Considering social work assessment of families

Abstract

Since the 1990s the way in which social workers respond to referrals of children to Children's Social Care departments has evolved. It has moved through a process that ‘screens families out’ of child protection assessment to a system aiming to ‘screen families in’ where necessary, and now uses a holistic assessment aiming to screen for both risk and need. The assessment framework developed to assess children in need and their families is the modern social work response to all referrals. Little research has been carried out to assess its suitability as a widespread social work response. This article considers the debates that have emerged in relation to its use and concludes that insufficient consideration has been given to evaluating assessment as an appropriate measure of need and risk. Wider provision of non-assessed universal services would reduce the need for assessment.

Keywords::

• refocusing
• assessment
• safeguarding
• child protection
• children in need

The shift to preventive strategies in relation to child welfare interventions in England after the refocussing debate of the 1990s has created a dilemma regarding family assessment. If assessments of need carried out under s. 17 Children Act 1989 are used as a means for also assessing risk they become potentially oppressive and intrusive. This raises issues of consent, coercion, the importance of accurate assessment and whether the experience of assessment itself causes harm. It also raises a more fundamental question about targeted early intervention strategies which require assessment before services are offered.

The problem of how best to assess families for need and risk was the impetus for the re-focussing debate in the mid- 1990s, resulting in assessment which treated all referrals as requests for services. Social work assessment of families became primarily a gateway to service provision. Families could be screened into child protection investigations if risk was identified during assessment, making risk assessment central to the modern assessment process. The debate was a response to the problems of screening families out of the child protection process as opposed to screening them in (Gibbons, 1995). Theoretically assessments can be contrasted with a child protection investigation for suspected child abuse, intended to be non-consensual and coercive.

Although this approach seems a solution to the problem of treating all parents of referred children as suspected child abusers, the question remains whether there is much material difference in the post-refocussing approach and whether this approach is appropriate. Whilst an assessment for need containing assessment of risk appears less draconian than an investigation focussing on suspected child abuse, the comprehensive ‘holistic’ assessment (Department of Health, 2000, p. 89; HM Government, 2013, p. 19–20) is intrusive and can be coercive. It places the family under intense scrutiny. Once under scrutiny it is difficult to see how a family removes itself from the assessment process.

This article examines the dichotomy between the coercive, child protection-focussed approach to family assessments prior to the mid -1990s, the change towards an ostensibly more supportive and less coercive framework and the recent integration of the two assessments (HM Government, 2013). The article considers models of English and international social work approaches to early intervention and assessment and offers a discussion of the assessment model used in England. It concludes that refocussing family assessments towards support has not fully resolved fundamental problems, including those of consent, coercion and effective evaluation of the impact of assessment, suggesting a more universal provision of support services.

Social work response to child welfare referrals in England

Before the mid1990s families referred to social services because of child welfare concerns were routinely screened using child protection procedures regardless of the reason for referral (Gibbons, 1995; Parton, Thorpe, & Wattam, 1997). Following the review undertaken in Child Protection: Messages from Research, (Department of Health, 1995) assessments underwent a process of refocus away from an emphasis on investigating child abuse. This change meant families were assessed primarily to ascertain whether they were in need of services, as opposed to an initial focus on whether the referred child was at risk of significant harm.

Child protection investigations screening out the possibility of child abuse were supplanted in the late 1990s by the Initial Assessment and Core Assessment, derived from the Framework for the Assessment of Children in Need and their Families guidance (Department of Health, 2000). These assessments were intended to be consensual and to create a sense of families and social workers ‘working together’, a key feature of early intervention strategies (Skinner, 2003). Munro recently investigated whether the aims of refocussing have been successfully achieved (2011). Following her final report (Munro, 2011) the distinction between Initial and Core assessment has been removed and the time allowed to carry out assessment has been extended to 45 days from the point of referral (HM Government, 2013, p. 23).

Local Authorities have a duty under the Children Act 1989 (c.41) to investigate suspected child maltreatment pursuant to s.47 and to assess children in need of services pursuant to s.17. The inter-relation between these statutory provisions is complex as although children who are maltreated can reasonably be described as also being ‘in need’; children who are ‘in need’ are not necessarily ‘at risk’: S. 17 concerns children in need of services and in the legislation is not directly linked to the identification of parental failure. S. 47 has been interpreted as being more focussed on parental failings and is coercive as it confers on local authorities a duty to investigate suspicions that a child ‘is suffering or is likely to suffer significant harm’ and to consider whether there is a need to intervene to prevent it.

Prior to the mid 1990s the investigative/policing approach dominated social work response to all referrals, favouring the more coercive route envisaged by s.47. Research findings showed that families were ‘screened out’ rather than ‘screened in’ to this process (Gibbons, 1995). The model sought primarily to establish whether a child ‘is suffering or is likely to suffer significant harm’ (s.47) before considering whether they were in ‘need of services’ (s.17). This model was developed despite the legislative intention to remove families from the threat of excessive non-consensual state interference unless there were clear grounds for it (Levey, 1991).

Concern over the dominant child protection approach to assessments crystallised in the Conservative Government-sponsored research, Child Protection: Messages from Research (Department of Health, 1995). This research coalesced growing opinion that the focus of social workers on risk and investigation of suspected child abuse tended to exclude the possibility of a more supportive service-provision approach. It concluded that social work was focussed almost exclusively upon risk and investigation of suspected child abuse to the exclusion of assessing and providing services to children in need (Department of Health, 1995). This problem was so acute that some children in need were deliberately wrongly categorised by social workers as being at risk of abuse in order to prioritise their need (Department of Health, 1995, p. 27). Having highlighted these issues Messages from Research advocated refocussing child protection policy and practice (Parton, 2008, pp. 169–170). This involved a prioritisation of s. 17 in response to referrals of children and families to social services, thus removing a perceived cause of conflict between families and social workers during assessments.

Family assessment as part of the early intervention agenda

The refocussing debate laid the groundwork for the Labour Government's early intervention strategies of the late 1990s (Skinner, 2003, p. 19). A range of criminological studies just before the Labour Government was elected in May 1997 argued that family influences and factors were linked to increased risks of offending as a result of a correlation between early problematic childhood behavioural problems and poor parenting (Graham & Utting, 1996, p. 88; West, 1996; Farrington, 1996). Consequentially child welfare policy has developed to encompass increasing opportunities to provide targeted services to families where poor parenting is identified aiming to reduce the risk of child abuse and adverse life choices in adulthood (Allen & Duncan Smith, 2008; Gross, 2008; Allen, 2011). The policy focus has been aimed at supporting and working in partnership with families on a multi-level basis to ‘reduce child poverty, build stronger communities and reduce crime’ (Skinner, 2003, p. 19) as well as predict and prevent incidents of significant harm.

The new approach to assessment was easily incorporated into the wider early intervention debate with its emphasis on targeted ‘early prevention strategies’ (Parton, 2008, p. 172). Allen's report Early Intervention:The Next Steps, An Independent Report to Her Majesty's Government takes as an axiom that early adverse experiences are shown in research to impact on a child's future:

We have [previously] illustrated the negative influence of adverse experiences on the infant brain and the subsequent negative effects in terms of crime (especially violent crime), poor examination results, higher rates of teenage pregnancy, lower rates of employment, higher rates of depression and suicide and substance misuse. (Allen, 2011, p. 23)

The refocussing debate signalled the emergence of a new discourse; that of partnership between the state and families, rooted in social welfarist ideology. This approach emphasised the need to examine family difficulties in a wide context, rather than focussing on investigating single events which were unlikely to capture the complexity of family problems. It also supplanted the idea of the ‘dangerous family’ with the more complex concept of the family in need of support. There is an inference, however, in early intervention literature that poor parenting is at the heart of the need for early intervention (Peters, 2012, p. 413) and there is a failure to separate out categories of parents where there is an element of parental insufficiency and families where there is a need for services for reasons other than parental failings, for example where a child is disabled or a parent is disabled or ill.

Prevention via early intervention strategies became important not ‘simply to combat risks but also to enhance the opportunities for child development via maximizing protective factors and processes’ (Parton, 2008, p. 173). It was assumed that a shift in policy away from the investigatory/policing was an improvement:

In general terms, the refocussing trend has received support within the social work profession, despite some concerns that it could reduce vigilance in relation to child protection. (Carrington, 2000, p. 26)

However, if a supposition exists that poor parenting will (or at least is highly like to be) present in all cases the assessment of risk must be a component of the assessment. The early intervention strategy's aim is to intervene before problems escalate to a level which could be categorised as child abuse. It assumes an escalating level of risk. Without some identified risk it could be argued there would be no referral in the first place. On this basis all parents referred to social services face being labelled at high risk of abusing their children regardless of the reason for referral.

Although the post-refocussing differences in assessment may be obvious to social workers, the move from one model of social work response to another may be invisible from the perspective of families:

Separation into policing and helping cases may be possible from the perspective of those delivering service, but not from the perspective of parents on the receiving end of service. Indeed, many of the parents in this study who described power being wielded over them were voluntary clients. From a parental perspective, a differential response may not separate intervention that uses power over them from one that employs power with them. (Dumbrill, 2006, p. 36)

This highlights the somewhat artificial distinction between the two models of assessment as they currently stand, particularly as Davies notes (2010), there is an element of child protection investigation inherent within the assessment framework.

Rather than conclude it is unimportant to families which model is used, the research suggests that both approaches are problematic when applied as a universal response to referrals that fall under two separate statutory provisions. Placing all referred families in a category where parental insufficiency is assumed to exist albeit on a continuum may be a bar to access to families who have genuine need. Assessment can be a source of significant stress for families particularly if parental failure is assumed from the outset (Amphlett, 1991; Howitt, 1992; Cleaver & Freeman, 1995; Bell, 1995, 2002; Spratt & Callan, 2004; Davies, 2010).

Areas of uncertainty for evaluation in the ‘dual purpose model’ of assessment

Early intervention assessment and service provision does not yield sufficient data to enable comprehensive analysis of the consequences of social work assessment in relation to its impact on welfare or in reducing child abuse. In areas of state intervention it is important to be able to evaluate the costs and benefits of the interventions. There has been surprisingly little detailed research into the wider issue of how to test the efficacy and success of the post-refocussing approach. Millar and Corby (2006) note that: ‘To date, there has been a limited amount of research into social work with children and families since refocussing’. One reason for this may be the problems of evaluation caused by mixing the purpose of assessment.

One aim of ‘psy’ discourses from which the assessment framework originates is that some therapeutic value is possible, either as a result of the process and dialogue involved in assessment itself, or as a transaction that follows assessment. One of the few qualitative studies into the effect of assessment considered whether the Initial Assessment could have its own therapeutic value (Millar & Corby, 2006). The results raised an interesting point: perception of therapeutic value is only evident where the process is perceived as consensual. The therapeutic value is gained by those who perceive the assessment as part of the help that feel they need, rather than by anything inherent within the method of assessment itself.

Millar and Corby's findings encapsulate the problem of using the same assessment process in coercive and non-coercive situations: Once assessment was perceived as coercive there was no therapeutic value. Their study concludes that:

… where we found persisting disagreement between parents and social workers about the need for external intervention, these parents tended to view the process of assessment as negative and in no way therapeutic. (Millar & Corby, 2006, p. 890).

The issue of consent merits more consideration. This includes whether it is overtly considered prior to assessment, the nature and quality of any consent gained, including whether consent can be said to be informed, and whether consent can be withdrawn. Clarity about what exactly is being consented to in terms of assessment participation and how the information is used is inadequately considered in relation to the new version of Working Together (HM Government, 2013). Alongside the issue of consent is the question of what can, or should be done if consent is withheld or if assessment is coercive. It is known from a posteriori observation that behaviour can be regulated by the existence of a framework of surveillance. A reasonable question about the effect of assessment therefore is whether the threat of, or act of assessment itself causes parental compliance with social work expectations, and whether this is desirable. At the very least it would seem to remove any therapeutic value in the process.

The level of parental compliance with social work expectations is important to the assessment process. If parental compliance is used as a measure of success, assessment could be argued to be a positive transaction between the state and the family. Implicit, however, in concluding it is positive is agreeing that the parental behaviours were in need of modification, and that the modifications brought about improved parental behaviours, as opposed to enforced compliance for the duration of social work involvement.

The phenomenon of enforced and temporary compliance has been widely noted (Provincial Project Committee on Enhancing Positive Worker Interventions with Children and their Families in Protection Services, March 2006, 2006, p. 413; Dumbrill, 2006; Callahan, Field, Hubberstey, & Wharf, 1998; Corby, Millar, & Young, 1996; Corby, Millar, & Pope, 2002; Cleaver & Freeman, 1995). This phenomenon casts doubt on the extent to which social workers are meaningfully able to work with parents particularly as it is clear that some parents consider compliance is their only option. There is foundation in this fear as disagreement with social workers has been shown in research to result in an adverse conclusion for parents:

… a major element in the construction of parenting considered not good enough was where parents were deemed to be ‘not working with the social worker’. (Woodcock, 2003, p. 98)

The rationale of refocussing family assessment includes the aim that partnerships could be developed between parents and social workers working together. In addition to an enforced compliance response two other categories of adverse response to assessment have been identified:

The first was a sullen, passive refusal to listen to the social work explanations or services being offered. The second was a blatant aggression, involving shouting and swearing. In both instances the social workers felt that the parents were misrepresenting the information as criticism, and ignoring the spirit in which it is offered – as the social worker wanting to ‘help’ or to ‘support’. (Woodcock, 2003, p. 98)

Regardless of how sensitively they are carried out, family assessments for the purposes of providing support services are potentially in conflict with assessments for the purpose of investigating child abuse. Issues of consent and coercion are more obviously problematic when child abuse is alleged. The consequence of coercive assessment is known to contribute to poor relations between families and social workers (Donzelot, 1980; Gibbons, 1995; Cleaver & Freeman, 1995; Corby et al., 1996; Corby et al., 2002; Dumbrill, 2006). Whereas a family which seeks assessment for their child on the basis of a disability, for example, may be extremely willing to work with social workers for the purpose of obtaining an assessment, a family which feels there is a pre-existing negative inference in relation to their parenting may not be. The position for social workers is also uncomfortable. As much as social workers may strive to build positive working relationships with families during the assessment process this may be unattainable. Added to this the fear of failing to correctly assess risk, potentially leading to a child fatality, leaves the dual purpose model vulnerable to criticism in respect of incorrect assessment of both risk and need.

In relation to measuring the impact of interventions, post-refocussing assessment practice contributes to a well-documented lack of ability to separate out categories of referral for numerical analysis and evaluation. These issues had already been identified as a problematic area during the refocusing debate (Gibbons, 1995, Parton et al., 1997). Separating out referrals is important in order measure the number of child abuse allegations, the number of children referred because of need and the success of interventions. The most recent assessment process (HM Government, 2013) compounds the problem by extending the integration of assessment. This adversely impacts on the ability to evaluate how well assessments are working. Prior to refocussing Gibbons (1995) note that the social work response to referrals was to filter referrals through the ‘child protection’ system where a large number are filtered out but ‘filtered out’ seems to mean that many were moved from one process to another:

The child protection system might be considered as a small-meshed net, into which are caught a large number of minnows as well as a smaller number of marketable fish. The minnows have to be discarded but no rules exist about the correct size of the mesh. Each fishing fleet may therefore set its own. The meshes are the organisational filters operated by the child protection systems. (Gibbons, 1995, p. 51)

Messages from Research highlighted the problem by concluding that many of those filtered out may be children ‘in need’ and produced a flowchart implying that the majority of children ‘filtered out’ at each stage of the process are assumed to be part of the class of children considered to be ‘in need’. The flowchart is not conclusive as it is noted that ‘the results are not precise’ (Department of Health, 1995, p. 28). The mechanisms for accurately separating out categories of allegations from categories of more general need simply did not exist. The data for the flowchart were based on research by Gibbons et al. who confirmed that the majority of these families were more correctly categorised as requiring services under Section 17 and were thus needlessly screened for child abuse (Gibbons, 1995, p. 115).

The cost of England's model of assessment and subsequent interventions is very high. It is difficult to ascertain accurate comparable year-to-year numbers as a result of varying levels of data collection, thresholds and definitions across areas, making evaluation of financial costs and benefits difficult. In 1996, during the refocussing debate, the annual cost of child maltreatment in the UK was estimated to be £735 million (National Commission of Inquiry into the Prevention of Child Abuse, 1996).

A means of evaluation of the system is important for fiscal as well as welfare reasons. Any state intervention into private life should provide a means by which interventions (including the assessment process) can be evaluated to see whether their outcomes are the intended outcomes, to measure any unintended harms, checks to ensure adequate remedies are available and a justification for the required budget.

Government data indicates that year on year more children are assessed (Department for Education, 2011, 2011). There is no conclusive evidence, however, that following assessment and interventions there has been a significant reduction in the incidence of child abuse directly attributed to such interventions. Although Radford et al. (2011) note some areas of reduction in the incidence of serious child abuse in their recent report there is no evidence to correlate this with a change in social work assessment practice, and there are other reasons why the numeric findings could have changed: the ‘Baby P effect’ has been suggested by CAFCASS (Hall & Guy, 2009) as a possible reason why s.31 (Children Act 1989) refocusing care applications increased following the Serious Case Review: ‘Child A’ Executive Summary which was made public in November 2008 (Haringey Local Safeguarding Children Board, 2008). There are also methodological reasons in Radford et al.'s study which could account for the change, including for example the use of retrospective self-reported data over a period where acceptable parenting practices have significantly changed.

Discussion of England's assessment model

Fundamental to the UK's system of child service provision is the use of threshold criteria for interventions at an early stage of need. Hardicker et al. (1996, 1999) and Bromfield and Holzer (2008, pp. 53–55) categorise levels of state intervention as primary, secondary, tertiary or quarternary. Based on these explanations the Department of Health (2000, p. 90) adopted these descriptors in relation to family interventions, relying on family assessment to enable decision making about the appropriate level of intervention.

Primary interventions, or self-regulation, include those interventions providing support and education, offered to everyone as a universal action to promote conditions to avoid problems arising. Secondary interventions, or supported regulation, include those interventions targeted at families in need. Such interventions provide additional support or help to alleviate identified problems and to prevent escalation, focussing on individuals or families who are considered to be high risk, but may not yet have problems. Tertiary interventions are categorised as coercive and include those interventions involving coercive statutory care and protection services. They provide services where abuse and neglect have already occurred, targeting individuals or families who are known to have problems to minimise their adverse effects. Hardicker, Exton, and Baker (1996:Col. 2) elaborated the usual three-fold levels in relation to intervention services by adding a quaternary level, described as optimising the prospects for children where family problems have resulted in their placement in substitute care. Quaternary intervention thus constitutes the most coercive and final interventions. They are not intended to be rehabilitative, and amount to removing children from their families.

The holistic assessment framework in both the Framework for the Assessment of Children in Need (Department of Health, 2000, p. 89) and Working Together to Safeguard Children (HM Government, 2013, pp. 19–20) sets out the comprehensive nature of the assessment, designed to situate issues that pass the thresholds for state involvement in the wider context of the family and its internal and external functionality. It is described in the Framework for the Assessment of Children in Need as:

… a conceptual map which can be used to understand what is happening to all children in whatever circumstances they may be growing up. For most children referred or whose families seek help, the issues of concern will be relatively straightforward, parents will be clear about requiring assistance and the impact on the child will not be difficult to identify. For a smaller number of children, the causes for concern will be serious and complex and the relationship between their needs, their parents' responses and the circumstances in which they are living, less straightforward. (Department of Health, 2000, pp. 26)

However, under these criteria the boundaries of need are unclear: The Audit Commission considers that a ‘reasonable standard’ is not defined (Audit Commission, 1994, p. 5) and neither are the ‘indicators which would suggest such a standard is being met.’ (Brandon, 1999, p. 11).

Part of the problem is that the assessment is designed to gather data from families in relation to practically every aspect of their lives under the broad headings of ‘a child's developmental needs, parenting capacity and family and environmental factors’ (Department of Health, 2000:Ch.2; HM Government, 2013, p. 20). The assessment is based on a form of diagnostic questionnaire derived from the ‘psy’ disciplines (Rose, 1985), intended to enable a psychiatrist to give a diagnosis. Assessments are carried out by a social worker who is not a diagnostician, using their professional skill, intuition and subjective interpretations of what they hear and see. This includes consideration of health, education, emotional and behavioural development, identity, family and social relationships, social presentation, self-care skills, community resources, family's social integration, income, employment, housing, wider family, family functioning and history, basic care, ensuring safety, emotional warmth, stimulation, guidance and boundaries and stability (Department of Health, 2000, p. 146; HM Government, 2013, p. 20). The assessment also includes secondary data from other statutory sources which may or may not be reliable, or where the family may dispute the accuracy of the data. The assessment includes data concerning family history and information about wider family which is difficult to verify. The majority of the headings include some level of subjective interpretation or opinion. The assessment framework follows the model used in Children in Need and their Families (Department of Health, 2000:Ch. 2, particularly Section 2.2). It was not designed to identify child abuse.

The combined assessment may be useful as a ‘conceptual map’ but this is not an adequate basis on which to conclude parents may be at high risk of, or actually, abusing their children. A conceptual map could be used as the basis for agreement with a family in relation to general areas where support may be useful, and where parental insufficiency is not the issue, but this is not sufficient to ‘understand what is happening to children in whatever setting they may be growing up’ (Department of Health, 1999, p. 26) and this could be particularly problematic where families disagree with social workers carrying out an assessment over what constitutes a ‘reasonable standard’ and disagreement over social work interpretation during assessment.

The combined model and dual purpose of assessment in attempting to assess need and risk simultaneously ignores the differences between the necessary methods and priorities. Dalgleish identified these differences as occurring mainly in terms of how comprehensively situations are assessed, the methods of undertaking the assessment and how the information is aggregated (Dalgelish, 2003). Dalgleish found sufficient similarity between the two types of assessment to argue that ‘both are essentially linked. One should not do one without the other’. (Dalgelish, 2003, p. 88), but this conclusion does not resolve the problem of ‘common scale, varying thresholds’ (Dalgelish, 2003, p. 91) which occurs because social workers are using the same scale for assessment, but operate different thresholds for statutory intervention with different issues of consent and coercion.

A further problem of combining assessment is the risk of the alienation of families from social work involvement, particularly when it is perceived that child abuse is suspected and is being investigated. A ‘one size fits all’ approach to assessment does not address this problem; arguably it could compound it. Assessing all referrals via extensive intrusion into all aspects of private family life leaves families unable to be clear about what exactly is being scrutinised. There is a substantial amount of information concerning families' accounts of the assessment processes, and although much is anecdotal it suggests there is a need for review and research into this specific issue (Howe, 2001). The aim of removing stigma and considering the child in as wide a context as possible via holistic assessment may in some instances have the opposite effect to that which was intended.

There is little research considering the issue of coercion including consideration of whether consent is routinely sought for assessment, the nature of any consent that is obtained, and to what extent it can be concluded to be informed (Radford, 2010, p. 167). In relation to the inherent power imbalance involved in family assessment this is surprising. Dumbrill considers the power imbalance in relation to assessment once it becomes a child protection investigation even when such power is intended as a form of support:

Child protection casework ideally involves workers and parents formulating goals in a partnership process (Campbell 1997; Hall & Slembrook 2001; Healy 1998; Waller 1995). This study suggests that given the power imbalance parents perceive between themselves and workers, an equitable partnership may not be possible in child protection casework. Even when parents perceive workers using this power with them as a form of support, they remained mindful and cautious of the potential for this power to be used over them. (Dumbrill, 2006, p. 36)

Parton et al. argue that the refocus does not ‘fully grasp the nature of practice in child protection’ (1997, p. 241). Davies argues assessment is still fundamentally located in child protection and this argument seems to apply even more strongly to the updated assessment model (HM Government, 2013, p. 20). It is de facto screening for issues relating to child protection and thus a threshold schema predominates (Davies, 2010; Department of Health, 2002, p. 4, 6 and 7). Davies advocates a return to clear separation and boundaries between coercive and non-coercive assessment arguing that even non-coercive assessments amount to child protection assessments. A further ground for critique of the use of the same assessment in respect of all referrals is that its dual purpose renders it ultimately a ‘tick box approach’ and a performance management tool rather than a sound basis for child protection investigation (Davies, 2010, p. 142). If this is the case it arguably weakens its effectiveness as a means of assessing risk. These arguments attack both the use of a consensual assessment for social work purposes that may not be clear to families, and the adequacy of such assessments to assess for a risk of significant harm.

This discussion is not a rejection of the concept that responding to referrals using assessment will inevitably cause problems. However, it does question the effectiveness of the shift from a s.47 to a s.17 response as a means of resolving the tensions caused by state interference into private family life, particularly when it is coercive or when families feel they lose control of the process without strong justification that they should do so in order to protect children. Consideration of a pulling back of the use of an investigative approach at such an early stage would simplify these issues.

International models of early intervention and welfare assessment

There are broadly speaking two international categories of response to child welfare concerns in jurisdictions that provide state welfare interventions. Some countries operate a residual and selective provision whereas others tend towards comprehensive and universal provision.

The UK, North America, New Zealand and Australia fall into the former category where child protection processes are separated from family support services following assessment. The system is characterised as being bureaucratic, investigative and adversarial with an emphasis on professionals' primary responsibility being the child, not the family. This approach prioritises children's rights as paramount. The purpose of risk assessment is in order to instigate a plan for coercive intervention where necessary (Hill, Stafford, & Lister, 2002). While there are significant variations in implementation, particularly in relation to mandatory reporting, the UK, Canada, the USA, New Zealand and Australia all operate within a shared model (Waldegrave, 2006).

In contrast the continental, Western European model encompassing Belgium, Sweden, France and Germany tends towards comprehensive and universal provisions where the concept of child protection is embedded within and normalised by broad child welfare or public health services. This model is voluntary and collaborative. Its focus is on family unity as opposed to individual rights, and professionals aim to interact with the family as a whole. The purpose of assessment is to provide a supportive or therapeutic response to meeting needs or resolving problems. Resources are available to a higher number of families at an earlier stage (Hill et al., 2002).

The key distinguishing characteristic of the Anglo-American model relies on a threshold being reached before secondary or tertiary interventions can be justified. Evidence gathering via assessment in order to justify the intervention is an important aspect of the system. Waldegrave (2006) found this emphasis on investigation to justify a threshold had been reached minimised or overlooked broad family needs. A consensual approach to families, which primarily focuses its resources on enabling parents to create safe environments for their children, and coordinated co-operation of legal, welfare and non-government organisations, may offer valuable pointers ‘to improving child protection work …’. (Waldegrave, 2006, p. 69). There is evidence that practice in England may be moving away from reliance on thresholds:

We came across local authorities which were moving away from the use of thresholds in favour of a more integrated model in which all children receive appropriate help. This is assisted by a multi-agency co-location model, and we strongly encourage all local authorities to consider the merits of moving to this. (House of Commons Education Committee, 2013, p. 6)

In Australia a review of the UK and North American model (PeakCare Queensland Inc., 2007) led to argument in favour of a broader base of primary intervention in the form of universal services which are:

… firmly focussed on those services which are already connected to families such as: maternal and child health services; early childhood education and care; schools; adult mental health services, and drug treatment services. (Scott, 2006, p. 5 in PeakCare Queensland Inc., 2007, p. 29).

This was proposed as offering a multi layered system with a reduced focus on statutory child protection work (Scott, 2006). The Western European model of providing universal services without suspicion and specific risk assessment was seen as a means of reducing the load of statutory child protection focussed work and also simplifying the processes required to filter families through risk assessment systems. Scott concludes the UK and North American model is ‘unsustainable’ as a consequence of the high number of referrals (Scott, 2006, p. 5 in PeakCare Queensland Inc., 2007, p. 29).

The rationale of a scaling back of assessment is that by embedding early intervention services into existing family focussed services with the emphasis on supporting the family the services would, by analogy, support children. Universal provision avoids the need for threshold assessment of need and the associated bureaucratic costs and time constraints. The amount of data collected by the state on many families falling below the threshold of ‘significant harm’ would be reduced. Concern about this type of data and the legality of its collection and use was raised in the UK by Anderson et al. (2006, 2009). Concerns have also been raised in the UK about the over-use of surveillance and data gathering in relation to the approach to assessment (Wrennall, 2010). There is a stronger justification in child protection investigations where there is a reasonable suspicion of significant harm to a child forming the basis for the intrusion into private life, extensive data collection, sharing and retention. There is less justification in cases where it is the need for services that is being assessed and there is insufficient justification to suspect the risk is that of significant harm.

Referrals are identifying specific families for intervention into private life via an intrusive assessment at a very low stage of concern. Assessment is creating a mass of data and bureaucracy identified as problematic (Munro, 2011). Requiring assessment that includes an inference of parental failings simply in order for services to be provided raises issues of coercive consent if services are withheld if assessment is not carried out, or if parents are threatened with child protection processes if they do not comply. The exposure of a family to comprehensive data gathering and intrusive assessment simply in order for services to be provided seems excessive. Munro and Parton for example consider England is moving towards a highly intrusive system:

England is in the process of introducing a mandatory reporting system but not based on any notion of child abuse but on the basis of ‘a cause for concern’, which is not defined in the legislation. The new policy of ‘safeguarding’ children has a much wider remit than just ‘protecting’ children from abuse or neglect. It aims to ensure all children reach the government's ‘preferred outcomes’ in terms of achievement at school, health, and behaviour. (Munro & Parton, 2007, p. 14)

A Western European model of universal early intervention services such as that suggested in Australia would involve a re-evaluation of the current framework for service provision. This would enable a refocus of early intervention away from assessment towards increased universal provision. This would in turn enable a new model of investigation for cases to be implemented where significant harm is reasonably suspected. This model could be more explicitly investigative, such as that proposed by Davies (2010) and would enable an explicit balancing of coercive, investigatory actions with a framework of rights and controls to balance the rights and responsibilities of the state and the family.

Conclusions

Finding an effective model of social work assessment is an important issue. Assessments form a critical decision-making function which can profoundly impact upon families. Accuracy of assessment is vital to any effective system that operates threshold criteria if social work errors are to be avoided. Assessments have attracted less attention than they merit given their critical function since the refocussing debate. As levels of referral have significantly increased in recent years the pressure on social workers to carry out assessment is growing despite cuts in funding and services (House of Commons Education Committee, 2013, p. 58, paras. 162 & 163).

The UK's early intervention agenda inherently involves some risk assessment. However, the risk that is being assessed and by what process is not satisfactorily addressed by the use of a holistic assessment as a response to all referrals. If there is a reasonable suspicion of a risk of significant harm to a child this is a child protection concern and should be referred on to a specific investigation to establish whether a parent (or anyone else) is causing a risk of significant harm or not.

Although this would require restructuring, a move towards this approach would enable a more robust, investigatory approach with clear boundaries and controls focussing on establishing facts related to the concerns raised together with a clear, evidence-based recorded outcome. A process that evaluates parental behaviours has serious legal and social implications for the whole family which may not be apparent when consent is given: an assessment model that separates out cases where parental fault is the issue and addresses this in a framework that takes account of the actors' rights and responsibilities is desirable.

The aim of the 1989 Children Act was not to intervene into private family life unless there is a good reason to do so (Levey, 1991). However the expanding number of referred concerns, assessment, comprehensive data gathering and early interventions are pushing the boundaries of intervention away from this aim. The Western European model of welfare provision which provides wider universal service provision would reduce the need for assessment in cases where families require some support but there is no reason to suspect their children are at risk of suffering significant harm. By expanding universal services a reduction in the volume of assessment of families could be achieved. The emphasis could be on family services that aim to expand universal support and help to pull back the threshold at which targeted assessment would operate.

Disclosure statement

No potential conflict of interest was reported by the author.