Green Lights and Red Flags: The (Im)Possibilities of Contextual Safeguarding Responses to Extra-Familial Harm in the UK

Abstract

Young people experience significant harm in a range of social contexts and from adults and peers unconnected to their caregivers. The recognition of this by policymakers in England, Scotland, and Wales has resulted in child protection policy frameworks increasingly requiring social work responses to the extra-familial contexts where such harm occurs, as well as to the young people affected. This paper presents results from an embedded research project in which five local children’s social care departments used a Contextual Safeguarding framework to respond to this shifting policy direction. The data collected via ethnographic methods over three years included meeting and practice observations (n = 65), meeting participation (n = 334), reviews of young people’s case files (n = 122), interviews (n = 27) and focus groups (n = 33) with professionals, focus groups (n = 6), interviews (n = 2) and surveys (n = 78) with parents and young people, and analysis of local policies and procedures (n = 101). At two stages in the project, the researchers used this dataset to review the progress in each participant site against the Contextual Safeguarding framework. Reporting on the progress made across the five sites, this paper identifies elements of the system change that appeared most feasible or challenging. The results demonstrate four ways in which current policy reforms fall short in creating national contexts that are conducive to the implementation of Contextual Safeguarding, despite local progress towards this goal. The implications for the policy and practices are outlined, with fundamental questions asked of the statutory systems which need to protect, but all too often criminalise, young people abused beyond their front doors.

1. Introduction

When a child is relatively safe with their caregivers but at risk of significant harm from peers or adults unconnected to their families, who should coordinate a response? Since 2018, policymakers in England, Wales, and Scotland have increasingly situated responses with the child and family social workers and the child protection systems in which they work. Concurrently, the policy has been updated to recommend that social workers assess and intervene with peer groups, schools, and public places where extra-familial harm (EFH) occurs and support the young people and families affected. These policy developments have been influenced by the concept of ‘Contextual Safeguarding’ (Firmin 2020), which was introduced to illustrate the gaps in the child protection responses to EFH. While many have welcomed these developments (Brandon et al. 2020; Independent Anti-Slavery Commissioner 2020), the feasibility of Contextual Safeguarding is under-researched. Is national policy recognition of Contextual Safeguarding reflected in the local responses to EFH? To what extent are social workers practising in systems that can, as a Contextual Safeguarding framework (CS Framework hereafter) suggests, assess (and respond to) risks that are faced by, as well as in, families? This paper offers a system insight into these questions. Presenting data from five social care departments that implemented Contextual Safeguarding over three years, we report on where implementation has been most feasible; where the challenges persist; and what this tells us about the sufficiency of child protection policy frameworks that promote Contextual Safeguarding responses to EFH.

2. Background

In addition to the challenges they may face within their families, young people during adolescence may also be harmed in extra-familial contexts. As they enter their first romantic, or intimate, relationships, a significant minority of young people report physical, emotional, or sexual abuse (Barter et al. 2015; Foshee et al. 2014). Some are sexually exploited by adults unconnected to their families (Berelowitz et al. 2013; Dierkhisinga et al. 2020), as well as by peers (Firmin 2017; Radford et al. 2011), and others are exploited into transporting/distributing drugs (Hudek 2018; Turner et al. 2019). The significant, and sometimes fatal, consequences of these forms of harm to young people’s welfare has resulted in their gradual framing as child protection issues by policymakers in England, Wales, and Scotland (HM Government 2018; Scottish Government 2021; Welsh Government 2021).

2.1. Challenges of Child Protection Responses to Extra-Familial Harm

Multiple inquiries and policy reports have identified fundamental limitations in the child protection responses to EFH in the UK (Brandon et al. 2020; Child Safeguarding Practice Review Panel 2020; Firmin 2017; Hanson and Holmes 2014). These limitations are wide-ranging and include the focus of child protection systems on younger children rather than adolescents (Brandon et al. 2020; Hanson and Holmes 2014); on the presumed distinctions between the victims and the perpetrators of abuse, which are often blurred in cases of EFH (Cockbain and Brayley 2012; Turner et al. 2019); on the system responses to the agentic nature of young people affected by EFH, who ‘choose’ to spend time with those who harm them (Lefevre et al. 2017; Lloyd 2019); on the nature of some EFH requiring social workers to coordinate plans to support young people impacted by organised crime (Child Safeguarding Practice Review Panel 2020); and on the fact that EFH often occurs in (extra-familial) contexts that are not the traditional focus of child protection systems (Firmin 2017).

2.2. The Development of Contextual Safeguarding

While there is no silver bullet to resolve these challenges, a range of innovations have attempted to address them (Ofsted 2018; Scott et al. 2017). In this paper, we focus on efforts to resolve the challenge presented by child protection systems focused on individual children and families affected by EFH and not on the contexts where it occurs (Firmin 2017). This mismatch was brought to life through a review of nine cases where young people came to severe or fatal harm within peer, as opposed to familial, relationships (Firmin 2017). In every case, the social workers assessed the capacity of parents/carers to protect their children from extra-familial risks, while the risks escalated in peer groups, schools, and public places; this is a pattern which has since been noted in multiple national publications (Brandon et al. 2020; Child Safeguarding Practice Review Panel 2020; MacAlister 2022).

As a result of the nine-case review, in 2015 the term ‘Contextual Safeguarding’ was coined and converted into a four-part framework (Firmin et al. 2016) to conceptualise the features of the child protection and wider safeguarding systems capable of targeting the extra-familial contexts where EFH occurs. The CS Framework recommended four domains against which to organise system responses to EFH;

Table 1. Domains of Contextual Safeguarding against traditional child and family system.

Domain	A Contextual Safeguarding System	A Child- and Family-Focused System
1	Targets contexts where EFH occurs: identify, assess, and intervene to change the social conditions of those contexts that are associated with EFH	Targets (in)action by parents/carers to build their capacity to protect their child from EFH
2	Uses child protection and wider child welfare legislation, as opposed to community safety and criminal justice frameworks, to organise their responses to extra-familial contexts where EFH occurs	Draws upon community safety and policing legislation to organise responses to contexts beyond families – reserving the use of child protection legislation for family intervention
3	Involves partnerships between those who lead child protection responses and individuals/organisations who have a reach into, or influence over, extra-familial contexts where EFH occurs	Relies on key statutory partners (the police, health, and education) to coordinate plans for children and families affected by EFH
4	Measures the contextual, as well as the individual, impact of safeguarding responses to EFH; in particular whether safety has increased in contexts where EFH has occurred in addition to safeguarding any young people affected	Measures the impact of safeguarding responses on the behaviour of young people and parents/carers

details these domains in comparison to the existing UK child protection systems.

Policy frameworks in England, Wales, and Scotland have been adapted, in response to the development of Contextual Safeguarding, to promote social work responses to extra-familial contexts (peer groups, schools, and neighbourhoods). In 2018, England’s statutory safeguarding guidance was amended to recommend that in cases of EFH:

[Social work] interventions should focus on addressing these wider environmental factors, which are likely to be a threat to the safety and welfare of a number of different children who may or may not be known to local authority children’s social care.

(HM Government 2018, p. 25)

In 2021, revisions to Scotland’s child protection guidance resulted in a recommendation that social work plans:

…seek to create the conditions in which young people can make safer choices rather than simply focusing on changing young people’s behaviour in persistently harmful contexts.

(Scottish Government 2021, p. 79)

In the same year the All Wales Safeguarding Guidance on Child Sexual Exploitation was updated to recommend that:

Interventions, together with support, are required in the places where children socialise and spend time, such as shopping centres and transport hubs.

(Welsh Government 2021, p. 26)

These amendments were presented as additional, something newly required of social workers and the child protection systems in which they operated (Firmin 2020).

Accordingly, the children’s social care departments in England (since 2017), Wales (from 2019), and Scotland (from 2020) have been using the CS Framework to build child protection systems capable of responding to contexts affected by EFH. Policy adaptations to promote contextual child protection approaches to EFH have created an authorising national environment for these local efforts. This paper explores whether such authorisation has created the conditions in which Contextual Safeguarding is feasible.

3. Methodology

The paper presents findings from a three-year, multi-site project in which five children and families social care departments in England and Wales applied the CS Framework. The project ran from May 2018 to June 2022. The participating sites aimed to design, test, and embed a Contextual Safeguarding approach to EFH across children and families services, to assist the research team in understanding what Contextual Safeguarding looked like in practice, its feasibility, and the conditions that facilitated this. Four research objectives were developed in accordance with the four domains of the CS Framework. How, and in what ways (or not), did the sites create system responses to EFH that:

• Targeted the social conditions in which extra-familial harm occurs (Domain 1);
• Used child protection and wider child welfare legislation, as opposed to community safety and criminal justice frameworks, to respond to extra-familial harm (Domain 2);
• Involved individuals and organisations that have a reach into, or influence over, extra-familial contexts (Domain 3);
• Measured the contextual, as well as the individual, impact of any response offered to extra-familial harm (Domain 4).

3.1. Research Process

Thirty local areas responded to a call for interest to participate in the project. Five were selected as research sites based on their readiness to undergo extensive system change and variation in geographical spread; authority type; social work models being used; types of EFH identified locally; and local demographics, such as poverty rates and ethnic diversity. The sites were:

• A unitary city council in the southwest of England covering a small urban area;
• A unitary council in southwest England covering a large rural area;
• A unitary authority council on the south Wales coast that included urban and rural areas;
• A metropolitan borough council in the northwest of England;
• A large county council with 12 district councils in the southeast of England covering rural, coastal, and urban areas.

A project team of eight researchers across two universities worked alongside the sites. They adopted a methodology informed by an embedded approach to the research, working collaboratively with the sites to co-create shared project plans and outputs (McGinity and Salokangas 2014; Lloyd 2021). The research was conducted over three phases:

• Phase one: understand the site’s current response to EFH and develop a plan to adapt that response against the CS Framework;
• Phase two: run two pilots at each site that build responses to extra-familial contexts, and assess their alignment with the CS Framework;
• Phase three: embed pilot results and disseminate learning.

Across phases one and two, the research team used a range of methods to capture the extent to which the sites were working towards the four domains of the CS Framework. These methods included:

• Observations of 65 meetings and the practice related to cases of EFH;
• File review of 122 cases and assessments;
• System mapping (n = 10): to understand the journey of children and contexts through the system;
• Participation in 334 meetings to inform developing work;
• 27 Interviews and 33 focus groups with practitioners;
• 6 focus groups, 2 interviews, and 78 surveys with parents and young people to capture their experiences and perspectives on the new approaches to EFH;
• Documentary review of 101 new and existing policies and documents relating to EFH.

3.2. Ethics and Limitations

The ethics for the research was granted by two universities. The consent to participate was secured throughout. The initial consent was agreed by the Director of Children’s Services and a multi-agency partnership. The consent for individual participation was gained individually. The consent considered the impact of the research, the confidentiality, and the right to withdraw.

Several limitations are relevant to the study. Firstly, COVID-19 significantly impacted the project. Prior to this, the researchers had been physically present at the sites; this was not possible from March 2020, impacting the development of the relationships with the practitioners and the provision of informal support and reflection. Moreover, COVID-19 put restraints on the involvement of the children’s service practitioners in creating and innovating system change, as they were responding to the immediate needs of young people during the pandemic. Secondly, the project did not aim to develop a single approach to responding to contexts associated with EFH. The systems created in these five sites are not necessarily feasible in other UK or international locations, which may operate in different political, social, economic, and cultural contexts. The study sought to understand the possibilities and constraints of Contextual Safeguarding in the current policy landscape in five exemplar sites.

3.3. Analysis

An iterative approach to the analysis was undertaken, punctuated by formal analysis sessions. Throughout the project, the researchers documented thoughts and reflections which were discussed during fortnightly reflective team meetings. Emerging findings were considered monthly with the sites and through governance arrangements. A systematic approach to analysis was then undertaken at several points throughout the project. A System Review framework (Firmin 2021) was used, whereby the research team considered the data collected against each domain of the CS Framework and for different parts of a site’s local child protection and wider safeguarding system:

-

‘referral’: activities to (a) support referrals of contexts associated with EFH and (b) record contextual information when young people affected by EFH were referred for support;

-

‘assessment’: activities to (a) assess contexts associated with EFH and (b) gather contextual information when assessing the needs of young people affected by EFH;

-

‘planning’: activities to agree plans (a) for contexts associated with EFH and (b) that support young people by identifying and addressing contexts where they are unsafe;

-

‘response’: activities that (a) respond to changes in the social conditions of the contexts associated with EFH and (b) respond to young people affected by EFH in ways that recognise the impact of contextual factors on their decision making and behaviour.

The data were organised against the framework to be illustrative of approaches that were ‘green’ (appropriate or emerging), ‘amber’ (some elements but not yet integrated), and ‘red’ (no evidence or limited evidence) practices as they aligned to the CS Framework (see

Table 2. example system review framework.

	Referral	Assessment	Planning	Response
Domain 1: Target	Red - Case reviews highlight that assessments of young people do not consistently include information on contexts Amber - Lack of clarity of threshold for cases of EFH - No formal established mechanism for logging contexts in at the front door Green - Front doors have systems in place for referring locations and peer groups - Referrals are being utilised for level 2 by multi-agency partners	Red - Individual assessments do not routinely ask questions about broader contexts Amber - Variation in thresholds for location assessments Green - Have developed a pathway for assessing locations and peer groups	Red - EFH meetings are not seen as priority so less attendance at meetings and less time for assessment Amber - While planning meetings are able to articulate context most in need, this is not necessarily evidenced in the target of interventions Green - Panel process builds relationships with partners and reduces individual practitioner anxiety	Red - lack of knowledge or understanding of potential interventions Amber - Evidence of some peer group intervention work but does not fully consider the peer group dynamics or address them Green - Use of partnerships facilitates responses in the places and spaces they spend their time with the people they are friends with
Domain 2: Legislative framework
Domain 3: Partnerships
Domain 4: Outcomes

below). Each field of the table was populated with evidence. Table 2 provides a pared-down example of this. Once the data were organised, the researchers used the CS System Review framework to decide on the overall colour of each unit. In the example in Table 2, the unit for Domain 1, at the point of referrals into the system, would be scored as ‘amber’ because while there is evidence of emerging green practice, many elements of the system at this practice point are not yet developed or integrated. The project co-investigator was present at every rating session to ensure consistency in the application of the framework.

The findings were presented to the site representative and then at the ‘system review’ meetings for input from the site participants. This RAG rating process was undertaken at two points in the project: at the end of phase one in 2019 and at the end of phase two in 2021. The researchers produced two colour-coded tables for each site to illustrate progress in implementing Contextual Safeguarding from 2019 to 2021 (see

Table 3. Findings of stage one and two RAG rating.

			Site U				Site V				Site X				Site Y				Site Z
			Part of System				Part of System				Part of System				Part of System				Part of System
			Ref	As	Pla	Re	Ref	As	Plan	Re	Ref	As	Pla	Re	Ref	As	Pla	Re	Ref	As	Pla	Re
Stage one	Domain	1	0	1	1	1	1	1	0	1	0	1	1	1	0	0	0	1	0	0	0	0
		2	0	1	0	X	1	1	1	0	0	0	1	1	0	1	0	1	X	0	0	0
		3	0	0	1	1	1	1	1	1	0	0	0	1	1	0	0	1	0	X	1	0
		4	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Stage two	Domain	1	1	1	2	1	1	2	1	1	2	2	1	1	2	1	1	1	2	2	1	1
		2	2	2	2	2	2	2	2	2	2	2	2	2	2	1	0	1	2	2	2	2
		3	1	1	2	2	2	2	1	2	2	2	2	2	1	1	1	1	1	1	1	1
		4	1	1	1	1	1	1	1	1	2	2	1	1	1	1	1	1	1	1	1	1

Note: X indicates insufficient data to provide a score.

below).

For this paper, all ten colour-coded tables were analysed. The colours were converted into numbers (red = 0, amber = 1, green = 2) to calculate the sites’ total points for implementation of each domain at phase one and two and the differences in each site’s performance between phase one and two for each domain and to provide an implementation rank for each domain from the highest to the lowest score.

4. Findings

Table 3 shows the RAG rating results after stage one and stage two for all five sites (referred to as U, V, X, Y, and Z for anonymity). The findings at these two stages demonstrate site-level developments across the project. The four domains of the CS Framework run vertically, and the part of the site’s system under analysis runs horizontally (Ref = referral, As = assessment, Plan = planning, and Re = Response).

By drawing together the total number of points scored across the sites for each domain of the CS Framework, we illustrate the progress the sites made between stages one and two of the project (

Table 4. Total points and difference across stage one and two.

	Total Points		Difference
	Stage One	Stage Two
Domain 1	10	27	17
Domain 2	8	36	28
Domain 3	10	29	19
Domain 4	0	22	22

).

At stage one, the sites scored highest for alignment with Domains 2 and 3 (10), with Domain 4 scoring the lowest. At stage two the sites scored highest for Domain 2 (36), with Domain 4 still scoring the lowest (22). The sites made the greatest improvement in developing systems that aligned to Domain 2 of Contextual Safeguarding and the least in their system’s ability to align to Domains 1 and 3. These results are examined below; they are contextualised with reference to the dataset used throughout the RAG rating process.

4.1. Most Progress: Domain 2

Domain 2 of the CS Framework recommends that the responses to EFH be built upon child protection and wider safeguarding and legislation/guidance, as opposed to community safety or wider criminal justice policy. The local systems aligned to this domain would likely feature social work responses to young people affected by EFH even when the parents/carers were assessed as protective (so as to respond to extra-familial factors), as well as the contexts in which EFH occurred (rather than police-led responses to groups or locations).

4.2. Alignment to Domain 2 at Stage One

At stage one, the sites scored a total of eight points for Domain 2. Every site was assessed as red for Domain 2 in at least one part of their system. Two sites (U and Y) were unable to demonstrate evidence of a child protection and wider child welfare approach to EFH at the point where young people, and the associated contexts, were referred for support. The two sites that were making progress towards a child welfare approach to referrals for EFH (X and Z) were unable to provide evidence of this through the assessments. In three of the five sites, there was no evidence that the plans to respond to EFH were child-welfare-led (as opposed to criminal-justice-led). Site Z scored red in every part of their system for Domain 2, except for referrals where the data were insufficient to reach a conclusion. No site scored green for Domain 2 in any part of their system.

The reviews of the social work case files at stage one evidenced how some sites only drew upon safeguarding frameworks when the EFH was assessed as being attributable to the (in)action of the parents and therefore the home context. Without such attribution, young people affected by EFH were closed to children’s social care:

B is still at risk of CSE (child sexual exploitation) and as such she is receiving support at school from [VCS organisation] on a weekly basis. B’s mother has a safety plan in place and supervises B. There is no role for the Children ‘s Social Work Team at this moment.

(Dip-Sampling Data, Site Z, 2019)

I had a positive feeling about this family, despite the incidents leading to this referral. N (brother) has agreed to referrals to PREVENT and Skills for Work... I see no further role for Children’s Service.

(Dip-Sampling Data, Site U, 2019)

The role of the social work and the child protection processes were largely reserved for intervention with the home, as opposed to the extra-familial, contexts.

The responses to EFH that went beyond parenting intervention were heavily reliant on community safety and policing responses (counter to Domain 2). This reliance was acutely evident when the sites responded to extra-familial contexts. The tasking of interventions into extra-familial contexts was managed through panels that themselves were heavily influenced by crime data. The role of social workers was limited to assessing the parents/carers of young people at risk of harm in those contexts – not the contexts themselves. At Site V, for example, an outreach team had been commissioned to support young people who were being criminally exploited or were at risk of serious violence and to work in the places and spaces where this harm was occurring. This team was based within a community safety directorate and had no social work involvement. The young people were identified by an algorithm which drew heavily from police data. The social worker responses to the young people concerned were viewed as being independent of that team and were parent-focused.

As such, at the stage one sites the responses to EFH were not aligned to Domain 2 of Contextual Safeguarding. The child protection systems and the social workers within them:

• Assessed, and developed plans to address, parenting/home contexts of young people at risk of EFH;
• Had an undecided, contested, or wholly absent role in responding to extra-familial contexts; a responsibility held by the criminal justice and community safety professionals.

4.3. Alignment to Domain 2 at Stage Two

By stage two, most sites were working towards or had resolved both of the aforementioned challenges. Various activities were undertaken to achieve this progress. We highlight three:

• Revision of the social care pathway for young people at risk of EFH;
• The increasing of the social work involvement in the assessment of extra-familial contexts;
• The building of a welfare-based understanding of EFH across a workforce.

The first two of these were consistent across multiple sites; the third demonstrates the contrasting progress made against Domain 2 in Sites Y and Z.

All the sites revised their social care pathway (and associated planning processes) for young people who had experienced EFH. They used child protection and wider safeguarding guidance to frame responses to EFH and to put in place social-work-led planning and oversight for the young people (and contexts) affected. For example, one site (code not provided to protect identity) piloted a ‘Risk Outside of the Home’ (ROTH) pathway and an associated conference. This pathway was to be used to coordinate plans for young people where the principal source of significant harm was extra-familial. The social workers were responsible for completing assessments and authoring reports ahead of a ROTH conference and a meeting, chaired by a child protection chair, to agree a plan. In these meetings, the social workers and partner organisations focused on the context(s) in which young people were at risk of harm, as opposed to the (in)action of parents. Parents/carers attended the meetings as safeguarding partners.

In one ROTH conference, the professionals were observed discussing the welfare of a young person at his college. He had been attacked on the college premises and was no longer attending due to safety concerns. The social worker leading the assessment stated that actions were required to build safety at the college itself, and the college leadership were asked to account for any work undertaken to date or planned in the interim; a shared safety plan was also devised for this young person and two of his peers; the plan was co-owned by them and their parents to identify safe ways (and places) for them to spend time together until the wider community-based risks had been addressed. The plan therefore was largely focused on his peer group and his college—as opposed to creating safety in an already protective family home.

Not all the sites introduced such extensive changes. Some made small amendments to existing systems or clarified the role of social workers in these cases. For example, Site Z reviewed and refined a pathway for supporting adolescents at risk of harm, auditing its use to ensure access to support due to the EFH not being reliant on the identified risks within the home context.

Sites V and X piloted social-work-led assessments of extra-familial contexts, including peer groups and public spaces, such as streets, beaches, and parks, and all the sites attempted to assess the welfare of the young people in extra-familial contexts during the test period. These processes meant that if young people were believed to be at risk of significant harm in an extra-familial context, a social-work-led assessment of that context was undertaken. Site V piloted five assessments and Site X piloted ten. The prior context responses had been the remit of community safety partnerships. The participants noted the distinct perspective a social work assessment brought to how the risks in extra-familial contexts were understood:

We couldn’t have done it without social care and having that, this piece of work ongoing, I don’t think, I don’t think it would have happened without the pilot running … to get people involved that I hadn’t really thought of. I don’t think I’d have thought about involving housing, if I’m honest.

(Site V, FG10)

Across all the assessments, the young people’s welfare, rather than crime, was a central concern. This shift in focus informed the development of child-welfare-orientated planning for contexts. For example, following a serious disturbance involving serious violence between young people on the seafront, the beach was referred to children’s social care for assessment. The area had experienced the closure of over-18 establishments due to COVID-19 restrictions, which reduced the business oversight of the area and increased the number of adults using the beach and mixing with young people. Social workers in Site X led an assessment of the location to increase young people’s safety at the beachfront. Social workers used this assessment to coordinate a multi-agency response to the beach involving activities, including increasing detached youth work provision, social media engagement, and increased guardianship via bed and breakfasts and cafes to help keep young people safe.

As with the efforts to develop social work pathways for young people at risk of EFH, introducing social work oversight in response to extra-familial contexts increased the site alignment with Domain 2. Such activities were built upon child protection legislation and wider safeguarding guidance, increasing the extent to which the sites took a child welfare approach to EFH, while simultaneously decreasing community-safety-framed interventions.

By reporting the progress the sites made in aligning to Domain 2 via structural or practical changes, we risk overlooking the culturally focused system change. Comparing the progress of Sites Y and Z is helpful here. Site Z made the most progress on Domain 2 between stages one and two, whereas Site Y made the least. At stage one, Site Z scored red; a child welfare response to EFH was absent. The stage one system review raised significant questions about the culture of the service. The data collected demonstrated that the practitioners were struggling to recognise young people/contexts affected by EFH as being in need of a social work response. Consequently, Site Z focused on training 391 social care professionals, 91 professionals from statutory partner agencies, and 7 from a fast-food outlet where young people were spending time. These efforts offered consistent communication that EFH was a child welfare, and not solely a community safety, issue and provided the foundations upon which to introduce the type of structural or practical changes outlined previously.

Comparatively, and as will be explored in later sections of this paper, Site Y did not build a collective understanding amongst its workforce that EFH was a child welfare issue. By stage two, there remained multiple examples of how social workers and their partner agencies framed responses via policing and community safety policy and practice. This meant that even when they introduced some contextual practices into their assessment and planning processes or attempted to undertake welfare-based assessments of extra-familial contexts, the social work plans remained predominantly concerned with parenting. For example, while the risks to one young person appeared to be community-based, the statement read at the end of their child protection conference stated:

… we want to live in home where you don’t witness fighting in family or community, we want you to feel supported and have consistent parent boundaries and to engage in support and interventions available so mum can live in a safe a supported community and we want him to be better able to manage his behaviour in school and in community.

(Site Y, Observation 48)

4.4. Least Progress: Domains 1 and 3

Domain 1 of the CS Framework recommends that local responses to EFH target the contexts where such harm occurs—by assessing and intervening in ways that change the social conditions of those contexts. In short, it is about identifying and addressing the social norms conducive to EFH, not simply changing people’s behaviour in, or the design of, those contexts. Domain 3 of the CS Framework recommends that any efforts to implement Domain 1 occur through partnerships between those who traditionally lead child protection responses to harm (children’s social care) and the individuals/organisations who can influence extra-familial contexts. These partners may include people who work in libraries, shopping centres, transport hubs, youth clubs, schools, etc., as well as the young people and families in contexts where EFH has occurred.

4.5. Alignment to Domains 1 and 3 at Stage One

In 2019, the sites scored a total of ten points for their alignment with both Domain 1 and Domain 3 of the CS Framework, making them the highest-scoring domains at stage one. For Domain 1, four of the five sites were assessed as demonstrating amber practice in at least one part of their system, and at three sites (U, V and X), the majority of the system was assessed as amber. For Domain 3, all the sites could demonstrate amber practice in at least one part of their system.

Perhaps it is unsurprising that at stage one the sites scored highest for Domain 1. While not a correct interpretation, the researchers in the Contextual Safeguarding team have found that practitioners often associate Domain 1 with social care interventions in places outside the family, such as when intervening to change the behaviour of individuals in extra-familial contexts rather than the social conditions of those contexts. As such, the data from stage one illustrated multiple activities with ‘amber’ alignment to Domain 1, such as:

[Early help service] have been doing peer-group eco mapping e.g., to inform working with a peer group following a suicide. Another related to a location.

(Amber, Stage one, Domain 1, Site X)

Have analytical capacity to peer group map and location map.

(Amber, Stage One, Domain 1, Site U)

As the test site practitioners developed responses to locations outside the home (Domain 1), they needed to do so with the agreement and cooperation of their partners in those locations (Domain 3). Consequently, the fact that the sites received the same score for alignment with Domain 3 as they did with Domain 1 was somewhat expected. England’s Disrupting Exploitation Toolkit (Home Office 2019) and the Welsh CSE guidance (Welsh Government 2021) both recommend that agencies intervene in locations associated with EFH in collaboration with multi-agency partners. At stage one, we saw multiple examples of this:

Youth Community Safety group meeting has positive welfare outputs—disrupting gangs work in parks etc.

(Amber, stage one, domain three, Site Z)

[Safeguarding Service] is developing a service with schools.

(Green, stage one, domain three, Site Y)

[Youth Services] in development provide group interventions.

(Green, stage one, domain three, Site U)

At stage one, Domains 1 and 3 of the CS Framework appeared to be interdependent. Emboldened by national policy frameworks that promoted location-based interventions and local senior commitment to Contextual Safeguarding, the sites developed approaches to working in extra-familial contexts with multi-agency partners as the first steps in creating their system change.

4.6. Alignment to Domains 1 and 3 at Stage 2

By stage two, the sites had made the least progress with respect of Domains 1 and 3. While no sites scored red for Domain 1 at stage two, they also failed to progress to green in more than two parts of the system.

To be considered ‘green’ for Domain 1, the sites had to do more than undertake safeguarding work in contexts outside the home (enough to potentially score amber); this work needed to tackle the social conditions of abuse. To align with Domain 3, the sites had to move beyond work with statutory partners deployed into extra-familial contexts (such as policing, schools, and health services) and instead engage non-traditional partners based in the contexts in which the young people spent time (such as with young people, families, businesses, and residents). The ability to create systems that aligned to Domains 1 and 3, or the challenges of doing so, therefore remained interdependent.

The implications of this interdependence are exemplified through the data collected from Sites U and Y, which piloted responses to extra-familial contexts between stage one and two. Site U’s response was a community meeting for schools to raise concerns about EFH, and Site Y assessed a neighbourhood and peer group associated with EFH.

Site U introduced a ‘community meeting’ process for schools to raise trends related to EFH. These community meetings brought together schools with the local authority adolescent team and other local agencies. In one meeting, a school raised concerns about a group of young people coming to the attention of the local police (present in the meeting) and where there was “a lot of bad feeling [directed at] these young people in the community” (school staff, meeting observation 44). In the meeting, different partners discussed the concerns that had arisen (such as criminal damage and being ‘abusive to the public’) and attributed them to a lack of activities for young people. In the meeting, the partners—the police, the youth workers, and the school—discussed how to pool funding to resource a few days away for the group to build relationships and address some of the challenges arising in the community. The meeting concluded with each organisation seeking funding and the youth workers and police agreeing to watch out for the welfare of the group when not at school.

Site Y undertook a context assessment in response to concerns that a group of young people were being criminally exploited. This involved an assessment of an estate, the surrounding area, and a group of young people associated with the concerns. The area was identified by the crime data received by the Community Safety Partnership in relation to, and as the product of, a police operation. The aim was to undertake assessments to reduce the perpetration of criminal exploitation in this area (by adults and young people) and increase the safeguarding of the young people. The information supplied by the young people to the safeguarding practitioners who were completing a peer assessment was triaged through a police website, and the lead social worker overseeing the assessment noted that the assessments “may lead to more arrests of young people” (notes, observation 46). The assessments supported the identification of guardianship in the area (youth workers) and highlighted options for other interventions to disrupt the crime, such as increased police patrols, CCTV, and licensing incentives (issuing notices to businesses where young people may be being harmed to disrupt activity). The practitioners noted that they were unable to engage young people in discussions about their peers but felt the peer assessment had helped them to understand the group more.

The examples from Sites U and Y illustrate two very different approaches to addressing EFH. While both pilots engaged similar partners, there were clear differences in the focus of the assessments. Site U’s pilot evidenced a commitment to ensuring the welfare of the young people; to understanding the context in which concerning behaviours had occurred; and to identifying activities/opportunities for the young people—thus changing the social conditions of the context (Domain 1). While the police were involved, this was to explore their guardianship role in looking out for the young people who may be vulnerable. At Site Y, assessments were led by Community Safety, focused on mapping crime data and considered opportunities to prevent crime via disruption (for example, by arresting adult perpetrators) to protect young people who may be exploited. This response reinforced the social conditions of the contexts where policing enforcement was dominant.

To align to the CS Framework, it is not enough for safeguarding professionals to work in contexts beyond the home with traditional safeguarding partners. Where the focus of the work is still individualised, it is likely that interventions will focus on changing the behaviours of individuals, not the contexts where the behaviour occurs (as Domain 1 requires). This, in turn, highlights the significant influence of the partnerships promoted via Domain 3 of the CS Framework. While neither site particularly engaged with non-traditional partners or young people and families (and as such were mainly marked as amber for Domain 3 at stage two), the task of the partners also influenced the approach taken to targeting the contexts/groups associated with EFH. For example, the police were involved at both sites. At Site Y, they were engaged in assessments focused on crime reduction, where arrests of young people were seen as a possible impact of the work—reinforcing existing relationships. At Site U the intervention plan included activities to change relationships between the police and the local young people.

At its most extreme, a site’s performance at the intersection of Domains 1 and 3 can influence whether they align with Domain 2—the extent to which the work is framed as a child protection and wider safeguarding response. This was evident when Site Y still scored red for Domain 2 in their system’s approach to planning at stage two. It is worth noting the contextual differences between Sites U and Y. At Site U, the practitioners were concerned about, among other things, a peer group ‘pulling up flowers’ in the high street (notes observation 44). At Site Y, the peer group were being exploited to traffic class A drugs and had experienced multiple stabbings. In this context, Site Y struggled to shift the dominant narrative of crime prevention in favour of welfare when planning in cases of EFH (Owens and Lloyd Forthcoming).

5. Discussion

The efforts of the five sites featured in this paper demonstrate that an updated policy framework does not a Contextual Safeguarding system make. While most sites managed to increase the social work oversight of both the contexts where EFH occurred and the young people affected (and therefore align to Domain 2 of the framework), this did not always result in child welfare responses that (a) targeted the social conditions of those contexts (Domain 1) or (b) featured partnerships with individuals and organisations who could influence those contexts (Domain 3). Four themes emerged across the findings detailed above which offer some explanation of these shortcomings.

Firstly, social care systems in England and Wales (as with many other countries in the Global North) are designed upon behaviourist principles (Gilbert et al. 2011). Domain 1 of the CS Framework requires local services to address the social conditions in which harm occurs—not just target behaviour-change interventions at individuals who comprise a peer group or spend time in a public place. The individualised nature of the UK’s child protection systems has been well documented (Featherstone et al. 2018; Parton 2014; Rogowski 2012). Individualised systems ask people to behave differently in persistently unsafe and unequal conditions; their behaviour (and not the context in which it occurs) is the target of the system. Through recognition of ‘extra-familial’ interpersonal harms, the researchers behind Contextual Safeguarding highlighted the limitations of that ask. Yet, expanding social work intervention to peer groups and locations has not necessarily shifted the focus of the social work interventions offered within individualised systems. The challenges the sites experienced in moving from amber to green in Domain 1 of the CS Framework were largely associated with their persistent focus on individuals who spent time in extra-familial contexts rather than the wider social conditions which informed their behaviour.

Secondly, child protection systems in England and Wales largely target the (in)action of parents in order to create safety. Parents, and the young people they support, are the target of the system—not a partner to it (Featherstone et al. 2018). While many areas have sought to redress this balance, introducing models such as Family Group Conferencing to create spaces where families can be partners in creating safety for children (Mason et al. 2017), these efforts do not characterise the systems more widely. Although the sites made some progress in pushing against this by attempting to target the contexts associated with EFH, the legacy of parent-focused systems meant that most sites struggled to establish collaborative social work relationships (and associated cultures) with non-statutory organisations and individuals in extra-familial contexts. The two sites that performed best in this respect (V and X) had long-standing relationships with community organisations that they were able to leverage in response to extra-familial contexts. In the remaining three sites, the partnerships with young people, families, and their wider communities were not in the foreground of the responses to EFH, nor were they a consistent feature of their wider safeguarding system, and this undermined the alignment with Domain 3 of the CS Framework.

Thirdly, the national policy frameworks for EFH have long straddled criminal justice and children’s social care departments (Cockbain and Brayley 2012; MacAlister 2022); the same patterns are borne out in local services. The jury is still out, it seems, as to whether EFH is primarily a criminal justice or a child protection matter, particularly when responding to serious violence. This tension was particularly apparent at Site Y from our dataset, but one that also challenged most sites. A clear steer from social care leadership in Sites U, V and X and a focus on shifting to a welfare-based culture at Site Z supported the four sites in aligning well with Domain 2 despite these tensions. However, all the sites are also operating in national policy contexts where community safety partnerships hold oversight of some forms of EFH – such as ‘serious youth violence’ (Firmin and Knowles 2022). In this sense, they have been authorised to develop social work responses to the contexts where EFH occurs in one set of policy documents, whilst another situates EFH as a community safety issue. No site appeared to fully resolve this by the end of the test period.

These three challenges within national systems/policy frameworks coalesced to produce a fourth and final challenge encountered by the sites—the desired outcomes for service responses to EFH and their (mis)alignment with the CS Framework. Most sites struggled to make any progress in consistently measuring the contextual impact of their responses to EFH (Domain 4). We have not reported progress against Domain 4 in this paper as this was the most undeveloped element of local responses at the start of the project and remained that way at the point of writing. All the test sites are operating in policy and practice environments that measure impact individually. How are social workers to assess and address risks faced by families when the systems they work in and the interventions they utilize are designed to measure change in the behaviour within those families instead of the contexts in which they are raising their children? Statutory commissioners and voluntary sector funding bodies resource interventions aligned to individual outcome measures, and government departments evaluate a range of responses on the same grounds (Preston et al. 2021). This focus is baked into so many structures in which these five research sites operate that it is interesting that they made any progress aligning to a CS Framework. However, as this project demonstrated, ensuring social work responses to EFH (Domain 2) does not mean that responses will seek change beyond the behaviour of individuals (Domain 1).

As well as highlighting where the sites were challenged, the four thematic barriers outlined above suggest the conditions required for Contextual Safeguarding to flourish, and these conditions go beyond the policy reforms outlined at the outset of this paper. It seems that for all four domains of the CS Framework to be implemented the sites need to be operating in national and local conditions that recognise and resource responses that are not underpinned by behaviourist principles; that (re)situate social work within communities, enabling relationships with non-statutory and community partners to be a common feature of the practice; that position criminal justice responses to EFH as secondary to (or subsumed under) safeguarding responses in both culture and practice rather than something delivered in parallel (and potentially undermining the safeguarding outcomes); and that routinely incorporate contextual outcome measures into commissioning frameworks and service reporting structures.

The variable absence/presence of these dynamics in local sites, and their general absence in national policy frameworks, is illustrated by the points of consistency/divergence across the sites and raises questions about the feasibility of Contextual Safeguarding. At the very least, it is critical to question the extent to which Contextual Safeguarding will be implemented as intended within national policy conditions or whether local efforts to do so may risk the extension of individualised practices experienced by families to extra-familial contexts and the young people within them. Areas that wish to implement the CS Framework will need to consider how to mitigate these four challenges in local structures and cultures and in doing so provide conditions more favourable for implementation. For those involved in researching Contextual Safeguarding, this is an important point of learning. Phase 1 of this project involved mapping the structures that the sites had in place to respond to contexts associated to EFH—and the extent to which those structures aligned to the CS Framework; however, this mapping did not look for whether local conditions addressed the four thematic challenges discussed in this paper. While the challenges can, in some respects, be mapped onto the four domains of the CS Framework they are distinct from them; they are policy and cultural facilitators of implementation.

6. Conclusions

This paper explored the feasibility of Contextual Safeguarding responses to EFH in England and Wales by reviewing the implementation process at five test locations. The results suggest that some elements of the CS Framework have been more feasible than others and that progress across all the sites has been informed by four national shortfalls in creating the conditions conducive to Contextual Safeguarding.

Given these shortfalls, we can conclude that recognition of the CS Framework in national policies has not been sufficient to create conditions for the approach to be adopted in local systems. In particular, wider system characteristics that focus on behaviour, risk reduction, and individual outcome measures and are designed to ‘do-to’ children and families, rather than to ‘work with’ them, all require attention for national policy frameworks to facilitate the local implementation of Contextual Safeguarding. Furthermore, recommending social work responses to contexts and young people affected by EFH does not necessarily undo contrary policy frameworks that permit sanction-based community safety or criminal justice responses to the same young people and contexts. National policy frameworks currently promote both an increase in a child welfare response to EFH and the maintenance of a community safety response to EFH—without fully acknowledging that the latter may undermine the former. Any site wishing to adopt a Contextual Safeguarding approach will need to consider how they work within, and mitigate, these system constraints to stand the best chance of successful implementation.

The challenges identified also have implications that go beyond the implementation of Contextual Safeguarding. They create a situation in which contextual responses to EFH may do more harm than good. A fair criticism of Contextual Safeguarding is that it has the potential to extend the reach of state intervention into young people’s lives, particularly into their friendships and communities (Wroe and Lloyd 2020). Such a critique is particularly pertinent if intervention is characterised by a focus on behaviour and crime reduction. While these features would not actually align to the CS Framework (Domains 1 and 3), the data presented in this paper illustrate that they could be a feature of implementation efforts and would potentially align to wider national recommendations to take a contextual approach to EFH (exclusive of alignment with the CS Framework).

Despite these challenges, most test sites developed a child protection response to contexts (and young people) associated with EFH. This progress suggests a desire for social work leadership/coordination in response to EFH and some alignment between this direction of travel and social work skills/values. Indeed, efforts to intervene at the point where structure and agency meet are well aligned to the ethics of social work promoted by the International Federation of Social Work (IFSW 2014). To further test the feasibility of Contextual Safeguarding, therefore, it will be critical to better understand and communicate the conditions most conducive to the implementation of the CS Framework, both in other UK settings (with test work underway in Scotland) and international settings (with international test work set to commence in late 2022), where contextual (and community) interventions are more prevalent, or criminal justice responses less dominant. In the intervening period, it appears that social work responses in England and Wales that address risks faced by families, as well as in them, are being developed despite, rather than due to, the wider policy frameworks in which they operate.