Child Protection Social Work: Risks of Fears and Fears of Risks – Impossible Tasks from Impossible Goals?


Address for correspondence: Brian Littlechild, School of Social, Community and Health Studies, University of Hertfordshire, College Lane, Hatfield, Hertfordshire, AL10 9AB. Email:


This article examines the relationship between the causes and effects of fear in child protection social workers, and the effects of risk assessment and risk management policies on this area of work. The focus on risk assessment and risk management has become a major area of attention within practice, policy and management of child protection work in the UK in recent years. Concepts of risk as constructed by the media, government and the public are increasingly impacting upon professional practices. This article examines the basis and validity of risk assessments in the social professions field, and particularly within the child protection arena. The article goes on to examine the experiences of fear arising from the risk agenda, which affects frontline workers, managers and child protection agencies. This agenda arises from centrally produced risk assessment frameworks, alongside unrealistic expectations from central government of prediction of risk by the use of current risk assessment tools. Such controlling policies from central government can lead to fear and anxiety in social work professionals of not assessing and eliminating risk, as the government and their employing agencies are expecting them to do. The article also proposes that this risk agenda fails to address a key element in the assessment of risk – how social workers experience threats and stress in their work, and the pressures they can be subject to within it, particularly in relation to violence and threats from parent service users where their children are being investigated for possible child abuse.

Risk assessment and risk management have recently developed as major areas of attention within practice, policy and management of child protection work in the UK, with a growing literature base (Munro 2002; Calder 2003; Thom et al. 2007; Webb 2007). Concepts of risk as constructed by the media, government and the public are increasingly impacting upon professional practices (Denney 2005; Morgan 2007).

In recent years, developments in risk discourses within central government agency policy and guidelines have meant that the risk enterprise has gained momentum. However, in the area of risk in relation to social work and social care, little work has been carried out to produce a social model of risk (Titterton 2005). At the same time, the Royal College of Psychiatrists states that ‘evaluation of risk is an inexact actuarial science operating in a political arena’ (Morgan 2007: 5). This then raises questions about the validity, as well as reliability, of risk assessments in the social care field – and particularly within the child protection arena as generated by central government mandatory guidance – which are required to be acted upon by the local authority agencies which have primary responsibility for child protection in the UK (HM Government 2006).

The article goes on to examine the fear which affects frontline workers, managers and child protection agencies as a result of their perception of the operationalization of risk strategies. This is approached from two interlinked perspectives. First, in relation to unrealistic expectations of prediction of risk by use of current risk assessment tools, and the subsequent fear for agencies of being subjected to damaging publicity in the media for a child abuse death. Second, from the perspective of individual workers and their concerns at being judged by the media and by politicians as culpable for such children's deaths by not assessing risk well enough, in an area of work where they are already often subjected to threats and stress from parent service users (Ayre 2001; Littlechild 2005a). Such unrealistic expectations can produce fear in workers of getting things ‘wrong’ in child protection work (now officially known as ‘safeguarding children’–HM Government 2006), arising from methods of risk assessment which are based upon positivistic approaches, and the unobtainable certainties within modernity which are sought in such approaches, which, it will be argued, can provide only indicative areas for potential risks. Essentially, central government appears to have taken a view that in order to reduce the risk of child abuse deaths, the production of mandatory guidance and checklists for professionals will ensure that agencies and professionals carry out risk assessments and plan their work in standardized ways, and therefore reduce the risk to children, and risk to the government of negative and critical publicity. However, such controlling policies from central government can lead to fear and anxiety in social work professionals of not assessing, and eliminating, risk, as the government and their employing agencies are expecting them to do. Regulatory guidance can be seen to place social workers ‘between a rock and a hard place’, with a fear of not meeting all of the many government requirements set out in the regulations, while at the same time often being personally threatened by families where abuse of children is taking place. This last point – violence and threats from parent service users where their children are being investigated for possible child abuse – leads into the second main theme of the article. Such experiences of workers are ignored within media accounts and official government risk assessment policies and guidelines, and yet are known to have been a major feature in a high number of child abuse deaths where children have been known to the protection agencies (Department of Health 1988; Humphreys 1999; Littlechild 2005a, 2005b; Stanley and Goddard 2002).

The tensions, dichotomies and omissions in relation to the assessment of risk in current ‘official’ policy discourses are discussed, and how these impact upon fears produced in child protection work. In essence, it is argued that it is necessary to examine workers’ and agencies’ fears of risk, in part by the use of qualitative research methods, to understand all the relevant areas of risk in child protection work, and in turn, the risks engendered within child protection work of those fears.

Child Protection Policies, Risk and Government Fears

The main aim of the child protection system in the UK is to protect children from harm and to enhance their well-being (HM Government 2006). The discourses and political realities which have determined how we view children, and therefore how risks to them are viewed, have changed dramatically in the last century and a half. From a time in which children were seen as young adults, with no special protection from the demands of the labour market, from sexual abuse, or from abuse within families or in substitute care, the rapidly developing discourse has moved on to how children are developed and protected in terms of work, in terms of compulsory education and, most significantly for our purposes here, in relation to ideas about protection in the twentieth century from what now has come to be seen as abuse (Platt 1969).

Policies relating to the perceived need to protect children on an individual basis became more evident after the Dennis O’Neill inquiry in 1948, following the gross neglect and death of this young boy following his evacuation as a result of the Second World War to an unregulated foster family, and the public outcry and government inquiry following these events. This public inquiry was the first to be set up by the government, and was a mechanism reintroduced in the 1960s, which continues through to the present (Corby et al. 1998).

The next phase of development in policy on abuse relating to physical harm and neglect came about from the work of the National Society for the Protection of Children (NSPCC) and its promotion in the 1960s of the work of two paediatricians from the United States, the Kempes. Following the groundbreaking work of the Kempes in persuading the legal and social welfare authorities in the USA that parents could physically harm their children, the NSPCC brought these two doctors to the UK to develop policies in a similar vein. Their approach derived from a medical model of diagnosis concerning diagnosis of deliberately broken limbs, and led to child protection registers being presented as the main policy directive in relation to child protection from government at that time. The categories of abuse were extended and developed as time progressed, to include emotional and sexual abuse. The most recent addition to this list of types of abuse has been recognition of situations within families where children are emotionally abused, due to their living in an environment of domestic violence (HM Government 2006; Humphreys and Stanley 2006; Children Act 2004). The central overarching legal concept in relation to the abuse and protection of individual children is that of significant harm or the likelihood of significant harm as set out in the Children Act 1989.

The importance of these developments for the purpose of the current discussion relates to the socially constructed nature of child abuse, as acknowledged by the government in the publication of its document, Child Protection: Messages from Research (Department of Health 1995), and therefore the difficulties in agreeing what constitutes abuse. This document recognized that child abuse is more like pornography than whooping cough, with the consequent difficulties of determining within a socially constructed perspective what it is and how to respond to it, rather than the medically oriented model of diagnosis, and treatment, framed within a positivistic review of knowledge and intervention. Such recognition, it could be argued, might have led to the media and politicians understanding that it is not always possible to accurately predict, diagnose and therefore treat child abuse in a way which can lead to risk elimination; however, this is not the case (Ayre 2001; Cooper et al. 2003; Bostock et al. 2005). These formulations affect risk assessments in child protection work carried out within an environment where agency managers and professionals perceive that if they do not eliminate risk, they will be open to severe criticism and personal abuse, both from within their agencies, and from the media.

The Background to the Risk Culture

The development of concepts of ‘risk’ and of the ‘risk society’ (Giddens 1990, 1991; Beck 1992) have led to suggestions that the lack of trust in institutions and professionals has meant an increasing tendency to regulate professionals and their decision-making in order to determine and minimize risk. Parton (1997) argues that the problem with current risk assessment approaches is that they lead to an environment in which social workers and their employing agencies will tend to make defensible rather than ‘right’ decisions – or avoid taking ‘risky’ decisions at all. Such an approach can mean that social workers are not encouraged – or even allowed – to take risks, as a result of a negative view of risk-taking within an agency (Stalker 2003).

One of the problems for child protection social work in recent decades is that it has been positively vilified by the media and politicians (Corby et al. 1998; Ayre 2001). In addition, it can be argued that the Labour governments since 1997 have been fearful of the publicity arising from further child abuse deaths, and have taken the view that more regulations and policies over the local authorities responsible for child protection and over the social workers they employ would reduce the risks of such deaths. However, the corollary of this has been an increase in the fear within social work of being blamed for such deaths. McLaughlin notes how procedural attempts to reduce uncertainty by way of managerialist strategies can be criticized for leading to professionals’ anxiety and lack of confidence in their abilities to assess and make decisions, as ‘failure to follow the correct procedure can leave the worker vulnerable to disciplinary or judicial action if things go wrong’ (McLaughlin 2007: 1264).

Aldgate et al. (2007) argue that this state of affairs arises from the new managerialism promoted by New Labour, which consists of a controlling approach of technocratic micro-practice with an overwhelming focus on outcomes. The lack of trust in social work and social workers has been clearly evidenced by the increasing tendency of government to issue reductivist checklists for social workers to follow, a trend followed by local authorities employing social workers, despite the possibility of misplaced trust in the efficacy of such checklists, which are often at best incomplete in relation to potential confounding factors, for example in relation to the risks arising from violence against staff (Littlechild 2005a).

Within the burgeoning number of policy directives and guidance documents from central government in the child protection area, intended to reduce the risk of child abuse deaths, the first attempt at a risk assessment tool came in 1988 in the document known as the ‘Orange Book’ (Department of Health 1988). Within this guidance, there was only one brief reference to the nature and effects of violence from parents/carers in child protection work, which was removed in its replacement, the Framework for Assessment (Department of Health 2000). Understanding and acknowledging within policies and risk assessments the effects of fear produced by violence from parents/carers is returned to as a key area in relation to risks for such children later in this article.

In relation to such managerialist approaches which promote such risk assessment paradigms, Parton and O’Byrne state:

Social work . . . has become very defensive, overly proceduralised and narrowly concerned with assessing, managing, insuring against risk . . . [and] during the 1990s we saw the introduction of sophisticated attempts to make social workers accountable for, and subject their practice to, ever more detailed reviews, inspections, audits and managerial oversight and prescription. (Parton and O’Byrne 2000: 1)

Such managerial prescription and control based on government concerns are evident in relation to another policy area relating to children and young people – the youth justice field. This was illustrated following the murder of 2-year-old James Bulger by two 10-year-old boys, Jon Venables and Robert Thompson, and their subsequent trial. A media and political outcry in relation to these two young people led to calls for dire punishments for them based upon their being labelled as evil, and for policy changes towards a more punitive system of dealing with young offenders, resulting from ‘the process [of] a politics of fear, even of “child hatred” . . . concerned simply to demonize, promote hostility and pursue the politics of vengeance’ (Muncie and Hughes 2002: 12). A number of the features of the trial were criticized by the European Court of Human Rights, including the actions of the then Home Secretary, Michael Howard, who announced that the two boys would be kept in custody for a minimum of 15 years, in response to the media outcry following the murder. In 1997, the Court of Appeal ruled that this decision was unlawful, and the Home Secretary lost the power to set minimum terms for life-sentence prisoners under the age of 18 years. The High Court and European Court of Human Rights have since ruled that politicians can no longer decide how long a life-sentence prisoner can remain behind bars.

While this punitive turn of events started under the Conservative government of John Major in the 1990s, it was the subsequent Labour governments which delivered punitive policies, leading to the abolition of doli incapax in the Crime and Disorder Act 1998, as a direct result of the Venables and Thompson trials (see also Denney, this issue). Doli incapax had at the trial required the prosecution in criminal proceedings to prove that a young person aged between 10 and 14 knew that what s/he was doing was criminally wrong. This did not fit with the New Labour idea of holding young people criminally responsible for their actions in ways not seen in official policies for nearly a century. Such ideas of central state control as responses to media outcries have been evident in both child protection and youth offending policies.

Hetherington et al. (1997) in their study of child protection systems in Europe concluded that the highly bureaucratic, centrally controlled and proceduralized systems in the UK could not have evolved in the same way in any other European country. In youth justice the government created the Youth Justice Board in the 1990s. The Board has control over the assessments and decisions to be made by youth justice professionals by way of the ASSET checklist. The process is the same – the government wishes to eliminate risk of a ‘problem’ occurring, and reacts by attempting to control agencies and the professionals involved. This is achieved through greater direction over, for example, the judges and professionals involved in youth justice, and the social workers involved in child protection. The process is built upon fear, desire to control, and regulation, with a key element of such regulation being central government's construction, and requirement for use, of assessment tools, as evidenced in the youth justice and child protection fields in recent years.

Risk Assessment Methodologies and Tools: Reliable and Valid?

Developments in rationality and scientific endeavour are the basis and justification for current discourses which underlie contemporary risk-assessment and risk-management strategies. Kronenfeld and Glik (1991) saw the perception of risk in the medical sociology field as reflecting the shift in people's thought processes away from an emphasis on fate or luck, to notions of control. Scientific rationalism has become indisputably the dominant paradigm for explaining and predicting events in the social world. The risk assessment enterprise is based on the premise that we can know the world, and that we can determine cause and effect from observation of events within a positivist paradigm. It is these notions of predictability and control which become so important in risk assessment and decision-making. No longer, then, do accidents, incidents or tragedies just happen; they are seen to be predictable, assessable and preventable (Littlechild 2004). One key area of risk assessment within this paradigm is the knowledge arising from previous events, often predicated upon actuarial assessments.

Actuarially Based Assessments of Risk

Actuarial methods are the basis upon which many risk assessment strategies are based. It is used in many areas of assessing risk in business; so, for example, insurance companies predict the risk to a particular individual or situation, based on information they have built up in relation to a certain activity. Thus, in relation to driving a car, for example, risks to different age groups, dependent upon where they live, previous offences and accidents, etc., are collected and analysed over time. However, actuarial approaches do not try to predict or manage risk to individual drivers; they balance the risk to their business having to settle claims from drivers by assessing the likelihood of the number of individuals from within certain groups making claims on their insurance. What they do not try to do is to predict which individual, over a period of time, will make that claim on their insurance. In health and social care professions, this is what it appears that agencies and professionals are expected to do – to predict which individual will act in which ways over a given period of time – a process termed by Fitzgibbon as the ‘actuarial fallacy’ (Fitzgibbon 2007a). Such methods in relation to the social world are, however, not able to take into account the many and complex areas which intertwine to determine how a particular person will respond on a particular day, to a particular set of circumstances (Titterton 2005). In addition, Higgins et al. (2005) note that actuarial methods are unlikely to be of use in a population with low base rates, e.g. of certain types of violence, as such methods lead to large numbers of false positives being generated. This is the case for the child abuse and child protection fields.

Fitzgibbon proposes that the use of statistics as the basis of actuarial risk assessments can be misleading; so, for example:

A risk assessment which always predicts a lower likelihood of harm could be interpreted as being 95% accurate even if 5% of subjects assessed go on to cause future harm. (Fitzgibbon 2007b: 137)

In addition, it is very difficult to measure retrospectively the effectiveness of risk assessment models and tools. One recent media-defined ‘scandal’ concerned prisoners released from prison who had subsequently carried out serious offences after being risk-assessed. It was claimed that the Probation Service had assessed a 91 per cent chance of reoffending by the prisoner (Guardian 2006). However, it is not possible to test empirically the validity of such assessments/predictions, as it can never be known if it was an accurate prediction or not. Positivistic models arising from the importation of models of knowledge and prediction from the natural sciences, then, cannot be easily transplanted into the human sciences in areas such as offending or child protection.

The Place of Non-positivistic Knowledge

Government policy and guidance do not acknowledge that risk assessment policies and tools derived from positivist technical–rational approaches fail to appreciate the limitations on how possible it is to be prescriptive in relation to risk assessment. Such processes, which provide guidance for professionals, require to be based on a systematic review of the evidence, including untidy and distinctly non-positivistic sets of knowledge concerning how people construct their knowledge of the area they are involved in, and their attributions of the world and the motives of others. In the particular case of the child protection social work field, this includes how social workers in reality make their assessment and decisions as a result partly of power dynamics within certain types of abusing families, and threats of intimidation and violence against them by parents in such families. These areas are ignored in the Department of Health Assessment Framework (Department of Health 2000), and its successor, the Common Assessment Framework. These assessment frameworks are required to be used by local authorities, in relation to child well-being and protection. If such experiences and attributions are not included in decision-making based upon risk assessments in social work, it then becomes impossible to include guidance on these areas which affect the safety of workers, and the well-being of children with whom they work (Littlechild 2004, 2005a).

Goddard et al. (1999) present a number of criticisms of risk assessment procedures which they see encroaching upon professional social work assessments in child protection work in Australia, New Zealand, Canada and the USA, among other countries. They argue that assessments that do not recognize uncertainty cannot accurately reflect the complex set of factors which might present risks to children, and are no longer acceptable. Parton (1998) similarly argues that social work must move beyond such restricted and restrictive risk assessments, and rediscover ambiguity and uncertainty, as he terms it – i.e. that in the complex set of factors which drive human cognition, motivation and behaviour, we are often not able to predict risks fully, even with a wide range of actuarial information concerning risk factors.

Given these uncertainties, it would not appear to be in the clients’ or workers’ interests to make use of mechanistic and potentially problematic solutions, yet for the employing, responsible agency, the attractions of having a mechanistic system that will ‘hold up’ in court if sued for their actions is attractive (Carson 1996). This in turn may possibly lead to individual workers being reluctant to assess a case as ‘low-risk’, as this may create high risk for themselves individually as professionals. This then could lead to coercive intervention which is not appropriate for the needs and rights of that child, or family. The fear created by a system which expects risk elimination through increasingly centralized guidance – essentially checklists – cannot help professionals or clients if it induces fear.

Policy Changes from Individual Child Abuse Death Inquiries

In contrast to actuarial approaches, the other major influence upon child protection risk assessment policies is the findings of child abuse death inquiry reports, whereby policies in relation to child protection were significantly affected by the results of such inquiries based on actions or omissions, as found in one single case (Corby et al. 1998). There are significant problematic effects of trying to base child protection work and the risks within it upon single incidents, as so often happens after tragedies have occurred. Such responses can mean that the risk factors identified in that particular case have a disproportionate effect upon policies and practice, focusing only on the issues raised in that inquiry, and excluding more important general risks for the population served (see e.g. Butler and Drakeford 2003). Examining ‘mistakes’ retrospectively when they have led to tragedy is not necessarily a good way to assess and deal with risks within a particular area (Cooper et al. 2003; Bostock et al. 2005).

The Effects of Current Expectations of Risk Assessment Work on Child Protection Professionals: A Culture of Fear

While there is a need to ensure that social workers and their agencies are acting appropriately and without negligence in their work, the idea that risk can be eliminated is unrealistic and problematic for services, children and frontline workers. The fear of risk as set out in previous sections has to be confronted by social workers and their agencies in order to improve their services. This can only occur within a culture which accepts that such agencies and workers cannot always get every decision ‘right’, while acknowledging that workers and agencies alike need to take into account appropriately lessons from research which can aid ways in which risk is assessed and worked with. This recognition can then lead to other, more constructive ways of conceptualizing how social workers and social work agencies deal with risk. This can be seen in some elements of the health service delivery, and more recently considered for social work and social care: learning from ‘mistakes’ or ‘near misses’, as set out in the Social Care Institute for Excellence (SCIE) position document, Managing Risks and Minimizing Mistakes in Services to Children and Families by Bostock et al. (2005; see In the airline industry, there has been a realization that lessons can be learned from near misses, which has led to new reporting procedures. This has not occurred in social work and social care. Bostock et al. (2005) argue that mistakes happen in all forms of human endeavour, and what we should be attempting to do is to learn from mistakes. If we blame staff for what happens, and make them fearful of reporting difficulties, the reality of the problems can neither be systematically examined, nor action taken to remedy them (Kemshall and Pritchard 1996). Workers can become fearful of their work, of their clients, and of reporting their concerns, in case they are seen as too demanding, as troublemakers, or nervous and/or ineffective workers (Rowett 1986; Norris 1990). Such a culture of blame is unhelpful for agencies, workers, and clients (see Bostock et al. 2005). In a report for the independent political think tank, Demos, Cooper et al. (2003) have argued that the current trend of regulating social workers more and more by way of the use of such mechanisms as the previously mentioned Assessment Framework, may create greater risks rather than reducing risks for children. They propose that politicians and managers have to allow professionals to have greater space within which to make judgements, and for them not to be blamed if a child's abuse is not due to gross neglect of their duties.

The Effects of Fear on Child Protection Risk Assessments

One key area in relation to fear in child protection work relates to the impact of violent parent service users. This appears to have been ignored in policy and procedures. In order to be able to have a more comprehensive and effective assessment of risk in child protection policies, procedures and risk assessment tools, government bodies need to ensure that research and developments in these areas include the reality of the experiences of the workers involved. This is particularly important as we know that social workers construct their own realities and attributions within their work, which lead to actions which are not always foreseen by policy-makers and higher-level managers. Professionals can become ‘street-level bureaucrats’, actively changing policy goals in relation to their own beliefs and experiences (Gelsthorpe and Padfield 2003; Evans and Harris 2004). A vital element of any evidence base is knowledge and consideration of how social workers perceive their world of work and their professional agency within it. Yet this has not been a feature of any statements or assessment of risk in relation to child protection work in any of the government publications or guidance since the Orange Book document (Department of Health 1988). This is despite the range and depth of evidence demonstrating the extent and the effects of such violence and aggression in social work and social care in general, and serious child abuse situations in particular.

Research by Pahl (1999) on stress in workers in social services departments discovered that violence and threats of violence to social workers were one of the major areas of stress and fear for social work and social care staff, and particularly child care and child protection staff (see also Smith and Nursten 1998). Concerns about how violence from service users can negatively affect child protection assessments and decision-making processes have been raised by, for example, Reder et al. (1993), Farmer and Owen (1998), O’Hagan and Dillenburger (1995), and Stanley and Goddard (2002). Analysis of a range of child abuse death inquiry reports in the United Kingdom has highlighted how assessment, intervention and decision-making in child protection can be influenced by workers’ concerns about clients’ aggression in a small but critical number of threatening and violent situations. These sets of features are often present in the most severe forms of abuse, including child deaths (Department of Health 1991; Guardian 2002). This significantly affects workers’ well-being, capacity to carry out their work effectively, and their commitment to that work (Norris 1990; Brockmann 2002).

What these studies also demonstrated, along with the findings of Norris (1990), was that a high proportion of incidents of threats and intimidation is not reported by social workers within their agencies. This means that the extent of the problem, and the precise nature and effects of it, cannot be monitored, evaluated or dealt with. Workers, because of their fears and anxieties, can fail to recognize the threats against them, and might not believe that reporting them would mean that they were supported, or that the matter would be dealt with satisfactorily by their agency – leaving themselves and their child service users at risk (Littlechild 2005a, 2005b).

Conclusion: Risks of Risk Assessments

This article has examined how risk of harm in child protection work is currently constructed in government documents and child protection agencies. It has been argued that it is necessary to examine the risks of risk assessments, as the centrally developed risk assessment agenda and its associated tools have (probably) inadvertently induced fearful perceptions in social workers. This is due to their concerns about the unrealistic expectation that they can, by the use of such tools, eliminate risk. At the same time other risks, such as violence from service users, are ignored in risk assessment tools. Current formulations of positivistic risk assessment approaches are based upon fears of central government in wishing to try to eliminate risk in areas in which they are seen to have responsibility, such as child protection, by way of controlling guidance and regulation. This then leads to unrealistic expectations of centrally formulated risk assessment within the social work and social care field, while at the same time government has clear expectations that social workers, if only they apply them properly, would be able to avoid child abuse deaths. This article has set out the evidence as to why this cannot be the case in terms of actuarial and positivistic-based approaches, and issues arising within single child abuse situations. It has also been argued that there are areas of risks for children which are ignored within official policy and guidance, which have to be confronted in order to protect children in ways which currently formulated positivist approaches of risk assessment tools have not achieved.

When we come to consider ways in which we need to take into account knowledge to produce effective policies to reduce risk to children, government needs to include research into how professionals make their decisions and why – otherwise there will always be unintended consequences. This is necessary to understand the processes which professionals go through, and the pressures and influences on them, while also gaining their commitment to approaches which reflect the reality of the situations they are put into by government policies themselves. In the human sciences, it is dangerous to assume that professionals are not human beings who construct their own worlds, methods of working, views about their employing agencies, and attributions about service users and society. This then requires the incorporation of knowledge about how professionals can be engaged to produce the desired policy outcomes. In furtherance of this, the work of Ruch (2007) examines issues of managerialism, supervision and child protection cultures, and the problems associated with them. Ruch concluded that there is a need to produce a model for the development of policies, procedures and support mechanisms for staff that includes the views and experiences of both service users and professionals to aid the pursuance of policy goals. This is undoubtedly the case in relation to child protection as examined in this article. At the same time, the results of systematic analysis of risk factors, which take into account the reality of the untidy worlds of human beings which do not fit easily with positivistic paradigms, can provide guidance for professionals to consider when carrying out their assessments – and for agencies to support them in this. Canton (2005) notes that while it may be that risk assessment tools can help frontline workers take into account the different types of risk factors, it will always be a professional judgement which will have to be applied to these areas of risk, to the particular situation and to the family and child with whom they are working, an approach advocated by the Social Care Institute for Excellence (Bostock et al. 2005), and the Demos reports (Cooper et al. 2003), referred to earlier.

Anxiety related to mechanistic risk assessment tools has increased fear in social workers, while the tools themselves do not contain all the necessary elements which pose dangers to children. For enhanced, effective risk assessments, government policies and guidance need to take into account different methods of understanding how professionals approach their work, and are affected by risk factors currently ignored in policy and guidance. The fear of reporting and not receiving support in relation to intimidation, harassment and aggression from parents/carers of children where there are investigations concerning abuse is an important example of this. Government guidelines also need to allow professionals, where they can justify this, to move beyond restrictive checklists. Such a change in culture needs to take into account the effects of the work, and the power dynamics within it, upon the workers themselves. This requires qualitative research to add to the mix of areas to be taken into account in order to develop effective policy and practice. It is proposed here that there are alternative means by which government objectives in terms of reducing risk to children can take account of research and theory which are more likely to produce the desired outcomes than the current formulations. Government policies need to encourage local authorities to allow social workers to address their various fears openly with their managers in a culture of support, rather than a culture of blame in relation to decision-making.