Keywords

Introduction

Child welfare caseworkers operate in a professional field characterised by juridification and hard-law initiatives as well as increased standardised procedures (Ponnert & Johansson, 2018). Although the use of Barnahus is not mandatory by law in Sweden, the model has rapidly become a standard normative procedure.Footnote 1 Specialisation and professionalisation tend to result in increased functional and structural differentiation of organisations (Axelsson & Bihari Axelsson, 2006), examples of which include both Barnahus and increased organisational specialisation into different units within the child welfare services. This chapter examines these developments by using the perspective of child welfare caseworkers,Footnote 2 with a focus on child welfare investigations (henceforth “child investigations” or “investigations”) in cases of suspected violence.

For almost a decade, Sweden’s child welfare services have been required by law to initiate an investigation when they receive knowledge that a child might have been exposed to, or witnessed, violence, or other abuses by or directed to a close relative (SOSFS, 2014:4, 6:1). We might look at this legal regulation as an attempt to strengthen the protection of children from violence and to reduce the risk that children who are referred to the child welfare services will not be thoroughly investigated. But an obligation to initiate a child welfare investigation might also create legal, professional, and ethical tensions (Ponnert, 2023). As of November 2022, initiating an investigation is no longer mandatory, partly due to interpretation problems in practice discovered in a follow-up and analysis by Sweden’s National Board of Health and Welfare (2021, 2022).

Violence in close relationships, however, is an area where the law is still complemented by binding regulations and general advice from the National Board of Health and Welfare (HSLF-LF 2022: 39), including a reminder to initiate investigations of children when appropriate, and an obligation to provide support to children who have been exposed to or witnessed violence in close relationships. The child welfare service’s responsibility in the event of violence has accordingly come to be emphasised and regulated in more detail in legislation in recent years. Since August 2021, such legislation also includes a legal responsibility to ensure that those who have subjected relatives to violence or abuse will change their behaviour (Prop. 2020/21:163; 5:11a Social Services Act [SSA]Footnote 3). Subjecting a child to witnessing violent acts between people close to the child is also now a crime, as of July 2021 (Prop. 2020/21/21:170; 4:3 Criminal Code), which may result in an increased number of children being subject to parallel investigations and hence referred to Barnahus in Sweden.

Although one aim of Barnahus is to provide an integrated and child-centred model for investigating suspected child abuse by using a “one-door principle” to avoid secondary victimisation, the opposite development may occur in how the child welfare services are organised. Today the services are characterised by an intra-organisational specialisation (Blom et al., 2009; Grell et al., 2022). In Sweden, municipalities have specific intake units (sometimes called reception unitsFootnote 4), which specialise in making the immediate risk assessment of referrals/reports concerning children at risk, to make a pre-assessment, and (within 14 days) to decide whether an investigation should be initiated (National Board of Health and Welfare, 2019, pp. 8–9 and 53). If an investigation is initiated, the case is then transferred to a specific investigation unit, where another child welfare worker (henceforth “social worker”) conducts the actual investigation. Social workers at separate support/intervention units also provide support to children and their families and follow-up children in care. When they have reason to believe that a child has been subjected to violence, social workers thus need to collaborate within the child welfare services and to do so in accordance with their own legal obligations and organisational procedures; they must also collaborate with the police and Barnahus. The challenges found in child investigations from a social worker’s perspective are thus important to highlight.

While research on Barnahus from the perspective of social workers is scarce, Mosegaard Søbjerg (2017) has shown that, from a broader holistic approach, social workers in Denmark may experience challenges while working with children and their families when Barnahus is involved, and police reports may also challenge collaborations with families (Mosegaard Søbjerg 2017; Johansson, 2011). Johansson (2011, 2017) has found that the inter-professional collaboration in Barnahus in Sweden is generally characterised by juridification, in the sense that a logic oriented on criminal law tends to guide work and override an approach oriented on social work. In my own thesis (Ponnert, 2007), which was carried out before Barnahus and specific intake units became standard in child welfare services, I also found several aspects of juridification influencing the compulsory care process within the child welfare services. For instance, as decisions were handed over to the administrative court, social workers expressed a need to await enough “legal evidence” before intervening, although they were morally and professionally convinced that compulsory care was required.

One question that then arises is the question of what happens with social workers’ risk assessments today, when they are both filtered internally at different levels within the child welfare services organisation and in relation to Barnahus, and a logic that is oriented towards criminal law (Johansson, 2011, 2017). In this chapter, I will explore a few organisational tensions within the child welfare services (intra-organisational tensions) and in the collaboration with Barnahus and other agencies (inter-organisational tensions), and how these tensions might affect child welfare investigations and practice concerning suspected violence. I ask which dilemmas can be discerned when it comes to violence from a social worker’s perspective and in relation to Barnahus, and what are the possible consequences from a child’s perspective?

This chapter’s discussion and results are based on a Swedish research project on social workers’ interpretations of violence in child investigations.Footnote 5 The Swedish system has been described as being oriented towards family service (Gilbert, 2012), with low legal thresholds for mandatory reports and for child welfare investigations. A child investigation proceeds most interventions from the child welfare services in Sweden, and interventions are rarely possible as a “service” without previous investigation. Hence not only are the most serious cases or referrals investigated, but so are cases where the family applies for, or only needs, minor support, such as counselling. Voluntary support, however, is preferred and is often used in cases of maltreatment as well, if the child’s custodians give their consent. As an example, babies are primarily removed to foster care as a result of voluntary or emergency measures (Hestbaek et al., 2020), and the division between support/welfare and protection vis-à-vis interventions can accordingly be blurred in practice. Pösö et al. (2014) have problematised the family-service orientation in Nordic countries due to its principle of providing the “least intrusive” form of intervention, since this approach may result in an overly high threshold for providing out-of-home care, even when children experience adverse home conditions. The large number of teenagers with behavioural problems in out-of-home placements in the Nordic countries may be an unintended consequence of a lack of sufficient protective interventions at an early age (Pösö et al., 2014).

Theoretical Framework

The analysis presented below is grounded in two theoretical frameworks: institutional theory and perspectives on professional discretion. I argue that both approaches are necessary to understand the challenges and consequences of different collaboration practices regarding child welfare investigations.

Institutional Theory and Organisational Perspectives

Institutional theory focuses on how organisations adapt to the institutional context (Di Maggio & Powell, 1983; Meyer & Rowan, 1977). Researchers have elaborated on the concepts of integration and specialisation from different perspectives to understand such adaptions. Lawrence and Lorsch (1967), for instance, claimed that organisations tend to become differentiated into parts in order to correspond to changes in the institutional environment, but they also found that these parts need to be integrated and provide a unified perspective for the organisation to thrive. Today, the child welfare services in most Swedish municipalities are also characterised by an intra-organisational specialisation that can be described as both problem-based and function-based specialisations (Blom et al., 2009). In a problem-based specialisation, separate units manage different social issues or target groups (such as children), whereas a function-based specialisation usually refers to the separation between authoritative work and decisions (such as conducting investigations and making decisions) and the provision of social support and treatment (Blom et al., 2009).

Axelsson and Bihari Axelsson (2006) relate the increased specialisation and professionalisation in public health to increased differentiation on different levels, both within organisations and at the inter-organisational level. They claim that functional differentiation results in structural differentiation, which may cause a fragmentation of responsibility, which in turn may result in a need for horizontal and vertical integration between organisations or units through coordination, co-operation, contracting, and/or collaboration. Integration, however, is a broadly used concept (Fisher & Elnitsky, 2012) that has also been used to describe organisational structures and specialisation (Grell et al., 2022; Smith et al., 2018). In this chapter, “organisational integration” refers to an organisational structure that avoids the differentiation of work into different specialised units, while “integration practices” are defined as collaborative forms of integration work.

Drawing on previous studies on collaboration in Barnahus, one starting point for the analysis is also that the collaboration in Barnahus has often been characterised by juridification, in the sense that the power dynamics result in a logic oriented on criminal law primarily setting the agenda, whereas the treatment-oriented logic (represented by social workers) tends to take a back seat (Johansson, 2011, 2017). In Johansson’s study (2011) and in interviews with staff at six local Swedish Barnahus, she showed that social workers may be excluded from co-hearings due to the secrecy of the criminal investigation; she also found that Barnahus coordinators might propose that social workers not inform a child’s custodians about police reports before the investigative interview at Barnahus has been conducted (Johansson, 2011, 2017). King and Piper (1995, pp. 132–138) have described such a juridification phenomenon whereby the legal system becomes the dominant discourse once the law or justice perspective “enslaves” child welfare knowledge. But juridification is a broad concept that may relate to many different dimensions, including increased legal regulations and judicial power or a tendency for people to think of themselves and others as legal subjects (Blichner & Molander, 2008). In this chapter, I refer to juridification as a regulatory process that tends to result in increased “instrumental orientation” (Teubner, 1987), which occurs as a result of both legal and organisational rules and norms for practice.

Social Workers’ Discretion

Discretion has been described as a doughnut hole, surrounded by a belt of restrictions (Dworkin, 1977). Social workers belong to a professional group that is usually associated with considerable discretion or room for manoeuvring to make assessments and decisions based on their specific knowledge and judgement. But researchers often distinguish between the actual room for manoeuvre, or “discretion as granted”, and the “discretion as used” (Hupe, 2013), as well as between professionals’ “discretionary reasoning” within their “discretionary space” (Molander, 2016). Molander (2016) claims that when discretion is split up among several actors, the power of each person is reduced. Such “divisive mechanisms” thus narrow the discretionary space and may also (either by intention or otherwise) function as a delaying mechanism. Hood (2020) has elaborated on how sharing discretion can be seen as a way of spreading responsibility. He uses the concept of “pooled discretion” or “hanging together” to describe situations where the responsibility for decisions is shared but not avoided. The functional specialisation of social services into several units may be seen as an example. Another way to spread responsibility is by “semi-delegated discretion”, where the discretion is fully or to some extent, passed on and delegated to another actor, also referred to as “passing the buck” (Hood 2020). In this chapter, this situation may be related to how the main authority to talk to children who have disclosed exposure to violence is passed on or transferred between agencies: either vertically (from mandatory reporters to child welfare services, and from child welfare services to Barnahus) or horizontally, between different units.

Method and Empirical Material

The analysis in this chapter is based on qualitative interviews conducted with 16 social workers in seven municipalities of different sizes in Sweden. One interview was conducted in each municipality. Six of the interviews were minor group interviews with two or three social workers, and one was an individual interview. In all municipalities, the child welfare services were organisationally specialised, and reports/referrals were first handled by an intake unit and transferred to an investigation unit (if an investigation was seen as necessary). All municipalities were also connected to a Barnahus (three different local Barnahus). The participants in the interviews all worked with, or had previous experience of, conducting child investigations. At the time of the interviews, 11 social workers worked at intake units, four at investigation units (of which one had previous experience of intake units), and one with following up children in care. The interviews lasted approximately 60–80 minutes and took place during 2020 and January 2021. Six interviews were conducted at the caseworkers’ workplaces, while one was a digital interview. A semi-structured interview guide was used guided by different themes; the organisation of the investigative work, the interpretation of legal regulations for child investigations in cases of violence, the various risks and opportunities involved in child investigations, and any legal tensions. Specific questions were also related to if and how Barnahus and police reports were used, as well as possibilities and challenges related to that aspect. All interviews were recorded and transcribed verbatim. The author translated the quotes used in this article into English; the quotes were then lightly edited for clarity.

The focus in the analysis is on illuminating intra- and inter-organisational tensions in social investigative work in cases of suspected violence, including the pre-assessment that occurs prior to a child investigation being formally opened. The interviews were analysed according to the principles of thematic analysis (Braun & Clarke, 2006), guided by the specific questions and theoretical approaches described in this chapter. I searched for quotes that displayed organisational and legal tensions in relation to Barnahus and the work within the child welfare services, noting initial codes. I then reread the transcriptions and searched for any overarching themes before finally reviewing, defining, and naming the themes. The study has been approved by the ethics review authority in Sweden. In the presentation, each social worker is represented by a randomly chosen letter (A-P).

Analysis

I will start by discussing how the organisational specialisation within the child welfare services affects the work with investigations, as well as showing examples of the integration practices that social workers use to avoid performing fragmented work. I then show how the immediate protection assessment conducted at intake units may be affected (and delayed) by new intra-organisational interpretations of the legal framework, the division of work between different units, and the Barnahus procedure itself. This organisational process involves the risk of what I call “administrative thresholds” for taking the child into care. Instead, social workers use different integration practices to try to keep the child secure before and after the experience with Barnahus. Finally, I discuss child investigations and professionals who respond to children’s disclosure of violence from an inter-collaborative perspective, also taking mandatory reporting into account, and the risk of what I refer to as “accumulated silence”.

The Administration and Integration of Work

The tasks at intake units in most municipalities include receiving reports, making immediate protection assessments, deciding on investigations and police reports in the event of violence, and, if necessary, making immediate arrangements (National Board of Health and Welfare, 2019), which was also the case in this study. In one municipality, however, the social workers stated that whether or not a police report was to be made was decided at the investigation unit. The work involved at intake units also in general (not only in cases of suspected violence) allows limited time for contact with children and parents, since the pre-assessment is restricted by law to 14 days. Also, within a pre-assessment, the social workers are only allowed to talk to the children and their custodians and the person who made the referral; further contacts are only allowed if a child welfare investigation has been initiated (11:2 SSA). As one social worker said, “We usually only have the chance to meet with them once during a pre-assessment”. Since initiating a child investigation in the event of suspicion of violence was also mandatory at the time of the study, the function of the intake units primarily seemed to be to urgently assess and then administer the violence by passing the case on to the investigation unit and, where appropriate, reporting to the police and contacting Barnahus.

F::

So, with pre-assessments concerning violence, there it happens that the intake unit doesn’t meet the families at all; they only do the first bits of processing the care—to actualise the case and make a protection assessment if you choose to do so. And to make the decision to initiate [a child welfare investigation] or hand it over to the head of unit.

—Interview 3, intake unit

At the same time, some emphasised that immediate protection assessments required more professional expertise and experience, and that those working at intake units often had long experience as investigators. Not talking to the family or the child before transferring the case may also be seen as a form of integration practice, to avoid the involvement of too many social workers. But some social workers could view the limited time for pre-assessments at intake units (14 days, according to 11:1a of the SSA) and the organisational specialisation as problematic and as causing unnecessary delays, since each unit has its own routines for case distribution.

B::

I think [things are] going too fast, because we have a case for 14 days. As well as having an investigation for four months, we’d need to have [the case] a little bit longer than a few days, because that’s a process as well. I’d have to meet the woman or the child who’s experienced threats and violence one more time. But as you say [case worker], it should be farmed out [from our unit] as soon as possible.

A::

As soon as possible, because it’s so obvious that we’re going to start [a child welfare investigation], and then we’re not going to have the case anymore.

B::

And then it might still be in the pile, because they [the investigation unit] only distribute the cases once a week, and then it’ll still be a week [where nothing happens]. If you transfer the case [to the investigation unit] on a Friday, it won’t be handed over [to a social worker] until the next Thursday or Friday anyway.

—Interview 1, intake unit

The social workers above also noted that the process was set in motion once they had met a child and parent, and that this process then ended abruptly after a meeting; they also expressed how the division of work could result in what Molander (2016) calls a “delaying mechanism”. Hjärpe (2022) has analysed how social workers can relate in various ways to the time limit of four months for child investigations, including task-oriented, relational, or clock-oriented approaches. The discussion above indicates a task-oriented organisational specialisation, where the case should preferably proceed quickly when the task of the respective unit has been conducted. The social workers themselves, however, could perceive a need for a relational perspective that typically requires time. In general, the strictness of the formal boundaries between the intake unit and the investigation unit may differ slightly between municipalities. In some municipalities, the case could remain in the intake unit somewhat longer, which may also provide discretionary space for integration practices. Social workers at intake units and investigation units could also collaborate when a need existed for immediate protection of a child and urgent compulsory care.

Interviewer::

So, if it’s about immediate compulsory care, then you [social workers at the intake unit] do it [the out-of-home placement]?

C and D::

Mm…

E::

But usually we also connect a social worker from the investigation unit.

D::

Yes.

E::

We [case workers at the intake unit] do it [the out-of-home placement] together with the investigating social worker. /…/ [The case may involve] a child at school who doesn’t dare go home. Sometimes it’s us, the people from the intake unit, running [the case]. But a child welfare investigation will happen, since we take action. If you assess that the child needs urgent protection, then usually one of us [social workers at the intake unit] leave, and you connect a social worker from the investigation unit.

—Interview 2, intake unit

The organisational specialisation into separate units, as well as the limited legal authorities within a pre-assessment, seemed to result in an administrative focus for the work that went on at intake units, especially when cases concerned violence, since such cases are quickly passed on for further investigation at the investigation unit. In cases of urgent protection, however, social workers may also cross their intra-organisational boundaries and collaborate across units, which may be seen as an example integration practice.

Administrative and Normative Thresholds for Urgent Protection

One question is what consequences the internal specialisation may have for the first urgent protection assessment (the assessment if something needs to be done right away, based on the information in a referral) and for the pre-assessment that follows, given the limited legal authority and time available within a pre-assessment. In practice, social workers at intake units must decide relatively quickly whether immediate compulsory care needs to be provided, or whether the process at a Barnahus could wait. Several social workers mentioned the example of a child who talks about violence at school, and the dilemma of deciding whether the child will go home again the same day. When asked how they reached conclusions about this type of dilemma, several social workers emphasised that if the child expresses some form of fear, or regarding matters of serious or repeated violence, they could not let the child go home or await the child investigative interview at Barnahus.

But the interviews also showed, as illustrated below, how collaborating with Barnahus, as well as the internal organisational interpretations of the legal framework, might affect the immediate protection assessment in a way that would make social workers take a more passive approach in the immediate protection assessment.

E::

It [the question] is really about the children who signal fear. There’s a difference if they’re at school and say that they don’t dare go home. Those are the cases that become urgent for us.

D::

But all the other children who say that they’re beaten—

E::

—but walk home—

D::

—and walk home. I mean, how do you write a protection assessment on that? The child might be beaten the same day again. And before we had Barnahus, usually when we received that kind of report, we went to the school the next day and talked to the child. But now Barnahus and consultation are taken into account a lot, and then you have to wait for the child investigative interview by the police.

Interviewer::

Okay.

D::

So it’s not so—

C::

—so the children wait.

D::

Yes.

Interviewer::

[They] also [wait] on you?

D::

Mm.

C::

Yes, since we have Barnahus.

D::

Yes.

C::

So you wait until they [Barnahus] have time, not until we have time. Because when we talk to the child, we also have to inform the child’s parents.

D::

It’s a full day’s work.

E::

If we go out and talk to the child before, we have to inform the parents. And I’m thinking, after that, the child won’t say to the police that he or she is beaten at home. Because then the child’s been ordered not to speak by Mum and Dad.

Legally, as stated above, always informing parents before talking to a child within a pre-assessment or child welfare investigation is not a formal requirement. Later in the same interview, the social workers described how they felt that the requirement to inform guardians early had become stricter over time in their own organisation, which also created difficulties in talking to children before they went to Barnahus.

D::

When we started working, you were definitely not supposed to inform parents when we were about to do a protection assessment at a school, for instance. But we have new managers now, who say that we have to inform the parents whenever we’re about to go and talk to a child at school for a protection assessment: “We’re about to go out and talk to your child.” Interviewer: Okay.

D::

It’s complete madness. It’s happened more than once where a parent shows up and starts messing around and arguing at school. And how does that turn out for the child?

—Interview 2, intake unit

The interviewees described what once was a normal internal procedure—going out and talking to children at school whenever they disclosed violence—as having become more complicated due to Barnahus, but also based on the internal rules about when the guardians needed to be informed. This situation means that the organisation of the work in the event of suspected violence may result in bureaucratic/administrative obstacles where emergency protective measures become increasingly practical and ethically complicated. The protection of the child “here and now” is pitted against the protection of the child in the long term via the legal process and Barnahus. This scenario may be interpreted as a result of a combination of pooled (internally) and semi-delegated discretion (Hood, 2020) on an inter-organisational level, resulting in delay mechanisms regarding assessments of urgent protection (Molander, 2016).

Previous studies have shown that child welfare services with less specialised work organisation often increase the tendency to investigate children (Östberg, 2010, 2014). Research also indicates that protective measures as compulsory care are rare interventions, since a focus on parental consent tends to guide the assessment (Heimer et al., 2018; Leviner, 2011; Linell, 2017a; Ponnert, 2007, 2019). In a study based on 291 reports processed within 208 investigations of suspected physical violence in a Barnahus region in Sweden, the researchers found that only 4.5% of the reports were assessed to have resulted in a need for immediate protection (of which less than half resulted in placement), and none of the 208 investigations resulted in compulsory care for the child (Quarles van Ufford et al., 2022). In addition, research on Barnahus has shown that children usually go home to their parents after a visit to a Barnahus, even if the parents are suspected of violence, and that usually children are only placed outside their own homes in the more severe cases (Landberg & Svedin, 2013; Kaldal et al., 2010). One possible risk with suspected violence being administered and filtered through several actors and different organisational units could be that stronger administrative and normative thresholds will arise for social workers where immediate protective measures (such as temporary foster care) are regarded as something that can generally wait and be decided upon at a later stage, after the intake unit has passed the case on, or after the forensic interview at Barnahus has taken place.

Protection Through Potentially Risky Integration Practices

In accordance with the findings from previous studies (Kaldal et al., 2010; Landberg & Svedin, 2013), several social workers in the present study also reflected on the fact that children are often allowed to go home to their parents after a visit to a Barnahus. Deciding whether a child should go home or not after the disclosure of violence raises a form of moral uncertainty (Ponnert, 2015), since both taking immediate actions and letting the child go home will raise ethical concerns.

N::

I’m thinking [about] these cases of violence where we go to school. Mostly, the child still goes home. But we also contact the parents. You might get in touch later that day and decide on a time, so the parents can come and talk to us over the next day or so. You try to safeguard [the child]. And what will happen now, when you go home together this evening? How will you talk to each other? I think that’s important for the child, to feel safe and know he or she will return [to the child welfare services] the next day.

Interviewer::

Yes, right.

M::

But it’s sometimes difficult from an ethical perspective, because it’s pretty special to disclose something like that. And then, no matter what the parents say, you will go home.

N::

Mm. Absolutely.

M::

It doesn’t have to be the case that actual violence has happened, but sometimes I think…we let them go home to a large extent. And sometimes it doesn’t result in so much more.

L::

But we might also make a plan with the after-hours social welfare office for back-up—

M::

—Yes.

L::

—and get in touch an hour later or so.

M::

/…/ Then I’m thinking that children, depending on their age, might have their own phones, so we inform them, if something happens, that they can always reach the social services at these numbers and get in touch. [We tell them] that it’s important to inform [people].

/—/

N::

But these situations when a child is at school and is sad and scared and doesn’t want to go home, and you talk to all the people involved, and [the situation] still ends up that way—it’s not a good feeling.

M::

No.

N::

But you can’t deal with it in so many other ways either.

—Interview 6, intake unit

In this interview excerpt, the participants expressed that handling the matter in any other way was sometimes impossible, even if the child was scared and feeling unwell, and they perceived the discretionary space they had been granted as being quite limited. Fear of violence can be an incitement for compulsory care, so the organisational context and collegial norms appear to be the factors that restrict the perceived discretionary space. Instead, the social workers described their use of discretion as focusing on more informal ways to safeguard the child by different integrating practices, such as preparing the after-hours social welfare office or by telling the child to contact the social services if something happens. The interviewees described using similar informal safeguarding practices to compensate for the lack of direct contact with children while waiting at a Barnahus.

Interviewer::

But sometimes you have to make an immediate protection placement; you can’t wait—

P and O::

—Mm.

P::

—to safeguard the child’s protection. But you might also make a police report and then have a contact with the school, since the investigation’s been [formally] initiated. Sometimes you might check with the principal at the school who knows the child. Perhaps [the contact is] the person who reported about violence. This person might keep an extra eye on the child until the police have conducted the child investigative interview and can be a little vigilant to ensure that nothing happens. So, there are some ways to make sure the child isn’t hurt as we wait [for Barnahus].

—Interview 4, intake unit

This discussion may be thought of as including examples of integration practices, in order to fill the gap for children between the report and the child investigative interview at the Barnahus and until a child investigation can move forward. But such practices can also be risky, since the protection of the child is at least temporarily semi-delegated (Hood, 2020) to external actors who lack the legal powers that social workers have to actually be able to control or ensure that the child will not get hurt.

The Accumulated Silence When Passing the Buck

Similarly to previous studies (Johansson, 2011, 2017), the social workers in the present study expressed an adaptation to the criminal justice process and said that they avoided further social investigative measures prior to the child investigative interview at Barnahus, even if a child investigation had been formally initiated. If the time for the child investigative interview at Barnahus dragged on,Footnote 6 the initial process in the child investigation would be delayed and, in reality, compressed in time, since the child welfare services normally have only four months to conduct an investigation (11:2 SSA).

D::

Yes, Barnahus is amazing in these cases, with serious sexual abuse and more serious violence. Then collaboration is good. But in these “light” cases, it only slows us down, because there won’t be any conviction anyway. Because most children can’t say when it [the violence or abuse] happened anyway, so it rarely if ever leads to anything.

—Interview 2, intake unit

In a similar way that the social workers in this study could perceive the internal organisational specialisation into units as a delay mechanism (Molander, 2016), so could co-operation with Barnahus. A document study of 69 children referred to Barnahus has also shown that several of the children had to wait a long time for the interview at the Barnahus, some more than three months (Landberg et al., 2020), indicating that such delay processes are not unusual. Following the fact that the criminal justice logic has priority also means that professionals avoid talking to children before the child investigative interview at Barnahus has taken place, which previous researchers have discussed as well (Johansson, 2011; Kläfverud, 2021). Children are given limited information before the investigative interview when they are taken to Barnahus by a companion (Kläfverud, 2021), and the information about Barnahus they do receive is most often provided by their custodians and not professionals (Kaldal et al., 2010, 2017). Landberg and Svedin (2013) found that approximately eight of 23 local Barnahus in Sweden followed a clear structure and routines for who should inform and follow up with children after a visit at Barnahus, and some lacked procedures altogether. This situation may result in a lack of information and support for children or their families after professionals have conducted the investigative interview at Barnahus (Kaldal et al., 2017). A document analysis of child welfare investigations has also indicated that many children only have brief meetings with the child welfare services (Landberg et al., 2020).

The social workers in the present study also noted that professionals who had a legal obligation to report suspected violence to child welfare services did not talk sufficiently with children who expressed something related to violence.

A::

(…) The school says that when we hear [talk of] “violence,” we shouldn’t talk to the child but should report it to you at the social services. Yes. But they could have asked a bit more. But some are like that—no, if they hear “violence,” they only want to tell the child welfare services, then nothing else. And then we’ll make a police report.

(—)

B::

We can receive one sentence (in a report) saying that a child said today at dinner that she was beaten by her dad. End of story. And if I ask the preschool teacher or educator, “This child has picked you in this situation; why don’t you ask any questions?” “I don’t know,” they’ll say.

—Interview 1, intake unit

Such scenarios can result in social workers feeling that they receive overly vague reports of violence that are difficult to assess. At the same time, a similar process occurs within the child welfare services when social workers at both the intake unit and the investigation unit avoid talking to children or parents before the interview at Barnahus. As previously mentioned, social workers at intake units often expressed that they in fact talked less to children when violence was suspected, since it was then obvious that the case would be passed on to the investigation unit. Some also expressed concerns about the limited number of times in which social workers talked to children and the tendency to avoid talking about violence specifically in child investigations.

B::

There will be a hearing at the police, and before that, we’ve only done a fictitious immediate protection assessment based on the information from the reporter—where we’ll find some protection factors that will result in us not taking the child into care right away. That becomes the leading words here, “compulsory care.”

/…/

Interviewer::

Okay. Would you say that you talk less to children when suspicions of violence are involved than in other pre-assessments?

A.:

Yes.

B::

Yes.

Interviewer::

You do. Okay.

B::

And then [the case is] left to the investigators. And that’s where I think the problem is gigantic, because a child welfare investigation then consists of a period of four months where, in the worst case, you’ve only spoken to the child once, and in the best case, three times. But then these cases are closed without further action in most cases if the parent refuses support interventions. And in my world, if a child’s told a person that someone beats them, they’ve eased their heart, picked out that person usually, and then nothing more happens. So we sometimes get these cases back at the intake unit. And then we can see that in exceptional cases, you [social workers] talk about the violence, but it’s not part of the rule to do so. Instead, [they discuss] everything else around the child—family, what they do, their free time, etc.—which should also be included. But not the specific violence.

/—/

A::

Many times, the role at intake units, we don’t talk to the child about the violence. Since a child welfare investigation is initiated immediately, a social worker [at the investigation unit] will be able to create a relationship and talk about what is difficult during those four months. But I think it’s easier to talk about school and other things.

—Interview 1, intake unit

Previous research has shown how child welfare investigations into violence may be guided by a focus on investigating the violence, or by a more holistic perspective where violence is discussed with less clarity (Landberg et al., 2020; Mattsson, 2017). The tendency to talk about easier subjects than violence, as expressed by the social worker above, could be seen as an example of a more holistic approach. The child assessment framework in SwedenFootnote 7 also provides a holistic framework for child welfare interventions in which violence is not highlighted by a specific predetermined heading in the written investigation (Ponnert, 2017). Several social workers in the present study also expressed the ethical dilemma of encouraging children to talk about violence when their parents could refuse support measures, hence highlighting some of the challenges involved in having a family-service orientation (Gilbert, 2012; Pösö et al., 2014).

The intra- and inter-organisational tensions discussed in this chapter also create the risk for what I call “accumulated silence” from several professionals when children disclose experiences of violence. Due to organisational specialisation and professional collaboration, and the overall adjustment to the criminal logic when violence is concerned, the common outlook is that talking to a child about the violence is best done by somebody else, at some other time and place. We may think of this scenario as a result of what Hood (2020) refers to as “semi-delegated discretion”, better known as “passing the buck”. This approach is problematic from a child’s perspective, since disclosure of violence or sexual abuse is a process (Foster & Hagedorn, 2014; Jensen et al., 2005; Linell, 2017b; Thulin et al., 2020). Research from the perspective of children has also shown that they feel that they lose control after a disclosure of violence and that they experience a lack of information from child welfare caseworkers and Barnahus on what will happen, resulting in fear or anxiety (Thulin et al., 2020). If they do not feel listened to because of various delay mechanisms in professionals’ responses, then the accumulated silence from professionals might also silence the voices of children. The family-service orientation (Gilbert, 2012) and the legal focus on parental consent for interventions in Sweden can also add to this silencing of children.

Conclusion

Since violence is a clear incitement for starting child welfare investigations, either legally and/or normatively, meeting children and their families at the intake unit is sometimes regarded as unnecessary from an integrative perspective, to avoid fragmentation (Axelsson & Bihari Axelsson, 2006). The limited time frame and limited legal authorities within a pre-assessment may also result in a focus on “administrating violence” at intake units, since the case is quickly transferred to the investigation unit. But the analysis has also shown how the organisational specialisation into different units within child welfare—where the immediate protection assessment and pre-assessment is separate from the actual child welfare investigation—in itself can act as what Molander (2016) refers to as division or delay mechanisms in cases of suspected violence, since cases may be distributed during specific weekdays at investigation units.

The inter-organisational collaboration at Barnahus also adds a further layer of division/delay mechanisms, resulting in additional administrative and organisational thresholds for providing immediate protection to children who disclose violence. Together, these mechanisms produce an “accumulated silence” among professionals, where a professional’s response to a child’s disclosure can be delayed or even absent. Social workers’ integration practices to try to safeguard children under these circumstances can also be described as risky, since they might involve people without a mandate to protect the children in practice, or even the children themselves. We may view this situation as a result of a multi-layered juridification process that, in the long run, also creates a serious risk for silencing the voices and further disclosure of children. This risk is probably even more apparent in what Gilbert (2012) refers to as family-service systems, such as in Sweden and the other Nordic countries, where the “least intrusive” principle and the focus on parental consent to interventions also narrows the scope for interventions and social workers’ discretionary space.

The analysis presented in this chapter is based on a limited number of qualitative interviews with social workers in a particular country, Sweden, with a focus on legal tensions in child welfare investigations. How the mechanisms I have identified in this chapter translate to other contexts is a topic for further research; the way in which the collaboration between Barnahus and the child welfare services is affected by the use of different units at child welfare services in most municipalities is also a question that needs to be explored. This chapter however does contribute to an understanding of the complex and intertwined layers of pooled and semi-delegated discretion associated with intra- and inter-organisational collaboration in a highly juridified practice such as violence and child protection, as well as some of its challenges.