Towards geographies of child protection

The emergence of current and historic cases of child abuse across the globe has, in recent years, dominated the news, political agendas and popular discourse surrounding children. From serious case reviews to exploitation in post ‐ conflict zones, from sexual abuse of children by groups to trafficking of drugs across countries, the importance of protecting children is an increasing concern in many countries. Key to, and inherent in, all of these processes and phenomena are child protection systems, working in varying degrees of effectiveness. While geographic interest has touched upon many of these areas, the role of child protection systems, and the practitioners that work within these, do not explicitly feature within this work. In this article, we seek to develop an introduction to geographies of child protection, producing an initial critical review which points to future research avenues in this field. We adopt a Foucauldian approach and review four themes to illustrate the ways in which geographical approaches might yield important insights. Drawing primarily on England as a context, we consider the historical geographies and origins of child protection, relational practices in contemporary child protection, the impact of austerity and finally we consider what future directions might require a geographical approach.


| INTRODUCTION
The emergence of current and historic cases of child abuse across the globe has, in recent years, dominated the news, political agendas and popular discourse surrounding children. From serious case reviews to exploitation in post-conflict zones (Seddighi, Salmani, Javadi, & Seddighi, 2019), from sexual abuse of children by groups to trafficking of drugs across countries (Blazek & Esson, 2019), the importance of protecting children is an increasing concern in many countries. Key to, and inherent in, all of these processes and phenomena are child protection systems, working in varying degrees of effectiveness. While geographic interest has touched upon many of these areas, even specific engagement with social work (see Andrews, 2020), the role of child protection systems, and the practitioners that work within these, do not explicitly feature within this work. Child protection work involves spatial activities at a number of scales (Jeyasingham, 2014). Internationally, researchers outside of geography are recognising the diverse spatialisation of children's experiences of harm and abuse, including the spatially contingent nature of abuse (Firmin, 2019), but also the spatial practices of social work systems themselves, such as the mobility of practitioners and functions of child protection (Jeyasingham, 2018). This article explores geographical approaches to child protection social work, in response to nascent work by geographers engaging in this field Lloyd, 2019).
The absence of geographers from discussion of child protection is curious given the focus of such systems on childhood risk, a key facet of early work by children's geographies (Valentine, 1997). Similarly, these systems can reflect and are interconnected with other systems that have preoccupied geographers of late, such as closed or carceral networks that families, children and young people experience (Schliehe, 2014). Furthermore, while geographers have often not engaged with child protection practices in their writing, many will have reflected upon processes of safeguarding children when proposing research with children and young people through university ethics processes. The powerful resonance of child protection with geographical scholarship underlines the potential value of geographers to this area and the contribution they can make to these processes. To paraphrase Willis, Canavan, and Prior (2015), child protection is an absent presence in geography, something we intend to address in this article.
In this article, we seek to develop an introduction to the geographies of child protection, producing an initial critical review which points to future research avenues in this field. We begin by providing an explanatory overview of the operation of contemporary child protection systems in England, which geographers may not be familiar with.
In this article, we focus primarily on child protection practice in England specifically. There exists a tension between a global framework (notably the UN Convention on the Rights of the Child (UNCRC, 1989), in particular article 19: protection from violence, abuse and neglect) and the significant diversity in the forms of child protection systems across the world. While these ostensibly share similar aims, there are important cultural and social specificities which shape the varying typologies of systems in particular contexts (Gilbert, Parton, & Skivenes, 2011;Hearn, Pösö, Smith, White, & Korpinen, 2004). We suggest, therefore, there is value in considering a particular country context to understand how these systems operate. As we note, however, future engagements should move beyond this initial review to consider other contexts.
We adopt a Foucauldian approach to child protection systems and practices, specifically conceptualising these as a dispositif (see Gallagher, 2019); a constellation of web-like institutional, forms of knowledge, laws and discourse. Positioning child protection as a dispositif, we then discuss four themes to illustrate the ways in which geographical approaches might yield important insights. These include exploring the historical geographies and origins of child protection, relational practices in contemporary child protection, the impact of austerity and finally we consider what future directions might require a geographical approach.
There is much of value to geographers in engaging with child protection practice. As these systems reveal, the state seeks to produce citizens, shaping and (re)making childhoods and families in the process. We argue that geographers should be engaging in dialogue with social work, to map the spatialities of child protection practice.

| CONTEMPORARY CHILD PRACTICE IN ENGLAND
While child protection (the activity of protecting children) is inherent in multiple institutions, we focus specifically on the role of social work. In England, as in many countries internationally, contemporary child protection systems have been developed to protect children from abuse. The state's role in responding to harm is enshrined in the framework of The Children Act (1989). This is underpinned by the 'paramountcy principle' (the welfare of the child is considered paramount), partnership with parents and interagency cooperation. This legislation provides the underlying legal framework for protecting children from harm but is interpreted and implemented in combination with regularly updated statutory guidance (such as Working Together, 2018). The act covers many aspects of child and family social work practice in England.
Under Section 17 of the act, a 'child in need' is understood as a child who needs support in order to safeguard and promote their welfare through the provision of services. Children deemed to be experiencing or at risk of 'significant harm' receive statutory intervention under Section 47 of the Children Act (1989), whereby social workers conduct further enquires and assessments to determine the level of harm present and the extent to which a child's welfare is impacted. If it is determined that the child is at risk of significant harm then they may be subject to a child protection plan, which broadly outlines how a child can be kept safe and the support needed to do so. If, following a plan the risk does not reduce, children can-as a last resort-be taken into the care of the state.
While The Children Act (1989) underpins the legal framework for intervention and support from the state, thresholds used individually by children's services across England vary significantly and have been subject to mounting critique in recent years (Platt & Turney, 2014). Furthermore, a number of high-profile cases have highlighted challenges in the structure and focus of child protection systems themselves (Munro, 2011). For example, while the Children Act requires harm to be attributable to a lack of parental care or control, children are increasingly at risk of significant harm outside of their families in places and spaces beyond the home (Lloyd & Firmin, 2019). We argue that such critiques highlight the complex spatialities of child protection and sometimes contradictory operation of various structures aiming to keep children safe.
In contrast to recent post-human approaches to social work in geography (Andrews, 2020), we draw upon a Foucauldian approach which emphasises the organising and disciplinary functions of child protection social work but also encompasses the sometimes messy, contradictory discourses and processes inherent in these systems. We now explain how conceptualising the operation of these structures, systems and practices as a dispositif can help to shed light on these processes.

| CONCEPTUALISING CHILD PROTECTION AS DISPOSITIF
Despite the importance of the concept of dispositif in Foucault's work, it has sometimes been neglected in social research (Raffnsøe, Gudmand-Høyer, & Thaning, 2014), remaining a somewhat opaque concept (see Hardy, 2015). However, it has been noted as one of Foucault's 'most powerful conceptual tools' (Rabinow & Rose, 2003: xv). It can be understood as web-like institutional, administrative mechanisms and forms of knowledge, responding to a specific urgent need to shape, maintain and enhance the social body. As Frost (2019, p. 159) notes, the dispositif can be understood as a 'network of relations between elements that responds to an emergency and which organises, enables, orients fixes and blocks relations of force'. The translation of dispositif has been the subject of significant debate (see Bussolini, 2010). Although the term 'apparatus' is sometimes used, it is distinguished from the more standard 'assemblage-like' apparatus (ibid, 2019). Foucault explains the dispositif in the following excerpt:  Secondly, what I am trying to identify in this apparatus is precisely the nature of the connection that can exist between these heterogeneous elements. Thirdly, I understand by the term "apparatus" a sort of -shall we say -formation which has as its major function at a given historical moment that of responding to an urgent need. (Foucault, 1980, p. 194-195) Examples of dispositif might be the police or army, and the mechanisms, laws and discourses that constitute these institutions, not acting singularly but as a constellation of different processes of actors and structures, responding to emergencies. Understanding child protection as a constellation of different processes is useful because it captures the inherent messiness of these practices and, at times, the disagreement or contradictory approaches employed by various agencies.
People are organised and constituted as subjects by power relations and dispositifs. While the dispositif is a potentially productive force, organising and creating subjectivities, 'the nature of biopolitics means that those selfsame dispositifs also have a negative side, in that they control and order which lives are worth preserving and which are not' (Frost, 2019, p. 160).
Child protection systems are then a dispositif (see Gallagher, 2019); a complex network of actors, non-humans, discourses, and legal and administrative processes. These systems have emerged from complex moral and philanthropic propositions but with significant elements of control embedded within them. Child protection is often understood in individualised terms or decontextualized from wider systems and places. Conceptualising it as a dispositif helps to reveal the wider geographies inherent in these processes. For example, harm is interpreted by actors within a child protection system (such as social workers and family court judges). Yet the families and children which become subject to scrutiny (through child and family assessments and child protection plans and court hearings) and regulated through powerful sanctions and discourses are not independent of wider systems and structures, such as austerity and racism and at the same time, the response that a family can expect varies widely depending on geographic region and political sensitivities (House of Commons, 2012) as opposed to the type of harm a child is experiencing.
Foucault's emphasis on the importance of responding to an emergency situation is also critical in understanding child protection systems as a dispositif, necessitating the mobilisation of diverse interventions through institutions such as courts and actors such as social workers to protect children 'at risk' of harm. The emergence of these processes also has a complex and at times contradictory history.

| HISTORICAL GEOGRAPHIES OF CHILD PROTECTION
There are particular historical geographies to child protection, encompassed broadly within children's social work, and relevant to the spatialisation of its practice. In particular, there has existed a tension between more individualised practice with a focus on maltreatment, and then an attention to broader, contextual issues around child protection, in the form of more community-based models.
Although many would trace the establishment of child protection practice in England to the 1889 Prevention of Cruelty to, and Protection of, Children Act, these practices have a longer history, shaped by competing and paradoxical discourses around childhood, nation and class (Sales, 2012). The origins of formal child protection practice in England can be traced to the latter decades of the 19th century (Ferguson, 2004), in particular the operations of the Charity Organisation Society (COS). In her genealogy of adoption, Sales (2012) argues that child protection campaigners were initially reluctant to intervene in what was seen to be the private, bounded space of the Victorian family. As noted in the 1869 Poor Law: It is most important on all grounds to avoid severing or weakening in any way the ties of the family, even where, owing to the character of the parents, it be thought that children could be benefited by removal from their control. (George, 1970, p. 26, cited in Sales, 2012 Child protection practitioners concerned themselves with children outside the family, most commonly street children engaged in begging or other activities conceptualised as a threat to moral order. There was a distinctly spatial public-private binary in this approach to protection procedures, with children physically outside the family, on the street, understood as those in need of intervention. Alongside this public dimension, child protection was particularly targeted at the urban working classes; Foucault argued that in the 1830s, the 'conventional' family emerged as 'an indispensable instrument of political control and economic regulation for the subjugation of the urban proletariat: there was a great campaign for the 'moralisation of the working classes'' (Foucault, 1978, p. 122.) Sales (2012) notes that a paradox emerged here: the family was understood as both a site of salvation and as an institution through which problems could emerge. She argues that although family was initially understood as a site of salvation, by the latter stages of the 19th century, it became a site of potential pathology and dysfunction, moving from a model for government to an instrument of government (Sales, 2012, p. 31). The spatial boundaries of intervention began to change; with bourgeois family as the assumed norm of family functioning, and spurred on by theories of biological determinism fuelling concerns around the health of nation, the practices of child protection shifted from outside the family on the streets of the slum, to inside the homes the families of the slum.
Intervention measures into the lives of the poor crystallised in these latter stages of the 19th century, shaped by notions of the working class as 'other' and fear of the transmission of parental vices; while early discourses of attachment and separation were slowly emerging in relation to bourgeois families, working class families were characterised more broadly by child protection practices where 'the breaking of parental relation was accorded little psychological significance' (Sales, 2012, p. 34). Working class families were different and troubling, identified as objects of concern with different morals who could potentially risk the health of the future nation (Ferguson, 2004). Swain and Hillel (2010, p. 64)  While there were decontextualised approaches to child protection practice, there existed a tension in Victorian England between those who understood neglect in individualised and relational terms, and those who pursed a more community-based model of social work. Alternative approaches emerged through the settlement movements, which attempted to collapse the spatial divide between rich and poor. In this approach, the context of poverty, rather than inherent character, became the primary focus (Pierson, 2011). This community-based model, exemplified by university-affiliated institutions such as Toynbee Hall in the East End of London, embedded social workers in areas of need, offering education and forms of family support. In contrast to more individualised approach, such as those of the COS, the settlement movements engaged and collaborated with a wider network of services (ibid).
This competing approach is akin to the community development model of social welfare practice with a particular focus on neighbourhoods and context, representing a more (implicit) geographical approach to child welfare. These contradictory approaches are emblematic of the constellation of different processes inherent in child protection practice and competing understandings between an attention to wider contexts and more relational approaches. Echoes of these tensions reverberate in contemporary child protection policy in England, and we illustrate this with a focus on child protection relationalities, austerity and a discussion of future directions in child protection. DISNEY AND LLOYD -5 of 13

| CHILD PROTECTION RELATIONALITIES
A key facet of contemporary child protection practice has been a focus on relationships, both through its practice, and the target of, that practice. First, through the relationships between social workers and those they work with which is 'characterised by an emotionally close and informal style of professional relationship, compared to an emotionally distant one' (Reimer, 2017, p. 8). Second, within children's social work, familial relations themselves can be seen to be the target of some social work interventions. For example, working with families to improve relationships through interventions such as parenting support and guidance. This focus is termed 'relationshipbased practice' and is presented as a means of humanising what are often widely described as increasingly bureaucraticised systems (Broadhurst et al., 2009;Munro, 2005).
While relationship-based practice may be considered the heart of social work (Ruch et al., 2018), such relationships are often seemingly divorced from wider macro-scale processes and spatial scales within social work practice itself. While social workers have the potential to exert agency within wider networks of the dispositif that comprise child protection practice, as Featherstone et al. (2018) suggest, systems in England have for some time been preoccupied detecting abuse rather than offering support and forming relationships. Rather than considering relationships as part of, and formed through, relations at a range of scales, practice can be considered to be increasingly individualised-focussed upon deficits within families, as opposed to the wider structural issues that contribute to harm . For example, while child neglect is often the result of families living within increasingly deprived circumstances, parents-as opposed to the state-are often blamed for their inability to food and clothe their children. When children are living in situations of domestic violence, it is often the decisions of the mother-whether to leave the abusive partner or not-that are the subject of scrutiny rather than the perpetrator of that harm and the patriarchal power relationships which shape such decisions (Nayak, 2019). In these relationships the child protection professional is cast in the position of powerful expert, with the parent infantilised requiring knowledge to reshape and remake their subjectivities (Parton, 1985). Relational approaches, therefore, often focus on individual responsibility and are either distinctly aspatial or focus on place in reductive or problematic ways.
While geographers have not explicitly engaged with these issues, social work academics have begun to consider the spatiality of social work practice. In exploring social workers' engagement with place through mobile interviews, Jeyasingham (2018, p. 85) notes how practitioners visiting families were 'haunted by past experiences' of neighbourhoods. In the study, social workers problematically imagined particular relations as embedded in local spaces, for example, drug use, despite not being able to provide evidence of these activities. The relationship with place highlighted in Jeyasingham's work demonstrates the circulation of problematic discourses in the dispositif and the influence of them at the micro-scale in practice. These approaches to place can also become mobilised and scaled up into national discourses of child protection, however, which may draw upon problematic notions of how-or where-harm manifests or who perpetrates it. This can be seen in the ways in which certain neighbourhoods become presented as 'troubled', generative of social problems and identified for localised intervention (Crossley, 2018).
New developments within social work are exploring the relationship between the harm young people experience and the place in which occurs-for example, serious youth violence on the street or sexual harassment at school, employing geographical or contextual methods to assess places rather than people (Firmin, 2020;Lloyd, 2019). Yet, in the absence of thorough geographical contribution, social work practice itself can be seen to develop with limited recognition and understanding of the role of space and place. For example, the problematic ways in which the dispositif of child protection systems and networks become orientated towards certain places as 'nested' (after Hall, 2019b) rather than interconnected in, and produced through, other spatial processes and relations.
Employing contributions from geography would allow practitioners to consider how space is constituted through relations. Seeing space not as a container but altered by social relations within it (Anderson, 2008), formed through 6 of 13 -DISNEY AND LLOYD interactions between objects, people and processes, always in a state perpetual becoming, stretched out across a network of social relations (Massey, 1994).
A geographical relational approach when applied to child protection allows us to consider social work in a variety of spatially contingent ways. First, how social relations-such as those of families-are re-defined through social work practice. In what ways does the presence of social workers within the space of the home shift relations within that space and constitute families in relation to absence or presence? Second, by understanding social relations as part of a broader, macro-scale network of social, political and contextual relations. How are familial relations shaped by the politics of those spaces, in what ways do local services or legislation impact families? Third, how are the wider components of the dispositif of child protection, such as laws, policies, institutions and discourses, entangled in producing the spaces of intervention they seek to change? Particularly in response to, and in times of, emergency. This is particularly important when the wider networks of services are reshaped and remade through policies, such as the austerity policies of the past decade, to which we now turn.

| AUSTERITY AND THE RESHAPING OF THE DISPOSTIF
The global financial crisis of 2007-2009 ruptured economic systems across the world, remaking local, regional and national geographies in numerous ways. Following the economic dislocation of the 'crash', certain countries ushered in policies of austerity, which Hall (2017, p. 303) explains 'as the conditions resulting from the cutting of public expenditure by the state to reduce government budget deficit, or as a condition of severe simplicity and selfrestraint, thus having intersecting and multi-scalar meanings'. The United Kingdom has pursued a particularly extensive programme of austerity, which has seen local authority budgets cut further than other areas of government funding (Hastings, Bailey, Bramley, Gannon, & Watkins, 2015).
The impacts have not been even, and there are complicated geographies evident in the implementation of austerity policies across England. Local authorities are responsible for child protection social work, and importantly certain local authorities are more dependent on central government grant and thus vulnerable to cuts. This is particularly true of certain regions still recovering from legacies of deindustrialisation, which are now also seeing the largest reduction in household incomes (Beatty & Fothergill, 2016, p. 3). What is clear is that there is an uneven geography to austerity, which impacts certain regions more strongly than others and certain people more. Austerity policies have been described as leading to a 'perfect storm' of a reduction in benefits, income and insecurity in family budgets, compounded by service withdrawals; far from the intended reduction in child and family poverty, England has seen worrying increases (Ridge, 2013).
Geographers have recently increasingly turned their attention to the micro-scale impacts of austerity and its implications for everyday lives, considering children's and youth service closures (Horton, 2016;Jupp, 2013), inequalities in labour market access (McDowell, 2012), and food banks as liminal community spaces negotiating austerity (Cloke, May, & Williams, 2017). They have noted the ways in which austerity has contributed to a 'shrinking of worlds' and everyday spaces, leading to poorer mental health outcomes (Hitchen & Shaw, 2019), exacerbating loneliness even as the government aims to combat it (Stenning & Hall, 2018). In her work on the everyday, lived, relational experiences of austerity, Hall (2019a) demonstrates how austerity has the potential to redefine kin and familial relationships, and further points how it reverberates through wider social relations such as social care professionals. However, while geographers have explored the shifting landscapes of care following austerity (Power & Hall, 2018), they have largely neglected the impacts of austerity upon child protection systems and policies, and the implications of this for child and family welfare given how it is contributing to rising levels of inequality and poverty. Andrews (2020, p. 11) recently called for consideration of which 'types of political-economic, institutional, family, and socio-cultural processes aid or obstruct … social work assemblages?' The impact of austerity has particular implications for the operation of social work practices, particularly in reference to child protection. Describing poverty as the 'elephant' in the room in cases of child neglect, Gupta (2017) notes that current discourses of child protection largely ignore contextual factors, such as the contribution of austerity to rising poverty levels, and tend to blame both families and social workers for not addressing child neglect and harm. As Jupp (2017) notes, families have become the target of many cuts but also increasingly relied upon in social policy, reflecting the Victorian paradox of the family as a site of both salvation and pathology. Concurrently, child protection practice has become increasingly 'risk averse', with the 'model being promoted centred on the identification of and elimination of risk to individual children with little concept of the relationship to safeguarding to the economic or community context' (Bywaters et al., 2018, p. 54). Bywaters, Brady, Sparks, & Bos's (2016) quantitative analysis of Section 47 investigations (statutory assessments used to identify if a child is experiencing 'significant harm') in deprived areas in England has revealed a troubling relationship between place, poverty and intervention, whereby the areas of significant deprivation are seeing the highest increase in these investigations which can lead to the rupturing of family and perhaps even child removal. What is becoming apparent is that place is a significant determinant of child protection practice, and that there is a hidden geography of poverty and child welfare regulation that demands greater discussion. These geographies are complex and troubling (Disney & Schliehe, 2019); not only are the populations targeted by interventions 'troubled' by the engagements and wider contextual geographies of poverty, but also the institutions seeking to remake these populations are also increasingly themselves troubled by and through cuts and managerialist pressures. Here there is an insight into the ways in which the dispositif of child protection practice orders populations, classifying certain groups as problematic and requiring intervention. But also how the impacts of austerity also highlight the vulnerability of the constellation of practices that comprise the dispositif; while families struggle to negotiate the impacts of austerity and rising poverty, in the wider ecology of child protection services, the diversity of tools available to social workers to engage with families and to support them have been severely reduced. This can be seen through the reductions in third sector partners and complementary government services that child protection social workers could previously rely on to refer to. Austerity has also led to local authorities seeking new ways to practice in order to cope with these pressures.

| ANTICIPATORY GEOGRAPHIES OF CHILD PROTECTION
The multi-layered impacts of austerity will be felt for years to come and seem likely to deepen following the global outbreak of Covid-19. These policies have exacerbated increasing preoccupations of the dispositif with the identification and elimination of risk through early intervention. This agenda has been widely adopted across the political spectrum. Key in such approaches is a notion that early intervention is a cost saving measure designed to ward off future problems and is increasingly 'informed' by neuroscience (see Allen & Duncan Smith, 2009). It is argued that without early intervention to protect very young children, supposedly irreparable damage may be done to their developing brains. Approaches that are seemingly grounded in 'brain science' have had a strong allure for policy makers, but it is important to note that these techniques have been powerfully critiqued within literatures of social work and social policy (Featherstone, Morris, & White, 2014;Wastell & White, 2012), as Featherstone et al.

note:
the original neuroscience literature shows quite remarkable resilience and plasticity when exposed to ordinary patters of 'chaotic' neglect usually seen in the population referred to children's social care … if changes to the brain were criterion for the removal from parents, very few children would be removed. (2014, p. 1739, emphasis in the original) The fusing of problematic interpretations of neuroscientific knowledge with early intervention into child protection practices highlights the desire within the dipositif to eliminate future risk and do this in a fast, cheap and 8 of 13 -DISNEY AND LLOYD effective way. The urgency to do this has been heightened through austerity and has led to the dispositif of child protection practice seeking other cost-effective ways to identify and eliminate future risks to children.
A number of further anticipatory approaches are gaining traction in child protection. The era of 'surveillance capitalism' (Zuboff, 2019) has seen the rise of algorithms for use in child protection practices, with private sector organisations offering big data solutions to the identification of potential child maltreatment. Concerns have been raised around how these techniques over-represent certain sections of populations in their data harvesting. In the United States context of Eubanks' research, she notes that when referrals predominately concern 'child neglect', there are issues to be raised about the ability of such systems to understand context: Where the line is drawn between the routine conditions of poverty and child neglect is particularly vexing.
Many struggles common among poor families are officially defined as child maltreatment, including not having enough food, having inadequate or unsafe housing, lacking medical care, or leaving a child alone while you work. (Eubanks, 2018, p. 130) Eubanks goes on to explain that data is gathered about the use of means-tested public services and interactions with the juvenile justice systems and child protection services, all of which disproportionately represent people in poverty. In contrast, middle-class families are able to avoid their data becoming entangled in such systems meaning that the system over samples poorer families.
The exact use of such systems in England remains somewhat unclear to date. Redden, Dencik, and Warne (2020) found that 53 local authorities were making use of some form of predictive analytics in various aspects of their services. They argue austerity has created the need for councils to seek out new ways of operating and encouraged the turn to 'datafied' services. They raise concerns over the effectiveness of such systems and their ability to produce false positives. Concerns have also been raised at the way different forms of data, such as those of crime and social work encounters, are being used to build profiles on data and aide intervention (see Turner, 2019).
The activities of these systems in the England are not clear at present, which is perhaps cause for concern.
These activities suggest a form of anticipatory geographies (Anderson, 2010) at work in the dispositif of child protection, aiming to eliminate future emergencies. Place and context again are vital here in considering intervention, but how such systems would account for these nuances is not clear. We argue that an attention to geographies is critical here in revealing the spatialities at play in the dispositif of child protection.

| CONCLUSION
This short review has drawn together literatures in social work and geography to discuss child protection practice, an area that has escaped the attention of geographers to date. We began with an explanation of contemporary child protection practices in order to explore what may be an unfamiliar area to many geographers. Drawing upon Foucault (1980), we conceptualised child protection practice as a dispositif; a constellation of discourses, polices, laws, actors and networks that produces and orders lives, often in response to an emergency. This is a useful framing, we argue, for shedding light on the wider geographies of child protection practice, which are at times hidden. Through a discussion of the historical geographies of child protection, it is clear that there has long existed a spatial tension between more individualising practices and more community-based models, which attend to wider spatial and contextual issues facing families. Through a discussion of the relationalities of child protection and the impacts of austerity facilitating the search for more cost-effective, 'scientific' means of identifying and eliminating risks to children, we have highlighted that individualising and decontextualised approaches still strongly shape the dispositif of child protection practice in England. Geographers are well placed to contribute to these debates and we call for greater geographical engagement to build upon emerging work in this area Lloyd, 2019).

DISNEY AND LLOYD
It should be noted however that this short review has only initiated what must be a much further reaching dialogue between geography and social work to unpick the spatialities of the dispositif of child protection. To suggest just two: we have focussed only upon England as a context for discussion and there are important and complex historical geographies of colonial pasts that require consideration. Similarly, there is increasing effort in geography to reveal processes of 'everyday bordering' (Cassidy, 2019), where the state's border is outsourced to other agents. A consideration of the agency and resistance of child protection social workers to these activities is another area that requires attention.