The Views of Local Authorities in England on How to Prevent Children Being in Care

ABSTRACT

Purpose

The rates of children looked after by local authorities in England have been rising for more than two decades. This study was conducted to determine what approaches local authorities have adopted that they perceive to be the most effective in preventing the need for children to come into care. It also considers how they evaluate these approaches and how they assess cost-effectiveness.

Methods

An online survey was distributed to leaders of children’s social services departments in England in 2018 (n = 152).

Findings

Sixty (39.5%) local authorities completed the survey. Respondents were asked to select up to three types of services or approaches they deemed most effective in preventing the need for children to come into care. The most popular was a whole-system approach selected by 81.7%, with Signs of Safety most commonly cited. This was followed by edge-of-care services (61.7%), early help (56.7%), family group conferences (43.3%), parenting programmes (18.3%), short break services (15.0%) and “other” services (20.0%). Local authorities who had experienced increases in the numbers of children in care were more likely to discuss approaches introduced relatively recently. Whole-system approaches and parenting programmes were the approaches most likely to have had independent evaluations. Whilst most local authorities reported the use of economic analysis methods as part of their evaluation, there was insufficient detail for a full assessment of cost-effectiveness.

Originality

This paper provides a description of contemporary attitudes amongst leaders of children’s services to approaches that aim to keep children out of care. It also describes approaches taken by local authorities to evaluation and assessing cost-effectiveness.

KEYWORDS:

• Social services
• whole-systems approach
• edge of care
• early help
• family group conferences
• Signs of Safety

Introduction

The rates of children in out-of-home care have been steadily rising for some time in much of the UK. For example, the number of looked after children in England has risen 67%, from 1994 (n = 47,950) to 2019 (n = 80,080) (Thomas, 2018; Department for Education, 2019) and it is England that is the focus on this paper. Rates are also rising in Wales and Northern Ireland but in Scotland, they are falling from a peak in 2012 (National Society for the Prevention of Cruelty to Children (NSPCC), 2021). Internationally, trends vary, with some countries also experiencing rising rates, for example, parts of Canada and Australia (O’Donnell et al., 2016) and others experiencing a fall over time (e.g. Denmark—see Ubbesen et al., 2015).

A child who is accommodated by the local authority in England and in their care is termed “looked after” in law. The colloquial term “in care” is also still used to denote this status, so both terms appear in this paper. The large majority of looked after children live away from their parents, but it is also possible to have the legal status of “looked after” but in fact be placed with your parents. This would be as part of a reunification plan after a period of time in out-of-home care. It should be acknowledged that there is important international variation in social understandings of child protection (Hearn et al., 2004) and the systems that have been developed for out-of-home care (Ainsworth & Thoburn, 2014). For a court to issue an order that a child be looked after by the State in England, they have to be suffering or at risk of suffering significant harm, but of course what counts as significant harm will change with shifts in the social construction of social problems—for example, intimate partner violence and child sexual exploitation increasingly being regarded as forms of child abuse that warrant state intervention (Thomas, 2018).

The Care Crisis Review (Thomas, 2018) suggested there were many overlapping factors contributing to the increasing numbers of children in care over time in England and Wales. Some of these relate to changes in society including socio-economic factors, the nature of child and family circumstances and changing conceptions of risk, but some of these also relate to the social care system and associated practice, including the legal and policy frameworks, professional practice, and tensions in the system.

There are, undoubtedly, circumstances where entry into care is in the best interest of the child. In such circumstances, the Public Law Outline sets out the process that needs to be followed in England and Wales (see Leeds City Council, 2017). Although outcomes for looked after children are poor compared to the general population—for example, higher all-cause mortality (Murray et al., 2020) and poorer mental health (Ford et al., 2007)—the picture is more complicated when comparisons are matched and change is considered over time. On some measures children in care do better than children with similar adversities living at home (see Sinclair et al., 2019 on education). Overall, however, there is relatively little known from the UK about outcomes over time for looked after children compared with similarly vulnerable children living at home, or about outcomes for communities with high looked-after rates. Nonetheless, the ever-increasing rates of children in care and the possibility that some of the increase may relate to professional practice and tensions in the system (Thomas, 2018) suggests that at least some of the cases may be preventable.

There are moral, legal and financial imperatives to prevent children coming into care. The Children Act 1989 contains the principle that children should be cared for within their birth families where possible. A child’s right to be cared for by his or her parents, where it is safe for them to do so, is also an article (no.7) of the United Nations Convention for the Rights of the Child (UNCRC, 1989). What a local authority puts in place to prevent children becoming looked after is very open to them to determine. They can put more or less emphasis on help that is more material (e.g. money, furniture, housing advocacy) or behavioural (e.g. educational or therapeutic programmes to improve parenting). They can choose to target particular types of child maltreatment—e.g. emotional abuse (typically where there is domestic violence or coercion) or neglect—or particular problems for parents or young people—e.g. mental health problems, involvement in organised crime—depending on local need.

Despite rising care rates nationally, there are variations at a local level. Wijedasa et al. (2018) explored changes in the rates of children looked after per 10,000 of the child population in English local authorities at five annual census points between 2012/2013 and 2016/2017. While the majority of local authorities followed the national trend and experienced average increases in the rates of children in care during the period, there were also a number of local authorities where there was an average decrease. There are numerous possible reasons for lower rates of children in care in a particular local authority, including population variations such as migration and economic changes. However, the rates of children looked after vary between authorities with similar demographics. Lower rates of children looked after may also occur in response to changes in practice and approaches used to prevent children being in care.

Local authorities’ approaches and services for working with children and families range from the use of holistic system-wide approaches to more specific interventions (Lepänjuuri & Cornick, 2017), and many of these approaches could in theory help to prevent children being in care. Categories of approaches are not typically defined in national policy, so how they manifest will vary between areas. Whole-system approaches tend to emphasise improving social workers’ relationships with families so that help is better received and care entry can be avoided where possible. They use theory-informed models such as restorative approaches (Williams, 2019), systemic practice (Forrester et al., 2013) or strengths-based practice (Turnell & Edwards, 1999). They attempt to affect the children’s services response at all stages of concern about a child, from initial inquiry through to risk of a child coming into care. There is also the practice of family group conferencing which, either within a whole-system approach or as a stand-alone service, aims to reduce professional power and give more responsibility for decisions about children to wider family members. Many authorities target these on families with children at risk of care on the assumption that more family-based placements will result, preventing children becoming looked after by local authorities.

Early help services are designed to provide support as soon as problems emerge (HM Government, 2018), thus preventing family problems reaching such a crisis point that children need to come into care. Some early help approaches have demonstrated the potential to reduce the likelihood of poor long-term outcomes for children (Chowdry & Fitzsimons, 2016), but the relationship is not straightforward. Early help services may also increase risk awareness and hence increase pressure on the care system (Beecham & Sinclair, 2007). Many local authorities invest in parenting programmes, at various stages of concern about children, from very early lighter-touch prevention to more serious concern and intensive help. There is good evidence of many parenting programmes having a positive effect on families (Barlow & Coren, 2018; Furlong et al., 2012). Services also exist for when crises cannot be avoided and children are at the edge of care. The availability of empirical research evidence on the efficacy of these is mixed. Ward et al. (2014) suggest that some edge of care interventions are effective while others might be harmful (e.g. some approaches to preventing problem behaviour in young people have been found to have the opposite effect (Rhule, 2005)). Edge-of-care services are not clearly defined in policy but they tend to involve more intensive crisis intervention work with children and their families. They may include a specific therapeutic element. There is evidence that in some families, longer term foster and residential care placements can be avoided through the provision of short breaks or respite services (Strunk, 2010).

Whatever the formal evidence base, whether or not particular approaches are adopted depends to an extent on practice wisdom and the extent to which different ways of working are perceived to be effective by local decision-makers. Our paper reports on a survey carried out with local authorities in England about the perceived efficacy of different approaches to reducing the need for children to come into care. This work also sets out to understand the extent to which local authorities evaluate their services and calculate the cost-effectiveness of these services.

Method

Study design

The research was carried out through a mixed-methods survey distributed to all 152 local authorities in England in 2018. It was mixed methods insofar as the questionnaire included some items with multiple choice responses and other open questions with free text boxes. The primary aims of the survey were:

• To identify approaches and interventions employed to work with children and families within local authorities and perceived as effective in reducing the need for children to come into care in their area

• To gain knowledge of the evidence of effect held by local authorities with respect to these interventions

• To identify the extent of any economic analysis

• To explore how long approaches favoured by different local authorities have been in place

• To explore any differences in the adoption of approaches between local authorities according to whether or not their rates of children looked after were decreasing

A key element of the survey design came from Lepänjuuri and Cornick’s (2017) study which explored local authority leaders’ perceptions of a range of issues relating to children’s social care. This element was used because Lepänjuuri and Cornick’s study was multi-wave research with the same population—leaders of children’s services—therefore a familiar style and format for the targeted respondents, and recently conducted by a research organisation of repute (NatCen). Although its scope was broader than ours, it did include an item that was highly relevant and had been developed through consultation and piloted. Lepänjuuri and Cornick’s questionnaire used a list of approaches identified through consultation with local authority staff as helpful in reducing the need for children to enter care. This list was piloted in 12 local authorities, in advance of their Omnibus Survey distribution. The list was adapted in line with the aims of our study to provide a list of potential approaches from which local authorities could identify the three approaches they thought the most effective in preventing the need for children to come into care. The approaches from which the selection could be made were:

• Whole-system Approaches

• Edge of Care

• Early Help

• Family Group Conferencing

• Parenting programmes

• Short break/Respite Services

• Other

The survey asked questions about the underlying rationale for why the approach had been adopted by the responding authorities. It also asked for details of any evaluation undertaken, including the extent of any economic evaluation. Ethical approval for the project was obtained from Cardiff University School of Social Sciences Ethics Committee.

The questionnaire, containing a mixture of open and fixed response questions, was published online via Qualtrics. Once the questions and the parameters had been established, the survey was piloted. After this, minor changes were made and the questionnaire text finalised. A survey link was circulated to the Directors of Children’s Services of the 152 local authorities in England. The initial distribution was followed up with emails and telephone calls over a three-month period between May and July 2018 to encourage completion.

In total, responses were received from 60 local authorities. This represents a response rate of 39.5%. Responding authorities were spread evenly across English regions. Of the 60 local authorities, eight authorities provided only their contact details and a response to an initial question asking the types of interventions that they thought were the most effective. The remaining 52 local authorities provided sufficient detail to facilitate further qualitative analysis. Respondents identified as either directors or other senior managers.

Analysis

Responses to open-ended survey questions were analysed using thematic analysis. For each type of service, qualitative data were coded for over-arching themes which included description of service, families it is targeted at, criteria for use, which service providers (e.g. which sector; inter-or sole agency; statutory, voluntary, both), purpose, mechanisms, implementation and how it fits into the wider service provision. The current paper presents only selected results.

From the multiple-choice questionnaire items, descriptive statistics were produced. The relationship between the timing of approaches that local authorities considered to be effective being introduced and changes in the rates of children in care were investigated by linking the survey data with the data produced about rates from Wijedasa et al. (2018). The mean number of years that the local authorities indicated the approaches they favoured had been in place was calculated according to whether or not those authorities had seen decreases in the numbers of children in care. Mann Whitney U tests were used to test the statistical significance of these relationships.

It should be noted that whole-system approaches, by their very nature, can also include other items on the list above. We made the assumption that any particularly important individual elements within a whole-system approach would be specified in the free-text responses.

Findings

When all responses were considered, the most common approach perceived to be effective was a whole-systems approach, followed by edge-of-care services, and early help (Table 1). Edge of care was particularly common for respondents who did not complete the open-ended comments and without this data, the exact nature of the edge-of-care services cannot be determined. Hence, while table three presents data from the total sample (n = 60), this section below focuses upon the open-ended responses for the 52 local authorities who provided this data.

Table 1. Numbers and percentages of responses that indicated different approaches.

Approach to working with children and families	Number (%) of local authorities selecting this approach as in the top three most effective
	LA that answered open ended questions (n = 52)	All authorities (n = 60)
Whole-system approach	43 (82.7)	49 (81.7)
Edge of care	29 (55.8)	37 (61.7)
Early help	33 (63.5)	34 (56.7)
Family group conferences	20 (38.5)	26 (43.3)
Short break/respite services	8 (15.4)	9 (15.0)
Parenting programme	7 (13.5)	11 (18.3)
Other	12 (23.1)	12 (20.0)

Whole-systems approaches

The approach most commonly claimed to be effective was a whole-systems approach. When exploring why they had chosen such an approach, local authorities cited the values of collaboration, partnership and communication, and relationship-based, strength-based, solution-focused practice. All the respondents who completed the free-text responses used one or more of these terms. While some local authorities indicated that they had introduced these approaches in the early 2010s, for many the adoption had been relatively recent—i.e. since 2015. There was evidence that these approaches are gaining in popularity as 23 local authorities (44% of those that provided data on timing) referred to approaches introduced since 2016.

A range of different types of whole-systems approaches were discussed, as shown in Table 2. Signs of Safety (Baginsky et al., 2019; Turnell & Edwards, 1999), was favoured by around half the local authorities who discussed whole-systems approaches. Many respondents described the use of Signs of Safety as a strengths-based practice that sought to engage families better and help them recognise and utilise their own skills in ways that benefited them. There was a further emphasis on how the use of Signs of Safety promoted family empowerment and generated more partnership-based, collaborative and solution-focused styles of practice, that still ensured families remained aware of existent risks and concerns. A few authorities (n = 4) felt use of the approach had influenced practice from early help right through to edge-of-care services, including reunification.

Table 2. Numbers of local authorities that discussed each type of whole-system approach.

Type of whole-systems approach	Number of local authorities
Signs of Safety	21
Systemic practice	6
Restorative practice	5
Multi-model (draws on multiple models)	5
In house model	2
Strengths-based model	2
System leadership approach	1
Family safeguarding model	1
Total	43

While Signs of Safety is clearly popular, a range of other whole-systems approaches were also reported, as seen in Table 2. Overall, the approaches selected most often, Signs of Safety, systemic practice and restorative practice, suggest that local authorities are at least aspiring to move towards more strengths and relationship-based models of practice.

Early help approaches

Early help approaches were discussed by 33 local authorities. A wide range of different early help services were described and there did not seem to be an agreed consensus about what constituted “early.” Local authorities tended to describe flexible early help services, individually tailored to family needs. These included short, often practically focused early support and targeted programmes of longer duration with focused support delivered when needed. Many respondents talked of partnership; strengths-based, relationship-based and solution-focused practice; and team-around-the-family approaches. Seven responses highlighted multi-agency working as key in the delivery of their early help services. The use of children centres in the delivery of early help services was highlighted by six authorities. Some respondents discussed particular programmes focused on families with children of specific ages. Examples included a programme for children 7+ which focused on issues that are more evident in the teenage years (exploitation, self-esteem, caring responsibilities); and services via Children’s Centres with a particular emphasis on children under two in the 30% most deprived neighbourhoods. Some local authorities described early help services introduced in the early 2010s, but many were much more recent, with five authorities discussing approaches introduced in 2017.

Edge-of-care services

Twenty-nine local authorities selected edge-of-care services as effective in reducing the need for children to come into care. A range of different edge-of-care services were described, mostly bespoke services containing different elements of support and delivered in different ways in individual local authorities. Many authorities described support interventions targeted at adolescent or pre-adolescent young people, more frequently because they were at risk of care, but sometimes if reunification was being considered. Some described services delivered for a time-limited period. The earliest any of the edge-of-care services discussed were introduced was 2008. However, local authorities seemed particularly likely to discuss edge-of-care services introduced very recently, with 12 authorities having introduced the approach during 2017/2018. Two respondents reported ongoing discussions of developing and piloting new edge-of-care approaches.

There were a variety of services and support mechanisms provided through these services. Many interventions contained at least an element of a therapeutic intervention. Five responding local authorities specifically mentioned multi-systemic therapy and other examples of therapeutic elements were mediation and relationship-based practice. There were also a number of services described as intensive support services. Several authorities mentioned interventions that used respite care or family group conferencing as part of an edge-of-care service. Other elements of support that were mentioned, albeit less frequently, included mediation, parenting support, Signs of Safety, and advocacy.

Family group conferences

Twenty local authorities discussed their use of family group conferences in their responses. In comparison to the other approaches explored by the survey, family group conferences were more likely to have been used for a number of years in the authorities that selected them, with the practice introduced as early as 2005 in one authority. Nevertheless, there were still some authorities who had introduced this approach more recently; 5 of the 20 authorities had introduced it in 2016 or 2017.

Many respondents emphasised how they value family group conferencing and its emphasis on finding solutions through family networks. There were examples of authorities using both in-house services or commissioning them from outside. Most authorities were employing family group conferences with families on the edge of care. Five authorities indicated their use with children subject to child protection plans, seven used them in cases subject to Public Law Outline, and three discussed their use in relation to reunification. There were, however, a small number of authorities who were talking about extending the offer of family group conferences to all families across the whole spectrum of needs. Family group conferences were also said to be employed to explore the potential for kinship care in several authorities.

Short break and respite services

Seven local authorities specified short break or respite services among their top three approaches for preventing children being in care. Of all the approaches discussed in the survey, this was the approach most likely to have been in place for some years, with four of the local authorities indicating that they had been using these services since before 2000.

The services provided by three responding authorities who gave sufficient detail were for disabled children. One described their respite care as being for children with complex disabilities to allow families to recuperate. The services described included short respite foster placements for very disabled children that were focused on “activity-based provision and mentoring,” with the overall aim of helping families stay together for as long as possible.

In contrast, three other services were for families living with problematic child behaviours as exemplified by a weekend and holiday time respite service for children aged ten and over at the edge of care. This service, described as “a therapeutic crisis intervention” was said to fit with the authority’s overall practice model, Signs of Safety—no further information was given as to how. Another authority said it had been supporting families where a relationship breakdown was possible with foster breaks for nearly twenty years. One authority was offering respite for both children with disabilities and behavioural problems.

Parenting programmes

Seven local authorities selected parenting programmes as one of the top three approaches they thought effective for reducing the number of children coming into care, although only six of these gave full descriptions of the parenting programmes they were using. While one authority discussed a parenting programme introduced before 2000, others indicated the programmes were introduced more recently.

Half of these local authorities described the delivery of a single parenting programme, two of which discussed the delivery of the Triple P parenting programme, and the other a bespoke local programme. The other three authorities described the delivery of a range of different parenting programmes, generally targeted at families with different types of need, for example, one local authority had different parenting programmes targeted at BME communities, the parents of adolescents, young children and pre- and post-birth families. A range of different programmes were mentioned. In addition to Triple P, these included Incredible Years, Strengthening Families Strengthening Communities, Parents Under Pressure, Pause and the Family Nurse Partnership. There was evidence of parenting programmes being used both as part of an early help programme of support and in families when children were on the edge of care.

Other approaches

The survey also provided the opportunity for local authorities to indicate any other services or approaches not listed by the online questionnaire that they used and believed to be effective at reducing the rates of children entering care. Twelve local authorities provided responses to this question, and their responses covered a variety of different approaches or programmes.

Two responses focused on overall ways of working, as opposed to specific programmes. These reflected some of the comments made in the section above discussing whole-systems approaches, with respondents discussing whole family approaches, and the principles of approaches such as restorative practice and family group conferencing. Other responses discussed more specific types of practice. Three responses raised the Family Safeguarding Model, which contains several elements including motivational interviewing, the co-location of professional teams to work together and the use of group supervision (Forrester et al., 2017; Rodger et al., 2020). Three authorities focused on using panels to allow wider involvement of staff in the organisation in consideration of family cases where children were at imminent risk of coming into care. The remaining three responses considered different individual programmes.

Association between timing of new approaches and reduction in care rates

As described, one of the aims of the survey was to identify if there were any differences in the adoption of approaches between local authorities according to whether or not their rates of children looked after were decreasing. The analysis for this was based on work carried out by Wijedasa et al. (2018) which looked at changes in the rates of looked after children in English local authorities between 2012/2013 and 2016/2017. Twenty-one (40.4%) local authorities who responded to the survey had experienced a decrease in the rates of children entering care during this period. In considering the relationship between approaches favoured by local authorities and the changes in children in care, it is important to factor in the time that an approach has been in place in that local authority. Approaches in place before this period could potentially have affected the numbers of children in care; however, where approaches have been introduced more recently such an effect could not be possible to ascertain.

Table 3 shows the mean number of years that the local authorities indicated the approaches they favoured had been in place according to whether or not those authorities had seen decreases in the numbers of children in care. Figures are provided for the four most frequently discussed approaches and are based on those who provided fuller responses to the survey, including when approaches and interventions were brought in (n = 52).

Table 3. Mean number of years that approaches have been in place in local authorities according to whether or not they have seen decreases in the rates of children looked after.

Approach		Decrease	No decrease	Sig. of Mann–Whitney U tests
Whole-system approaches	Mean	3.5	2.1	.011
	SD	1.9	1.8
	n	19	24
Edge-of-care services	Mean	4.3	2.1	.044
	SD	3.1	1.9
	n	9	20
Early help	Mean	4.5	3.4	.068
	SD	1.5	2.1
	n	13	18
Family group conferencing	Mean	4.6	5.0	.824
	SD	3.2	3.5
	n	11	9

Results show that local authorities who had not experienced decreases in the rates of children in care were more likely to be discussing some of the approaches introduced more recently. With respect to whole-systems approaches and edge-of-care services, these associations were statistically significant at the 0.05 level.

Evaluation

Respondents were asked about the levels of evaluation that had been carried out for each of the approaches they discussed. For all the approaches, a large proportion of authorities indicated that evidence of effectiveness came from the reports of staff, children and families involved in the services. Evidence was also frequently obtained from analysing local authority data on children entering care after using a service. There were however a number of interventions that had been subject to more robust evaluations. Table 4 shows the numbers of authorities that indicated the approaches they discussed had been subject to either independent or in-house evaluations.

Table 4. Numbers of authorities indicating different types of evaluations.

Approach discussed	Number (%) of authorities that indicated
	Independent evaluations	In-house evaluations	Selected approach
Whole-system approach	15 (34.9)	5 (11.6)	43
Edge of care	5 (17.2)	7 (24.1)	29
Early help	1 (3.0)	2 (6.1)	33
Family group conferences	1 (5.0)	2 (10.0)	20
Short break/respite services	1 (12.5)	0 (0)	8
Parenting programme	3 (42.9)	3 (42.9)	7

The variation in the responses to the level of evaluation indicates that some approaches were much more likely to have been subject to robust evaluations. Robust evaluations seemed particularly likely in relation to parenting programmes. A similar trend is seen in the published international research evidence, where parenting programmes are more likely than many other social care approaches to have been subject to robust comparative studies, as can be seen from browsing evidence clearing houses, such as the California Evidence-Based Clearinghouse for Child Welfare (https://www.cebc4cw.org/). This may be because of evaluation feasibility the disciplinary orientation of the evaluators. Of the authorities that discussed parenting programmes, three indicated that their parenting programmes had been the subject of an independent evaluation, two of which were randomised controlled trials, and the other a single-group pre-post study. One of these local authorities and two additional authorities also indicated in-house evaluations. There was also evidence of a high level of evaluation among those who discussed whole-systems approaches, 15 of which reported independent evaluations. Two of these were randomised controlled trials and one had effectiveness data based on two-group non-randomised comparative studies. The remaining evaluations were single group studies. A further five authorities cited evidence based on “in-house” evaluation studies, all five being single group studies: two with data from one point only, and three which included both baseline and follow-up measures. There was also a reasonable level of evaluation among the edge-of-care services. Five out of 29 local authorities indicated that the evidence of effectiveness was based on independent evaluations (two randomised controlled trials, two single-group pre-post studies, and one with data from one time-point only). Seven out of 29 authorities had undertaken in-house evaluations of their edge-of-care service.

Evaluation was reported much less frequently in relation to some of the other approaches. In spite of the popularity of early help approaches with local authorities, evidence regarding their effectiveness was largely based on reports that staff, and/or children and families reported that the services were effective or from analysing local authority data. Only one authority indicated that an independent evaluation had been carried out, and this was a single-group pre-post study. Two further authorities indicated that there had been in-house single-group studies with data from one point in time only. Evaluations were also rarely reported for family group conferencing services, or short break/respite services, even though these were the interventions most likely to have been in place for longer. Only one authority indicated that there had been an independent evaluation of its family group conference service, a single-group study with data from one point in time only, while two local authorities indicated that they had carried out their own evaluation of the services, both single group pre-post studies. In one local authority, an independent evaluation of the short break/respite services had also been carried out. This was a single-group study with data from one point in time only.

Economic evaluation

Two-thirds of the local authorities surveyed indicated an awareness of economic evaluation. The focus was predominantly on quantifying cost savings associated with changes in outcomes such as the number of children avoiding care, reduction in child protection plans or changes in the number of looked after children without weighing this against the cost of the interventions themselves. This constitutes a partial economic evaluation and a full assessment of cost–benefit or cost-effectiveness cannot be gauged with this information alone.

Measurement and valuation of costs unaccompanied by the measurement and valuation of appropriate outcomes is not sufficient to draw conclusions on cost-effectiveness or cost–benefit. While an intervention or programme can increase costs overall, it may still be cost-effective if the economic value placed on the improvement in outcomes exceeds incremental costs. This highlights the importance of identifying, measuring and valuing costs and outcomes jointly of both the intervention aimed at reducing the need for children to come into care and a suitably relevant comparator such as standard practice or “do nothing” if no suitable comparator exists. Only when all this information is available can economic evaluation results be used to inform decision-making on the efficient use of resources in children’s social care.

Only one local authority referred to a comparator group in their feedback by mentioning benchmarking against a statistical neighbour. Unfortunately, they did not describe how the statistical neighbour had been used in the evaluation; it is, therefore, unclear whether they represented a suitable comparator or whether the same intervention had been evaluated from a different authority’s perspective. Two authorities adopted a more robust approach to a cost analysis by carrying out a return on investment analysis. Again, these are only partial economic evaluations as they only measure the amount of return on investment in relation to the intervention’s costs and do not take into account outcomes or present information for a comparator group.

More than one local authority, with good reason, stressed the challenge of carrying out economic analyses in their area of work. Owing to the novelty of applied economic research in this area, it is important that economists are involved early to guide the economic analysis plan, ensuring that all relevant costs and outcomes data are identified, collected and analysed appropriately.

Discussion

This paper reports the results of a survey to determine what approaches local authorities perceive to be the most effective in reducing the number of children entering care. A whole-system approach was most popular, followed by edge-of-care services, early help, family group conferences, parenting programmes and short break services, in that order. Overall, the most common source of evidence was that of self-report accounts from practitioners and/or families who thought the approach worked, although there were also references to independent evaluations. Many local authorities reported economic analyses, but what was described constituted partial evaluation due to insufficient detail for full assessment of cost–benefit or cost-effectiveness.

Many of the approaches discussed were reported as having been introduced over the last five years. This may indicate a high level of change in approaches in local authorities; perhaps because local authorities periodically revise their approaches. Also, although reducing numbers of children in care has not been an explicit national policy, there has been increasing policy interest in the issue in recent years—e.g. a report by the Chief Social Worker (Trowler, 2018) and What Works for Children’s Social Care prioritising this theme in its set up phase (e.g. Brand et al., 2019). The apparent recency of approaches may also reflect staff turnover and lack of organisational memory. It is also possible that a specific focus on reducing numbers in care has become more of a priority in recent years. Interestingly, there was a tendency for those local authorities who were not able to achieve reductions in the numbers of children in care to discuss approaches introduced more recently, a trend that was statistically significant for whole-systems approaches and edge-of-care services. Caution is advised when interpreting this finding, as it could be that local authorities who have been experiencing increases in the numbers of children in care are more likely to highlight approaches introduced more recently. Moreover, authorities that experienced decreases in children in care, may perceive the interventions introduced before those reductions to be effective, even if they are not responsible for the reductions. Further work would therefore need to be carried out to understand this trend more fully.

Other studies confirm the popularity of whole systems approaches, and Signs of Safety in particular. Baginsky et al. (2020) report Signs of Safety being used exclusively in one-third of English local authorities and alongside other approaches in a further third. Despite its popularity, there is, as yet, little robust evidence that its use reduces care entry and crucial to effectiveness is how it is implemented in local contexts (Baginsky et al., 2019; Sheehan et al., 2018).

Evaluation is of course important more generally, so that local authorities can know more about what they do is making a difference, to whom and why. Evaluations need to put the experiences of children and families to the fore and they need to compare the approach in question with an alternative through a comparative study design. Economic evaluation is especially under-developed in local authorities. The recent systematic review by El-Banna et al. (2021) makes several recommendations for fuller economic evaluation in children’s social care, with cost-effectiveness analyses and cost-consequence analyses being said to be especially relevant.

There were a number of clear limitations of our survey. It was about what local authorities perceive to be effective and not about approaches, and interventions for which we have reasonably objective evidence of effectiveness. The study relied on self-report, the accuracy of which could not be checked within the limitations of a small-scale study. Telephone interviews, rather than a self-completion survey, would have allowed more probing of responses and more contextual detail, although delivering interviews at a national scale would have been time-consuming and therefore expensive. The response rate was disappointing. Contextual information was limited to data on change in care rates, so did not include local social and economic factors or important external interventions such as Ofsted inspections. Responses were of course shaped by the categories of services offered to them in the questionnaire. There was a degree of variation in how the approaches were understood by respondents, and these were not clearly defined and bounded categories. It is quite possible that “whole-system approaches” would also include some other items offered as alternatives on the list, such as family group conferences, and that different local authorities would have approached this categorisation problem in different ways. We do not know the relative emphasis placed on local authorities on each of the top three selected approaches. We did not collect data on whether approaches had been partially or fully implemented and the timing of this. All of these limitations could be addressed in further studies.

It also needs to be noted that the survey did not encompass other reasons for change in care rates beyond those services provided or commissioned by local authorities. It is important to acknowledge that practice in children services is only one dimension of a complex problem. Many other intersecting forces affect change in rates of out-of-home care, including the role of other sectors (education, health, criminal justice) and structural factors which affect demand for services, such as changes in income (Wijedasa et al., 2018). Even where local authorities have strong and credible services, there can be barriers that challenge their effectiveness. For example, even where local services agree to the importance of early help, funding cuts can work against this preventative focus (Webb & Bywaters, 2018).

Conclusion

The survey presents a picture of the current popularity of different approaches to preventing the need for children to come into care, in English local authorities. This includes strong support in the sector for whole-systems approaches, and a clear trend for local authorities to highlight approaches introduced in recent years. More and better comparative evaluation of the various approaches used would be very helpful, including designs which allow for fuller economic evaluation.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by funding from the Department for Education, England, UK [grant number 41070002828; 16.10.2017]. The CASCADE partnership receives infrastructure funding from Health and Care Research Wales.

Notes on contributors

Cindy Corliss

Cindy Corliss, PhD is a Research Associate at CASCADE, the Children's Social Care Research and Development Centre at Cardiff University. Her research interests include children's social care, childhood mental health, young people and social media, and LGBTQIA+ identities.

Samia Addis

Samia Addis, PhD is a Senior Public Health Researcher at the Adverse Childhood Experiences (ACE) Support Hub, Public Health Wales. Dr Addis is an experienced academic researcher in population and public health, health services and social care.

Asmaa El-Banna

Asmaa El-Banna, PhD is a research fellow at the Centre for Health Economics at the University of Warwick. Her research focuses on economic evaluations of health and social care interventions using both trial-based and model-based methods.

Nina Maxwell

Nina Maxwell CPsychol. is a Senior Research Fellow at CASCADE, the Children's Social Care Research and Development Centre at Cardiff University. Her research interests focus on adolescence and criminality, workforce development and private family law.

Jonathan Scourfield

Jonathan Scourfield is Professor of Social Work at Cardiff University and Deputy Director of CASCADE, the Children's Social Care Research and Development Centre. His research has covered several different aspects of social work, including child and family services, working with men, social work education, and research capacity-building.

Nell Warner

Nell Warner, PhD is a Research Associate at CASCADE, the Children's Social Care Research and Development Centre at Cardiff University. Her research interests include children's social care, parenting and family support interventions, the emotional well-being of parents, outcomes for children in adverse family circumstances.

Annie Williams

Annie Williams, PhD is a Research Fellow at CASCADE, the Children's Social Care Research and Development Centre at Cardiff University. Her research interests are primarily focused on children and family service systems.