Risky work: child protection practice.

Abstract

The introduction of a differential response model to the New Zealand child protection system is an important social policy initiative. However, the differential response literature has yet to address the role that risk discourses play as organising and regulatory regimes in contemporary child protection work, and this paper addresses this gap. Child protection social work is strongly underpinned by discourses of risk, and this is best illustrated in the adoption of risk assessment tools that aim to assist the practices of risk assessment and its management. This paper traces the shifting and discursive functions of risk in child protection social work, and argues that Child, Youth and Family (CYF) (2) social workers are negotiating a complex and increasingly pressured practice environment where difficult decisions can be legitimised through the use of risk discourses. The author's doctoral study, which considered risk discourses and statutory social work practice decisions, is drawn on to illustrate how social workers may inadvertently compromise the differential response system--a system where the discursive functions of risk are likely to remain central and regulatory. There is a danger that CYF social workers might construct their role within such a system as increasingly the assessor and manager of high risk. This paper advocates for social work training and supervision as forums where practitioners can consider and better understand these risk discourses.

INTRODUCTION

The New Zealand child protection system is under siege from an anxious public and ever-increasing political scrutiny. In this way, New Zealand is in good company, as many other international jurisdictions struggle to cope with increasing demands placed on child protection services (Scott 2006). While social work practice has often been the focus of criticism in times of tragedy, and reviews and recommendations attempt to address practice deficits (Doolan 2004), these tragic events and swift organisational responses continue to significantly shape the work of child protection (Parton et al. 1997). This is best illustrated by the uncritical adoption of risk assessment and risk management policies and tools, both increasingly relied upon to assist the complicated and often uncertain work associated with child abuse and neglect.

Over the last 30 years, the discourses of risk have shifted and changed, and this paper traces these changes to consider the role that social workers have in constructing facts about particular risks. This paper is a discussion piece that aims to extend social policy debates around the discursive functions of risk in child protection work. It draws on interviews with social workers to illustrate this point (Stanley 2005, Stanley forthcoming). It is important for policymakers to consider the role that social workers actually have in carrying out policy initiatives in the area of child protection.

A recent social policy innovation in the New Zealand context is the adoption of a differential response model to assist with the organisation and service provision of child protection work (Connolly 2004, Waldegrave and Coy 2005). While this has significant implications for how the work of child protection will actually operate, there has been scant attention paid to how discourses of risk may serve the organisational and practice imperatives for social workers, while potentially rendering families less central to decision making. To date, this has been largely overlooked in the child protection literature (Webb 2006), and this is an important discussion in light of New Zealand developing and introducing a differential response to the work of child welfare.

Child protection work is now strongly underpinned by discourses of risk (Stanley 2005, Webb 2006), best illustrated by the introduction of risk assessment tools that aim to assist social workers to reach conclusions about risks for particular children and families. Risk assessment and its management are also organising principles that both structure social work practice and offer audit and monitoring functions. This paper draws on the family group conference (FGC) to illustrate how monitoring and audit functions operate discursively, and how statements about risk can be drawn on to support organisational imperatives. This paper advocates for social work training and supervision as forums where practitioners can consider and better understand these risk discourses.

In practice, the balancing of child protection and family support imperatives has produced a working context where ambiguity and uncertainty remain important--yet rather under-theorised--practice issues (Stanley 2005). Importantly, those engaged in social policy initiatives and practice debates must consider the role that CYF social workers have as strategists in making use of risk discourses. Such use has the potential to compromise policy initiatives like the differential response model.

THE DIFFERENTIAL RESPONSE MODEL

A significant development for New Zealand's child protection system has been the development and introduction of the differential response model (Waldegrave and Coy 2005). Designed to help determine the most appropriate service provision for families in a more timely manner, the model aims for CYF to make a preliminary assessment that would then lead to a range of outcomes, such as a child and family assessment or a statutory social work investigation. According to Waldegrave and Coy (2005), CYF social workers may face challenges in making initial assessment decisions to enact the possible pathways for families, given the societal pressure on CYF to protect and ensure the safety of children who come to CYF's attention. Further, risk discourses will be drawn on to assist in such determinations, and it is vital that policymakers, social workers and their supervisors are able to engage critically with such discursive functions of risk. The next section illustrates this shifting and discursive functioning of risk.

THREE DECADES OF CONCEPTUALISING RISK

   What we understand to be a "risk" (or a hazard, threat or danger)
   is a product of historically, socially and politically contingent
   "ways of seeing". (Lupton 1999:35)

Over the past 30 years, child protection work can be characterised by three risk periods. The 1970s was characterised by a growing anxiety towards children. Increasingly, children came to be seen as being "at risk". Risk was used to delineate those at risk from those posing a risk and, importantly, those not at risk. Risk entered the official discourse of child protection, and social workers were increasingly expected to diagnose and identify risks for particular children and families (Parton et al. 1997).

This was followed by a technological period in the 1980s and 1990s. This period was characterised by the development of risk assessment tools and risk management policies. Cases could be defined as high or low risk and, accordingly, receive particular responses and resources from state agencies. Formal risk assessment tools were introduced to facilitate this (Doueck et al. 1993, Gambrill and Shlonsky 2001). Increasingly, proceduralised models of practice were introduced to help social workers manage the uncertainty and ambiguity associated with assessment work. These procedures were able to be monitored and audited (Parton 1998), yet they served to mask the more unpredictable and uncertain parts of child protection assessment work.

Internationally, child protection systems favoured actuarial risk assessment tools to enhance certainty around risk assessment practice. Actuarial risk assessments take an insurance approach to assessment work (Kemshall 2002), where risks are aggregated and statistically calculated. Individuals gain access to services based on particular risk classifications, and resources are allocated accordingly. In contrast to actuarial models, New Zealand's child welfare system adopted a consensus-based approach to risk assessment work (Smith 1995).

More recently, a third risk period has emerged--a period of legitimacy, where discourses of risk can be drawn on to legitimise assessment decisions made about particular children and families. This is a significant shift in the relationship between families and social workers because participation is inhibited when there is an insistence on "objective facts" about risk. According to Ferguson (2004), child protection social workers and families may actually hold a different understanding of the purpose of their relationship. The subjective or local experiences of family members are less likely to be the subject of inquiry when social workers seek evidence and proof that a child is or is not "at risk". Moreover, this may lead to social workers inadvertently compromising the assessment relationship.

According to Bessant (2004), a "science of risk" has emerged in social work, providing a frame of "objective" risk knowledge. What she means is that risk is regarded as something that can be located and resolved, and this is influenced by positivist understandings of risk that propose credibility and scientific rigour. Evidence and certainty are privileged, and any ambiguity and uncertainty associated with assessing child care and protection are expected to be managed through risk assessment procedures. To summarise, New Zealand child protection is operating in a context of decision legitimacy, where risk operates discursively; an environment where certainty and objective risk measures are increasingly relied upon by social workers.

NEW ZEALAND'S RISK ESTIMATION SYSTEM (RES)

According to DePanfillis and Zuravin (1999), risk assessment tools were developed to reduce worker inconsistency in decision making. There are also compelling arguments for the use of formalised risk assessments to assist social work assessment work (Appleton and Craig 2006). Coohey (2003:821), for example, argues that without a risk assessment tool in cases of neglect, "it will be difficult to see how [social workers] can be expected to know whether a child is likely to be harmed in the future". Assessing risk, however, is a matter of judgement and not fact-finding. Thus, social workers need to maintain a professional and ethical approach toward it.

   At best, [risk assessment] instruments and models are wonderful
   tools in decision making and good case-work practice. At worst, they
   can negate practitioner responsibility and be used mechanistically
   and defensively. (Smith 1995:10)

The New Zealand Risk Estimation System (RES) was introduced in 1996, and is a consensus-based model. Introduced to assist social work decision-making and risk assessment practice, the tool has received little critical research attention (see Smith 1998). Importantly, this tool was designed to position social workers as central and active judgment makers about particular risks for children and families. It is important to note that the RES is a tool designed to have social workers build an analysis about risk, not locate it as a concrete state. In this way, the scores of vulnerability, severity and likelihood are categories that cannot be collapsed or aggregated to a particular risk level.

UNDERSTANDING HOW SOCIAL WORKERS UTILISE RISK DISCOURSES

In understanding how CYF social workers actually make sense of risk, and their associated decisions about particular children and families, the author discussed risk assessment work with 70 child protection social workers (Stanley 2005).3 The Critical Incident Technique (CIT) provided a model for data collection (Fook 1996), and this was coupled with a grounded theory approach to analysis (Strauss and Corbin 1997). During semi-structured interviews, social workers described both complex and straightforward cases of their child protection risk assessment work. Each case was classified as a unit of data (a critical incident) because this was the actual experience of doing assessment work, and thus an example of salient work experience to the research participant. The study did not aim to represent all CYF statutory social work assessment experience, but the research finding that discourses of risk can be employed strategically by social workers--and without their explicit recognition--to legitimise their practice decisions is relevant to developments in child welfare social policy in New Zealand. The next section of the paper illustrates how the strategic use of risk discourses by CYF social workers actually operates in practice.

DECISION LEGITIMACY--SOCIAL WORKERS MAKING USE OF RISK DISCOURSES

As discussed earlier, New Zealand's risk assessment tool employs a consensus model, and social work assessment analyses are generated through use of the RES and a professional supervision forum. Thus, definitions and determinations of "risk" are designed to be outcomes of social work assessment with families, analysis of case files, professional social work theory and models (Connolly 2006), support tools such as the RES, and supervision consultation. The study indicates, however, that social workers can make strategic use of discourses of risk when managing their dual responsibility of child safety and family support. They do this through a particular use and presentation of "at risk" claims. According to Webb (2006:154), risk claims are attempts to argue for or against a particular decision, and this point has been overlooked in the risk assessment literature.

To illustrate, consider two significant decisions: the decision to remove a child from their home and family, and the decision to leave them there. Many of the social workers who participated in the author's research described decisions to remove children from their home following an assessment of risk. We can see that the rhetoric of risk is useful in supporting and legitimising this decision.

"[The risk assessment] was helpful [in] confirming or affirming that you saw those risks, and it's okay [to remove the children]." (Social Worker 3)

"The [risk assessment] formalises things, [and] also gives you something to back up whatever assessment that you've come to." (Social Worker 27)

Another worker noted:

"[We] couldn't leave the children there after that thing had been said [by the other parent]. I suppose it would be knowingly leaving the children "at risk" even though it's not a quantifiable risk, it's a risk and that may be enough." (Social Worker 32)

Other social workers in the research described cases where they made decisions to leave children at home following an assessment of risk. This worker was typical in her explanation of this.

"[I made] the decision to leave the [teenager] there. I investigated the mother--I don't believe that [she] was at risk enough to uplift." (Social Worker 22)

To support and legitimise decisions to remove, or to leave, children, risk is rendered into an "objective" and quantified (measurable) state. This is a narrow focus on the risks facing these children. In both sets of decisions, risk is focused on as an act by an adult, or an event outside of the child's control. Risks associated with the removal of children to alternative care, and risks associated with being involved with statutory child protection staff, were not woven through an analysis of risks facing these children or families. Decisions to remove and decisions to leave children are rendered legitimate because they conform to a set of established rules; the rules are then justifiable, and consent--from both supervisors and family members--is achieved (Beetham 1991).

The majority of social workers who participated in the author's research explained that the work of risk assessment and its management were CYF social work functions, to be completed and then presented to families. While the Children, Young Persons, and Their Families Act 1989 is premised on family inclusion in decision making, there was overwhelming evidence in the research to show that families' understanding of particular risks was not focused on during assessment interviews. Other workers noted that when parents acknowledged the risks facing children they could be left at home as this was an indication that the parents were "on board" with the social worker's assessment. Families were frequently encouraged to "come on board", and accept risk assessments as presented by social workers.

"[The mother] never believed that her children would be removed, so this was like a huge wake-up call, which made her do a 360-degree turn and start recognising the issues, so she came on board and was doing the work. So it was around, I was kind of putting my faith in her." (Social Worker 5)

While working through these complex issues and the need to confront families with the realities of children's safety, there may be value in social workers more consistently talking through risk assessments with families earlier in the process. Rather than risk them becoming a "virtual object" constructed by social workers, they become something that is arrived at together with the family.

Under a differential response model the type of service offered will follow a preliminary assessment. Cases of high need may receive a child protection assessment, while cases of low need may be directed toward family assessment services (Waldegrave and Coy 2005). The cases that are assessed as being in need of a statutory CYF assessment may tend to be more complex than those assessed to be in need of a family assessment. (4) There is the potential, then, for CYF social workers to construct their role as one of assessing and managing "high risk" cases, when this in fact may not be the case. The case may be one of high need and low risk. (5) Treated as an ontological certainty, social workers can proceed to locate and manage "risk". When risk is treated as an ontological certainty, family definitions of what constitutes risk or risky behaviours are less likely to be incorporated into assessment work.

FAMILY GROUP CONFERENCING: PROFESSIONAL LEGITIMACY

The imperative of family inclusion in decision making about their children is enshrined in the Children, Young Persons and Their Families Act 1989, and operates as the family group conference (FGC). However, unlike the few social workers who described the FGC as a model of family decision-making, the majority of social workers who participated in the author's research held predetermined ideas about the outcomes they regarded as being in the best interests of the child. A small group of social workers described using the FGC as an increased intervention step, aimed at providing the social worker with more formalised power. This next worker described how she went into the reconvened conference to formalise her power as a professional worker:

"When we had the second family group conference, little had changed when we had [this] FGC review, and at that point, I had gone into the conference feeling that we needed to have more authority, more clout about this, and so the agreement was that we would have an application for declaration." (Social Worker 18)

A number of workers described how they used the FGC process as a way to formalise monitoring of the family. For others, the conference provided the mandate to formalise support plans around families:

"They're still in our care, we are going to have FGC next Tuesday, and my recommendation is for the department to take a 101, which is custody order, and the children have been placed with [a relative] they know really well." (Social Worker 23)

"So I'm currently doing another FGC referral whereby I'm going for declaration, because I don't believe mother can keep these children safe." (Social Worker 25)

These social workers were clear that there was both an ongoing role for CYF after the conference and a level of intervention necessary to ensure child safety.

The rhetoric of risk serves a legitimising function in this forum because the assessments and management of it can be audited by supervisors and managers after being presented in the FGC. Further, such practice of decision legitimacy can be theorised as risk aversion. In a practice context where increasing demands and scrutiny are placed on CYF social workers to account for their work, risk can be used strategically to serve the organisational requirements of social workers. For example, social workers may complete an RES for the purpose of case closure. This has implications for the clients of social work practice because children and families are often at the receiving end of professional intervention decisions based on risk classifications (Cleaver and Freeman 1995). Risk aversion requires a reflexive engagement in practice, and this issue is discussed in the next section.

MANAGING RISK AVERSION THROUGH REFLEXIVE SKILLS

According to Alaszewski (2006), the way risk is defined will lead to particular responses by practitioners. If risk is defined as an objective certainty, something that can be located and managed, then risk management and risk aversion are dominant discourses shaping the practice. On the other hand, if risk is defined in a more open and creative way, then the actions of practitioners will also reflect a more open and creative practice. For example, risk taking can be considered and theorised within risk assessment practice. One way to assist social workers, supervisors, policymakers and trainers to conceptualise risk in this more open manner is through the use of a risk continuum (Lupton 1999) (see Figure 1). The aim here is for critical conversations to occur that open up ways to conceptualise risk; importantly, discussions can focus on where practitioners place particular risks on the continuum and why.

[FIGURE 1 OMITTED]

Discourses of risk can be conceptualised as falling along a continuum, ranging from realist epistemologies, where there is little doubt that a child is at risk, to what Lupton (1999:35) referred to as constructionist epistemologies, where definitions of risk are mediated through historical, cultural and social processes. Both ends of the continuum and all of the possibilities in between are useful for social work practice. There will be cases where there is no doubt that a child is at risk. However, there will be many cases where risk is less tangible, more nebulous and slippery. In such cases, if social workers attempt to render risk as a certainty, this may lead to a premature closing down of assessment work, or reaching early decision points that may not take into account all of the risks facing children and families.

Working with uncertainty need not be scary or concerning, something to be swiftly managed; it offers the possibility of being reflexive about the many and varied risks associated with child abuse and neglect work. Increased debate and discussion on what "risk" is will not avoid tragedies in practice; generating more talk about what constitutes risk, and what this may mean for assessment practices, would enhance assessment work. Importantly, it would ensure that discourses of risk are not used to legitimise practice decisions.

A critical engagement with what is meant by "risk" is the first stage in the risk management process. When a narrow definition of risk is applied, the management options are also narrow. This social worker demonstrated this when she explained that managing risk requires the elimination of it.

"Somebody has to be removed from that situation to eliminate that risk ... I think the most successful part of [decision making] was actually [being able to] eliminate the risk for [the child], the immediate risk for him, at that particular time, and putting him in a safe environment." (Social Worker 3)

To talk of risk elimination is potentially misleading, in that risk can never be totally eliminated from the lives of families (Titterton 2005). Further, this may set up unrealistic expectations for social workers, leading to risk-aversive practices.

SUPERVISION AND TRAINING: FORUMS FOR REFLEXIVE CONVERSATIONS

Supervision and training offer forums for deep, critical and reflexive discussions about how "risk" is being conceptualised and applied. There is an opportunity to discuss the concept of risk here because social workers bring to these forums their practice experiences, education, internal training and a generalised knowledge of what they regard risk to mean. These discussions will allow supervisors and social workers to consider how their particular understandings of risk may be divergent or consistent. As Walker and Clark (1999:1439) noted, "clinical supervision can offer compassionate and cost-effective risk management".

Social workers need to maximise the supervision session to build critical skills in discussing, engaging with, and conceptualising risk. Being reflexive about risk assessment work would welcome the perspectives, experiences, voices and understandings of those who receive social work services--children and families. As Oaks and Herr Harthorn (2003:3) argue:

   Perception of risk is best argued to be socially and culturally
   constructed through a complex process that depends on a range of
   social and cultural factors and may be contributed to through
   processes of risk communication.

There is evidence from the author's research that developing reflexivity around how risk discourses operate through practice and policy initiatives may be easily achieved. Social workers reported to the author that the process of talking about risk assessment work was actually useful. Some remarked that the research process was like an in-depth supervision session. The following pieces of interview transcript highlight this point:

"It's good to have somebody ask you questions about what you do and the way you assess things." (Social Worker 27)

"I [also] got a lot of clarity. So thank you ... I actually feel like it's been a supervision session, even though you haven't given me any answers or direction, the questions that you've asked have given me food for thought and I feel quite cleansed or healthy, in being able to put it out there." (Social Worker 8)

"You know risk is, I guess ... one of those things that you can never consider enough." (Social Worker 1)

Risk is articulated throughout all phases of the investigation process and during an FGC. It may not be discussed outside of this context as a risk discourse in the way made possible by this research. Some of the comments made by the workers suggest that greater opportunities to explore the meaning and implications of risk-based decisions may afford valuable insights for workers and supervisors.

SOCIAL POLICY IMPLICATIONS

The challenge, then, is for policymakers and social workers, both inside and outside CYF, to be encouraged to value different conceptions of risk, rather than forcing them into rational technocratic frameworks. The RES was designed to facilitate this. Sadly, social workers are more likely to use risk discourses to order the uncertainties of practice, rather than explore uncertainty as a social construction--something that can then be theorised with families who remain the key stakeholders in the work of care and protection.

Risk discourses play a part in how assessment work is both set up and undertaken with families. This is an important argument for New Zealand social policy developments in the area of child welfare. While the current systems of child protection are unsustainable (Scott 2006), the wholesale introduction of an alternative model brings with it new challenges and risks. One major risk is that families directed toward a statutory child protection assessment may be regarded--that is, constructed by CYF social workers--as being "high risk". This may lead to predetermined or premature intervention decisions. Moreover, this is risky for children and families.

Dumbrill (2006) argues that policy analysts and social workers need to conceptualise power as a variable that will affect social work practice under a differential response model. His qualitative study of 18 Canadian parents considered the experiences of families who had received child protection services, and he found that families regarded social workers as having and maintaining more power than the families, regardless of whether they were operating in the context of voluntary family support or a coercive statutory response. Dumbrill questions the separation of child protection and family support functions of child welfare in light of this. Further, Warner and Gabe (2006) make a compelling argument that policymakers have tended to favour risk assessment models exemplified by objective and rationalised checklists: While there is some utility in actuarial models, where checklists and objective measures are favoured, it is limiting to divorce risk assessment work from the wider context in which social workers, families and organisations are negotiating their understanding of what constitutes particular risks.

Assessments of risk can be presented to families for their acceptance, and this strategically utilises risk discourses. This is a way of ordering the uncertainty that pervades child welfare work, and is a powerful strategy in securing child safety. Importantly, for the New Zealand context, social workers will continue to require organisational and managerial support to manage the ambiguity and uncertainty that is inherent within child protection work. This may prove more challenging under a differential response model because cases will be assessed at the front end of a child welfare system, and differentiated out to receive child protection follow-up or family support responses. Importantly, policymakers need to consider key actors like CYF social workers, who are more frequently at the operational end of policy initiatives. This paper has attempted to re-situate the CYF social worker as a key social policy colleague--someone who can be encouraged to critically engage in what a new initiative may mean for child protection social work.

CONCLUSION

The discourses of risk located in the child protection arena continue to shift and change, and new risks emerge. Social workers and policymakers need to consider the risks associated with foster care, risks in leaving children in particular homes, risks in removing children, risks to themselves, and risks to the agency. Contact with statutory child protection systems is indeed risky for families (Scott 2006). Further, risk discourses are likely to remain important organising principles for New Zealand's child protection work for two reasons. First, as argued in this paper, discourses of risk have entered child welfare policy and practice over the last 30 years with little critical attention paid to how they are utilised in practice. This is illustrated in the wide adoption of risk assessments tools and risk management policies within child welfare agencies.

Second, New Zealand is currently introducing a differential response model for its child welfare system, where CYF social workers may construct their role as one of "high risk" assessors. It is important that training and supervision continue to provide forums for critical conversations about the role and purpose of risk assessment. This is important because risk can operate discursively to assist the management of uncertainty and ambiguity. This being the case, the concept of "risk" requires a more reflexive engagement, because competing demands on social workers to assess and manage risk often support organisational requirements rather than building family solutions (Titterton 2005). Risk assessments can include families when the tools are used to engage talk around what we and others understand risk to mean (Kufeldt et al. 2003). Asking parents if they share similar concerns, and to find out what they think a risky situation would look like, offers social workers important information for their assessment work.

Risk discourses are useful for social work. They assist in the organisation of it. However, this paper has argued that CYF social workers are strategists, drawing on statements about risk to legitimise decisions made about children and families. They do this to manage any uncertainty associated with assessing child abuse and neglect, and it is this risk management that renders families less central to discussions about risks facing their children. Policymakers and CYF social workers need to pay more critical attention to how risk discourses contained within social policy initiatives actually shape the work of child protection, and how they, in turn, are also shaped by them. The place to start is to, organisationally and collegially, help shape a policy and practice culture where critical and reflexive questions about "risk" are both asked and encouraged.

REFERENCES

Alaszewski, Andy (2006) "Health and risk" in Peter Taylor-Gooby and Jens O. Zinn (eds.) Risk in Social Science, Oxford University Press, Oxford.

Alaszewski, A. (2006) "Managing risk in community practice: Nursing, risk and decision-making" in P. Godin (ed.) Risk and Nurshing Practice, Palgrave, Houndsmill.

Appleton, C. and A. Craig (2006) "Preparing human service workers for risk assessment and management: Training for child protection risk assessment in the statutory social work setting" in K. McMaster and L. Bakker (eds.) Will They Do It Again? Assessing and Managing Risk, HMABooks, Christchurch.

Beetham, D. (1991) The Legitimation of Power, Humanities Press International, Atlantic Highlands, New Jersey.

Bessant, J. (2004) "Risk technologies and youth work practice" Youth and Policy, 83:60-77.

Cleaver, H. and P. Freeman (1995) Parental Perspectives in Cases of Suspected Child Abuse, HMSO, London.

Connolly, M. (2004) Child and Family Welfare: Statutory Responses to Children at Risk, Te Awatea Press, Christchurch.

Connolly, M. (2006) "Practice frameworks: Conceptual maps to guide interventions in child welfare" British Journal of Social Work, advance access 16 June 2006.

Coohey, C. (2003) "Making judgements about risk in substantiated cases of supervisory neglect" Child Abuse and Neglect, 27(7):821-840.

DePanfillis, D. and S. Zuravin (1999) "Assessing risk to determine the need for services" Children and Youth Services Review, 23(1):3-20.

Doolan, M.P. (2004) A Life Too Short: Child Death by Homicide in New Zealand, MSW thesis, Canterbury University, Christchurch.

Doueck, H.J., M. Levine and D.E. Bronson (1993) "Risk assessment in child protection services" Journal of Interpersonal Violence, 8(4):447-467.

Dumbrill, G.C. (2006) "Parental experience of child protection intervention: A qualitative study" Child Abuse and Neglect, 30(1):27-37.

Ferguson, H. (2004) Protecting Children in Time: Child Abuse, Child Protection and the Consequences of Modernity, Palgrave MacMillan, Houndsmill.

Fook, J. (1996) The Reflective Researcher: Social Workers' Theories of Practice Research, Allen and Unwin, St Leonards, New South Wales, Australia.

Gambrill, E. and A. Shlonsky (2001) "The need for comprehensive risk management systems in child welfare" Children and Youth Services Review, 23(1):79-107.

Kemshall, H. (2002) Risk, Social Policy and Welfare, Open University Press, Buckingham.

Kufeldt, K., D. Este, B. McKenzie and B. Wharf (2003) "Critical issues in child welfare" in K. Kufeldt and B. McKenzie (eds.) Child Welfare: Connecting Research, Policy and Practice (pp. 395-428), Wilfrid Laurier University Press, Waterloo, Ontario.

Lupton, D. (1999) Risk, Routledge, London.

Oaks, L. and B. Herr Harthorn (2003) "Health and the social and cultural construction of risk" in B. Herr Harthorn and L. Oaks (eds.) Risk, Culture and Health Inequality, Praeger, Westport, Connecticut.

Parton, N. (1998) "Risk, advanced liberalism and child welfare: The need to rediscover uncertainty and ambiguity" British Journal of Social Work, 28(1):5-27.

Parton, N., D.H. Thorpe and C. Wattam (1997) Child Protection: Risk and the Moral Order, Macmillan Press, Basingstoke.

Scott, D. (2006) "Sowing the seeds of innovation in child protection" keynote address at the 10th Australasian Conference on Child Abuse and Neglect, Blossoming of our Children, Wellington, 14-16 February.

Smith, C. (1995) "More than a gut feeling." Social Work Now, 2:8-12.

Smith, C. (1998) "Risk assessment in child protection" presented at the Family Law Conference, Christchurch.

Stanley, T.W. (2005) Making Decisions: Social Work Processes and the Construction of Risk(s) in Child Protection Work, Doctor of Philosophy in Social Work thesis, University of Canterbury, Christchurch.

Stanley, T. (forthcoming) "Assisting risk assessment work: Social workers talking about the Risk Estimation System (RES)" Social Work Now.

Strauss, A.L. and J.M. Corbin (1997) Grounded Theory in Practice, Sage Publications, Thousand Oaks, California.

Titterton, M. (2005) Risk and Risk Taking in Health and Social Welfare, Jessica Kingslea, London.

Waldegrave, S. and F. Coy (2005) "A differential response model for child protection in New Zealand: Supporting more timely and effective responses to notifications" Social Policy Journal of New Zealand, 25:32-48.

Walker, R. and J.J. Clark (1999) "Heading off boundary problems: Clinical supervision as risk management" Psychiatric Services, 50(11):1435-1439.

Warner, J. and J. Gabe (2006) "Risk, mental disorder and social work practice: A gendered landscape" British Journal of Social Work, advance access 17 October 2006.

Webb, S.A. (2006) Social Work in a Risk Society: Social and Political Perspectives, Palgrave Macmillan, New York.

Correspondence T. Stanley, School of Community Studies, Christchurch Polytechnic Institute of Technology (CPIT), PO Box 540, Christchurch, New Zealand. Email: [e-mail]. The author, a registered social worker, previously worked as a social worker with Child Youth and Family.

Tony Stanley (1)

Lecturer Social Work, School of Community Studies

Christchurch Polytechnic Institute of Technology

(1) Acknowledgements An earlier version of this paper was presented at the 10th Australasian Conference on Child Abuse and Neglect (ACCAN), Wellington, 14-16 February 2006.

(2) New Zealand's statutory child protection agency.

(3) The study received ethical approval from the CYF Research Access Committee and the Human Ethics Committee, University of Canterbury. The study considered the sets of relationships between decision making and risk discourses in statutory social work practice in Aotearoa/New Zealand. It is important to note that this research took place in CYF offices nationally from late 2002 to mid-2003--well before the development and introduction of the differential response model.

(4) However, cases of high need may receive a child and family assessment. This is a comprehensive, family-inclusive needs assessment which may be more complex than many statutory investigations. Cases of low need may be referred for family support or other services. Those cases where it is considered that the child may be in need of care and protection will go on to a statutory CYF investigation. A further possibility is that two pathways may be chosen; for example, a child and family assessment and a CYF investigation where one child in a sibling group may be in need of care and protection combined with a range of complex family issues and dynamics (personal communication, Differential Response Model implementation team).

(5) For example, poverty can produce situations of higher need for families, while not necessarily meaning higher risk outcomes for children.