregarded indicates that it is perceived as not suitable to current organisational structures and other, for the practical administration more important values. According to the study, this partly boils down to a silo mentality between units managing SA and child protection respectively, where the latter is seen as responsible for the predicament of children and thus, by extension, their participation. In addition, the decoupling is underscored by value-based scepticism towards the suitability of approaching children in families on SA for the sake of the integrity of the parents/children. These two factors are likely to mutually reinforce each other, which contributes to rendering children's participation a lowpriority issue.
If children's participation in SA is viewed as something desirable, the findings of the study are grim. In contrast to international and national policy, there is no evidence to suggest that children in families on SA are acknowledged as legitimate stakeholders who are competent to express their experiences (cf. Corsaro, 2017). Rather, the findings corroborate research showing that children are invisible in SA (Fernqvist, 2011) and that children are treated as objects of concern who should be protected from involvement. This means that social workers are oblivious to how children experience the decisions. To obtain such information appears not least important in the light of studies indicating negative effects associated with economic hardship during childhood (e.g. Cooper and Stewart, 2017;Ridge, 2011).
On the other hand, children's participation may not be adequate in all situations (e.g. Tisdall, 2015). Decoupling tends to occur when professionals find certain practices not suited to their main working tasks (cf. Boxenbaum and Jonsson, 2017) and you may query whether there is any validity in the scepticism towards participation identified in this study. Here, it should be acknowledged that SA in certain aspects deviates from other social services (e.g. child welfare) subject to participation requirements. The main differences are that SA has traditionally been geared exclusively towards adults and that the service can be said to belong to the public cash benefit system. An argument against child participation in SA is that public institutions in general consider the way income is allocated in a family to be a parental responsibility. From an ethical standpoint, it could be regarded as stigmatising to sidestep such principles just because families are dependent on SA (cf. Smith et al., 2012). Further, given the established adult-centredness of SA, it is not evident that routine child interviews would lead to children having an impact on decisions in a substantial way (cf. Holland et al., 2010). Arguably, the SA practice would need to be reorganised and social workers' attitudes towards participation shifted, in order to avoid tokenism (i.e. a symbolic impact), which according to previous research is common (cf. Tisdall, 2015).

Conclusions
To conclude, consistent with studies from other domains of social work (e.g. Vis et al., 2011;Gallagher et al., 2012;Biljeveld et al., 2015;Ten Brummelaar et al., 2018;Toros, 2021), this study provides further evidence of the difficulties concerning the realisation of children's participation. Unlike most studies on children's participation, we have, based on institutional theory, analysed participation as an institutional pressure that in practice is difficult to align with the prevailing organisation and goals of SA. A further analysis of the prerequisites for participation in different social work settings is needed, since it is an issue that is likely to continue to be at the top of the agenda, both in social work theory and practice.

Implications for practice
For social work practice, the conclusions accentuate a continuous need to discuss the function of participation in different settings as well as viable strategies for professionals to realise a meaningful participation of children in adult-centred services. Here, it appears important to discuss the value conflicts that professionals may face, and the readjustments that welfare institutions have to make, when they implement principles that challenge how work is usually conducted. More clarity from supranational and national agencies regarding expectations of how professionals may execute participation in different social work settings would be welcome.

Limitations
Concerning the limitations of the findings, it is important to note that the article focuses on case file data and professional perceptions, and hence, service users' opinions on participation are not analysed. Further, the study consists of case studies and hence, the findings cannot be directly extrapolated to wider population and contexts. However, the study elucidates the difficulties of adopting children's participation in adult-centred settings comparable with Swedish SA.