Child Welfare Inequalities in a Time of Rising Numbers of Children Entering Out-of-Home Care

The relationship between poverty and child maltreatment and by extension of being placed in out-of-home care is a well-established one. However, this study goes beyond recent UK studies on the scale of child welfare inequalities in the likelihood of being placed in out-of-home care by considering such inequalities over time. The study is an analysis of longitudinal administrative data on children ‘looked after’ with a specific focus on children entering care in the two years that followed the death of Peter Connelly in 2007, a period that saw a rapid increase in numbers of children entering care. The analysis considers these increases using a child welfare inequalities lens. There is a ‘social gradient’ present within the overall rates of children entering care, with children in the most deprived neighbourhoods almost twelve times more likely to enter care than those in the least deprived. Such inequalities are compounded further in times of rapidly increasing entries to care with children entering care being disproportionately drawn from the poorest neighbourhoods, illustrated by a 42-per cent increase in rates between the two years in the most deprived neighbourhoods whilst rates in the least deprived neighbourhoods fell or remained the same.

S e e h t t p://o r c a .cf. a c. u k/ p olici e s. h t ml fo r u s a g e p olici e s.Co py ri g h t a n d m o r al ri g h t s fo r p u blic a tio n s m a d e a v ail a bl e in ORCA a r e r e t ai n e d by t h e c o py ri g h t h ol d e r s .

Introduction
This article will consider child welfare inequalities and what happens to those inequalities in relation to children 'looked after' during a period of significant, rapid increase in the numbers of children entering care.This will be done by exploring inequalities during a significant period in the UK for statutory children's services, the period from April 2008 to March 2010, which covers the events surrounding the death of Peter Connelly (Baby P) and their immediate aftermath.The use of a social inequalities lens is well established internationally in both health and education research.However, its use in child welfare research, particularly in the UK, is relatively under-developed although it does get some attention in the USA and elsewhere.This study aims to go beyond the small, but growing, body of literature in the UK that seeks to describe the 'social determinants' of child welfare and highlights the current lack of consideration given to the socio-economic drivers of contact with child welfare services by looking at such inequalities over time.
The relationship between poverty and child maltreatment is a wellestablished one (Conger et al., 1992;Pelton, 1994;Jonson-Reid et al., 2009;Berger and Waldfogel, 2011;Pelton, 2015;Conrad-Heibner and Scanlon, 2015;Slack et al., 2017).There is also, by extension, a substantial literature linking socio-economic factors to outcomes such as being placed in out-of-home care (Bebbington and Miles, 1989;Franzen et al., 2008;Simkiss et al., 2013) often as a result of abuse and neglect.Whilst this considerable body of research exists showing a relationship between poverty and child maltreatment and/or becoming 'looked-after', there is relatively scant research focused on identifying and explaining the causal relationship between the two (Slack et al., 2017).There is, however, the beginnings of an emerging literature that suggests the presence of a causal relationship (Shook and Testa, 1997;Fein and Lee, 2003;Cancian et al., 2013).In particular, the quasi-experimental study by Cancian et al. in the USA showed a reduction in child maltreatment reports within a group whose income was increased by as little as $100 a year relative to a control group, suggesting a causal relationship between household income and child maltreatment.However, within the UK in recent years, the message from some politicians and senior policy advisors has been to talk down ideas of a correlation, much less a causal relationship between poverty and child maltreatment.It has been argued that structural explanations can be used to portray abusive parents as victims of social injustice thus robbing individuals of the power of agency and breaking the link between their actions and the consequences (Gove, 2013).Similarly, it has been argued that whilst 'there may be a partial correlation between disadvantage and poor parenting ... there is not a causal link' (Narey, 2014, p.11).The understanding and even acceptance at a policy level of a relationship between poverty and child abuse is therefore contested and, by extension, the relationship between families' socio-economic circumstances and children becoming 'looked after'.
There are observable differences both between countries in the UK and at the level of local administrative areas (local authorities) in the numbers of children in out-of-home care.Whilst these differences have been attributed to a range of factors (Dickens et al., 2005), deprivation has been identified as a major explanatory factor (Oliver et al., 2001;Cordis Bright, 2013).Bywaters (2015) suggests that such differences should 'be reconceptualised as child welfare inequalities' (p. 6) borrowing ideas of social determinants from health inequalities research.Bywaters and colleagues have in recent years lead a number of studies in the UK on child welfare inequalities (Bywaters et al., 2016(Bywaters et al., , 2018a)).The original Bywaters study, undertaken using data relating to children in fourteen local authorities in the English West Midlands (Bywaters et al., 2016) and the larger four UK nation Child Welfare Inequalities Project (Bywaters et al., 2018b), which followed it, provide the impetus for this study.What these studies evidenced was the presence of an overall 'social gradient' (Marmot, 2010) in the rates of child welfare interventions, both in children being placed on the child protection register and children being placed in out-of-home care (becoming 'looked-after' children).Broadly speaking, what was shown is that for each step increase in relative deprivation there is a corresponding increase in the rate of children welfare interventions.Whilst present in all four nations of the UK, this social gradient was shown to be steeper in Wales (Elliott and Scourfield, 2017), where the study to be discussed was undertaken.Both of these earlier studies used data collected on a single census day (in 2012 and 2015, respectively).The intention of this study is to provide further insights into child welfare inequalities through the use of longitudinal administrative data to consider inequalities over time.
The findings to be described are drawn from an overall analysis of six years of child-level routinely collected administrative data, but will predominantly focus on entries to care during a two-year period between April 2008 and March 2010.This period has been chosen because, as shown in Figure 1, it covers the period prior to and during the initial rapid increase of entries to care in the wake of the death of baby Peter Connelly (also referred to as 'Baby P'), a high-profile child death in the UK.
Peter Connelly died, age seventeen months, in the London borough of Haringey in August 2007.Those accused of causing his death were prosecuted in November 2008.In the same month, a public summary of the Serious Case Review undertaken following his death was also published (Department for Education, 2008).In December of that year, the secretary of state for Children, Schools and Families used special powers to remove Sharon Shoesmith, the head of children's services in Haringey, from her post.In the same month, he also commissioned Lord Laming to urgently evaluate the practice that had been developed since the publication of the report of the Independent Statutory Inquiry following the last high-profile child death, Victoria Climbie ´, in February 2000.Lord Laming's report, published in March 2009, stated that there was a need for 'a step change in the arrangements to protect children from harm.It is essential that action is now taken so that as far as humanly possible children at risk of harm are properly protected' (Laming, 2009, p. 4).The impact of these events, which unfolded during the period from November 2008 to March 2009, was to prove significant in terms of the 'looked-after' children populations within both Wales and England.In the first three weeks that followed publication of the Serious Case Review, the Children and Family Court Advisory and Support Service (CAFCASS) identified a sharp increase in the numbers of care order applications made to the courts, relative to the same period in 2007 (CAFCASS, 2009;CAFCASS, 2012).This trend is one that continued in the years that have followed.This analysis is not intended to focus on practice or wider issues such as the rise of risk-averse practice (Hood et al., 2016); the impact of the media coverage of high-profile child deaths (Ayre, 2001); or other factors that led to the so-called 'Baby P effect' (CAFCASS, 2009) on numbers of children entering care, as these are discussed elsewhere (Jones, 2014;Warner, 2015;Shoesmith, 2016).These two data collection years cover the period of the largest singleyear increase in rates of children 'looked after' in Wales in recent times-a seven children in 10,000 child population increase in rates between those reported for 31 March 2009 and 31 March 2010-and as such represents an important period to explore from a child welfare inequalities perspective.What this article will therefore explore is the  (2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014) relationship between families living in deprived neighbourhoods and the likelihood of children becoming 'looked after' against the backdrop of these rapid increases in numbers.

Methods
The overall analysis used routinely collected administrative data on children placed in out-of-home care ('looked-after' children) over a six-year period.The child-level data used are based on the information about children 'looked-after' collected by each Welsh local authority.The data are submitted annually to the Welsh Government in the form of the SSDA903 return, a term which will be used to describe the dataset throughout this article.Access to the data was gained through a data access agreement between the researcher and Knowledge and Analytical Services within Welsh Government.Ethical approval to undertake the study was gained through the ethics committee in the school of social sciences at Cardiff University.
These data were organised to identify children at the first time they entered out-of-home care in Wales during the period from April 2008 to March 2014.The overall analyses conducted looked at children at the first time of entry during the six-year period.However, whilst some reference will be made to the findings of the overall analysis, data relating to two of these data collection years will provide the primary focus.
The overall six years of data produced a sample of almost 9,000 children and young people (n ¼ 8853).Due to levels of missing home postcode data, these cases were drawn from 18 of the 22 Welsh local authorities.Two collection years, however, provide the predominant focus of this analysis, covering the period from 1 April 2008 to the 31 March 2010 and representing almost 3000 cases (n ¼ 2957).These data were linked to both population data to enable the calculation of rates per 10,000 of the child population and, using the Welsh Index of Multiple Deprivation (WIMD) (StatsWales, 2014a), to sociodemographic characteristics at the level of small area geographies (Lower Super Output Areas) relating to the neighbourhoods from which children entered care.The WIMD, the Welsh Government's official measure of relative deprivation, is made up of eight separate domains: income, employment, health, education, access to services, community safety, physical environment and housing, with each domain compiled from a range of different indicators (Welsh Government, 2014a).These domains are used to calculate an overall deprivation score for each of the 1909 lower super output areas (LSOAs) in Wales and it is these overall scores that are used within this analysis to compare deprivation levels between 'neighbourhoods'.
The population data used to calculate rates for the whole six-year period were derived from the mid-year population estimate for 2011 from which the population of zero-to seventeen-year olds were established.These are derived from the 2011 Census (Office of National Statistics, 2012) and are arguably more robust than other population estimates for the period covered by the study data.Secondly, the data, collected around June 2011, also represent the mid-point for the years covered by the 'looked-after' children data.The analyses used the WIMD for 2014 (StatsWales, 2014a, b), which represent the most recent data for neighbourhood-level sociodemographic characteristics in Wales.Linked to the LSOA relating to the address from which each child entered care, the overall WIMD score for each LSOA enables some level of description of the levels of relative deprivation in the neighbourhoods in which children were living before becoming 'looked after'.This study, in common with those in the UK that have preceded it, used postcode level data that allows some consideration of neighbourhood level effects as a proxy measure for the household circumstances of families whose children enter care.

Findings
Rates per 10,000 of the child population were calculated for each deprivation decile based on the neighbourhood from which each child entered care during the six-year period covered by the overall data.The results are presented in Figure 2.
What the figure illustrates is the presence of an overall 'social gradient' (Marmot, 2010) in the rates of entries to the care system within Wales.As discussed earlier, what this illustrates is that for each step increase in relative deprivation, there is a corresponding increase in the rate of children becoming 'looked-after' for the first time during the observation window.As illustrated by Figure 2, a child living in decile 1, the 10 per cent of most deprived neighbourhoods in Wales, is almost twelve times (11.8) more likely to become 'looked after' than their peers living in the least deprived neighbourhoods (decile 10).This pattern replicates that found in the Bywaters et al. research on which this study builds.As part of the wider study from which the analysis presented here is drawn (Elliott, 2017), this pattern of increases in rate for each step increase in deprivation level is shown fairly consistently across years and regardless of how the data are disaggregated, i.e. age group, year of entry, sex, legal status, etc.
The question is what happens to those 'social gradients' that illustrate child welfare inequalities during a period of increasing entries to care, such as that witnessed in the wake of Peter Connelly's death?Using child-level data relating to entries to care during the year in which events such as the trial of those responsible for his death and the publication of the serious case review unfolded What Figure 3 shows is a 42-per cent increase between the two collection years in the rate of children entering care from the 10 per cent most deprived neighbourhoods (decile 1), an increase of twenty-two children per 10,000 of the child population.Similarly, there was a 35-per cent increase in the rates from the second decile, representing an increase of twelve children per 10,000.From the fifth to the ninth deciles, rates between the two years either remained the same or went down.Whilst the tenth decile does show a substantial increase because of the small numbers of children overall entering care from neighbourhoods in this least deprived decile, this only represents an increase of eleven children.This suggests that in addition to the overall inequalities in the likelihood of a child becoming 'looked after' illustrated earlier, the impact on local policy and practice in the immediate aftermath of the events The SSDA903 data as provided by Welsh Government included each child's age in the year of entry to care.This allows a comparison of age groups by deprivation quantile for each of the collection years to also be compared.Due to the small numbers in some age and deprivation categories, this analysis was undertaken using quintiles rather than deciles and age groups rather than individual ages.The analysis shows that as well as being drawn from the most deprived neighbourhoods, the increase in numbers of children entering care were predominantly drawn from families with younger children.We know that overall the youngest children (birth to four years) are disproportionately represented in entries to care (Bebbington and Miles, 1989;Elliott, 2017) and enter at the highest rates.What Table 1 shows is that the rate of children aged from birth to four years, living in the 20 per cent most deprived neighbourhoods (Quintile 1) in Wales, increased by 37 per cent in 2009/2010.Perhaps more surprising is the increase in the rate of children aged five to eleven years living in the poorest neighbourhoods between the two collection years.This age group saw the rate of entry increased by 81 per cent during a twelve-month period.This finding does, however, mirror that of the CAFCASS report on care order applications in this period, which also noted an increase in the percentage of children aged five to ten years (CAFCASS, 2009).In this most deprived quintile, the rate of twelve-to fifteen-year olds also saw a 27-per cent increase, whilst rates of sixteen-to seventeen-year olds decreased.Whilst five-to eleven-year olds living in neighbourhoods in the second quintile also saw a 27-per cent increase, all other age groups across the remaining quintiles saw either reductions in rates between the two collection years or very small percentage increases.
The data also include information on the legal basis on which children and young people entered care.For the purpose of this analysis, the two main categories under which children started their period in care, voluntary accommodation and care orders (full and interim) will be focused on as these represent 89 per cent of all entries over the years being considered.
Table 2 shows that the highest overall rates relate to voluntary accommodation, which in itself is an interesting finding given the predominant narrative around increases in care orders during the period.This would also merit further investigation regarding the issue of the use of voluntary accommodation and whether this demonstrates an ethos of working in partnership with families and providing them with support or as Lord Justice Munby suggested it is being used inappropriately as a prelude to care proceedings (Stevenson, 2015).This is, however, beyond the scope of this article to discuss.In the 20 per cent most deprived neighbourhoods, the increase in children entering care at the direction of the courts saw a 114-per cent increase in rates between the two collection years and a 39-per cent increase in voluntary accommodations from the same neighbourhoods.In the remaining quintiles, rates of entry through voluntary accommodation either reduced or remained unchanged.In the second quintile, rates of children entering under a care order saw an increase of 82 per cent.What these analyses suggest is that the increases in entries to care in Wales in the wake of the Peter Connelly case were concentrated in the poorest neighbourhoods and predominantly consist of younger children.Whilst there were increases in rates of entry under both of the main legal routes the starkest difference is in the rates of children subject to care orders in the poorest neighbourhoods, where rates more than doubled within a year.

Limitations
A limitation of the methodology used in this study is the use of measures of neighbourhood-level deprivation as a proxy measure for the household circumstances of families whose children enter care and the limitations of such an approach are acknowledged.One of the issues that this highlights is that no information is currently routinely collected on the socio-economic circumstances and other factors affecting the families with which children's services work.In order to fully understand the families with which services work and the outcomes of those interventions, it is clearly vitally important that in the future this information is collected or made more readily available through data linkage or other mechanisms so that it can be used to inform both policy and practice.
There is a possibility that what has been captured in analysing differences between the two collection years that provide the main focus of this article is naturally occurring fluctuation between years and this is acknowledged.The data as collected for the study from which this article is derived did not include sufficient data for the years before the period of interest to look at patterns both before and after the period discussed in order to rule this out.However, some additional analyses was conducted comparing rates per decile in 2008-2009 with the five years that followed up to March 2014 and the pattern of increase in entries to care from the poorest neighbourhoods was broadly seen throughout.
As discussed in the methods section and illustrated in Figure 2, the analyses of data relating to the period 2008-2010 that provide the main focus of this article are drawn from a larger study that used data relating to looked after children in Wales between April 2008 to March 2014.As a consequence, the deprivation data used for the overall study were the most recently released to the end of that period, being the WIMD 2014.
Had the analyses solely focused on the earlier period, use of WIMD 2008 would have perhaps been more appropriate, being more contemporaneous to the period being described.The study has, however, used the 2014 deprivation data as a proxy measure for deprivation across the six years covered by the overall study and the limitations of this approach are recognised.In the context of the larger study, the decision to use this approach was a pragmatic one, based on factors such as the lack of direct comparability between IMDs from different years, which would have made the use of multiple deprivation indices across the six years problematic.For example, WIMD 2008 contains 1896 LSOAs, whereas the later WIMD contains 1909, meaning the number and boundaries of units of comparison changed over time.The use of the later deprivation measure meant that direct comparisons between years could be consistently made.For example, it allowed comparison of rates by deprivation decile in the years following the period of interest, to explore whether the changes in rates observed were a consequence of yearly fluctuations.

Discussion
The overall premise of the child welfare inequalities approach is based on illustrating the presence of a 'social gradient' in the rates at which children are subject to child welfare interventions, such as being placed in care.This social gradient is present, both in the overall entries to care across the six years and in the two collection years that provide the main focus for this analysis.This finding is consistent with those of the original Bywaters et al.'s (2016) study in the English Midlands and those of the follow-up 'four-nation study' (Bywaters et al., 2018b).The data suggest that a child's likelihood of becoming 'looked after' in Wales is to an extent not a postcode lottery.It is instead part of a systemic pattern where there is a concentration of intervention in the lives of children and their families in the most deprived areas.
I would argue that the increase described here is predominantly drawn from families that were in the main already known to children's services.This supposition is based in part on comments by CAFCASS who suggested in their 2009 research report that 'a substantial proportion of the increase can be attributed to local authorities re-evaluating their involvement with families where they are already providing a service' (CAFCASS, 2009, p. 17).Whilst it is not possible to say with any certainty with the data used, the increase in entries to care seen in the months immediately after the events of late 2008 and early 2009 are therefore arguably not the product of the type of 'net widening' described by Bilson and Martin (2017) and others where more and more families are drawn into the child welfare system.Increasing number of families being subject to investigation, the 'investigative turn' (Bilson et al., 2 0 1 7 ) and children being brought into care as a consequence are I would suggest later developments that have seen this initial surge in numbers of children entering care maintained to the present day.If this re-assessment of existing families is taken to predominantly be the case then one of two scenarios seem likely.Either that such a re-evaluation of the support needs and risks associated with such families focused disproportionately on those in the poorest neighbourhoods, or if it is assumed that reevaluation took place across the socio-economic spectrum of families known to children's services at the time, then there is a relationship between the experience of living in poverty and the outcomes of local authorities re-evaluating their involvement with families.Not only are the increases observed located in the poorest neighbourhoods, but there is also a reduction in entries from most deprivation deciles outside of the 20 per cent most deprived.If the rates per 10,000 for each decile in 2008-2009 are taken as a baseline and compared to the rates for the equivalent decile for the next five years (up to March 2014), a broadly similar pattern emerges.The most deprived deciles show the largest increases with the least deprived deciles showing small increases or reductions in rates.This would suggest that the findings from this analysis are unlikely to be as a result of just yearly fluctuations.
Within child and family social work, I would argue as others are beginning to that there is and has been for many years a taken-for-granted understanding that the children and families with which social workers work are predominantly from the poorest families and neighbourhoods.That this is the case appears to be seen within the profession and more widely as unremarkable and goes therefore largely unremarked upon.It has been suggested that there is an extent to which poverty has now become the 'wallpaper of practice' (Morris et al., 2018).It is always there providing the backdrop to much social work with families, but also so familiar that sometimes it is almost forgotten or, alternatively, it is seen as too large to address.From this perspective, the findings presented here are perhaps unsurprising.What is presented here, however, is further empirical evidence of the scale of the relationship between neighbourhood deprivation and child welfare interventions.
To suggest that child abuse and neglect (the main reasons for a child becoming 'looked-after') are a result of poverty alone is to assume incorrectly that such behaviour is inevitable within families living in poor neighbourhoods and by extension that most parents in such neighbourhoods are abusive.This is clearly not the case and is not what the research is intended to suggest.Equally to focus on abuse and neglect as actions and omissions by individuals without considering the structural context of poverty within society is also to seek only partial explanations (Elliott, 2017in Family Rights Group, 2018).However, as Wright Mills suggests in the Sociological Imagination (2000), a sociological lens allows us to understand 'personal troubles' within the context of the political and economic institutions of society within which they are situated rather than just the character and personal circumstances of unconnected individuals, an approach which arguably currently predominates.In considering the findings of this study and possible responses to it, I have found Wright Mills' analysis of the nature of unemployment useful.Wright Mills suggests that: When, in a city of 100, 000 only one man is unemployed, that is his personal trouble, and for its relief we properly look to the character of the man, his skills, and his immediate opportunities.But when in a nation of 50 million employees, 15 million men are unemployed, that is an issue, and we may not hope to find its solution within the range of opportunities open to any one individual.The very nature of opportunities has collapsed.(Wright Mills, 2000, p. 9) The presence of such a clear pattern of intervention, replicated in the other UK nations as shown by the Child Welfare Inequalities Project (Bywaters et al., 2018b) suggests an urgent need not only for social work practice which works with the 'personal troubles' of individual families but a policy approach to address the 'public issue' of poverty and social inequality and its impact on the likelihood of families having need of support that will bring them into contact with child welfare services.What is perhaps sobering about these findings is summarised in the words of Bob Holman.Writing over forty years ago Holman, an academic, writer and community campaigner, suggested: It has been argued that social deprivations create problems which endanger the stability of some families.In response, governments have made little progress in reducing inequalities or removing poverty....B y contrast, the local authorities have been allowed to develop agencies well-equipped to deal with the results of the problems -to take children away from their parents.(Holman, 1974, p. 17) Decades later has a 'perfect storm' (Featherstone et al., 2014) of factors, including the impact of austerity both on child welfare services and family resilience, resulted once again in an era where local authorities are only able to deal with the results of poverty and inequality rather than having the resources to work alongside families to ameliorate their effects?William Schwartz (1979) cited in Dunk-West and Verity (2018) suggested that 'professions have a way of moving periodically through eras of rediscovery in which an old truth comes alive with the vigour and freshness of a new idea' (p.4).The recent traction gained in the UK by the findings of child welfare inequalities research would suggest we are entering such an era.The use of a child welfare inequalities lens can provide the catalyst for such a reinvigoration of conversations about the need to revisit ideas of social work practice that both works with the 'personal troubles' of families, but also places them in their structural context in a way that has perhaps been lost in recent times.Whilst this article has highlighted the compounding of child welfare inequalities during the period described, with the increased number of children entering care being disproportionately drawn from the poorest neighbourhoods, it is clear that more generally such inequalities are ever-present.Whilst, as already highlighted, there is a need to reinvigorate conversations amongst practitioners, policy makers and social work educators about the role of structural factors in the lives of children and their families, there is also a need for action.At the level of social practice, there is a need for a renewed focus on practice that is poverty aware.The anti-poverty practice framework for social work (Morrison et al., 2018) developed in Northern Ireland and the poverty aware paradigm of social work developed by Krumer-Nevo (2016) both highlight ways in which social work can be practiced in ways that both recognise and respond to the role of poverty in the lives of many of the families with which we work.However, such structural inequalities also need to be acknowledged and acted upon beyond the sphere of social work.At a national and local government level, structural change is also required, in ways which recognises and gives equal weight to tackling child welfare inequalities in the same way they already are in health or education.
Pl e a s e n o t e: C h a n g e s m a d e a s a r e s ul t of p u blis hi n g p r o c e s s e s s u c h a s c o py-e di ti n g, fo r m a t ti n g a n d p a g e n u m b e r s m a y n o t b e r efl e c t e d in t his ve r sio n.Fo r t h e d efi nitiv e ve r sio n of t hi s p u blic a tio n, pl e a s e r ef e r t o t h e p u blis h e d s o u r c e.You a r e a d vis e d t o c o n s ul t t h e p u blis h e r's v e r sio n if yo u wi s h t o cit e t hi s p a p er. Thi s v e r sio n is b ei n g m a d e a v ail a bl e in a c c o r d a n c e wit h p u blis h e r p olici e s.

Figure 1 :
Figure 1: Rates of children looked after as at 31 March (2003-2014) For the comparisons presented here between children entering care between April 2008 and March 2009 and April 2009 and March 2010 (Office of National Statistics, 2013), the mid-year population estimates for children and young people aged zero to seventeen years for 2008 and 2009, respectively, were used.
(1 April 2008 to 31 March 2009) and the year after (1 April 2009 to 31 March 2010), it is possible to compare the socio-economic composition of the cohorts of children entering care in each year.

Figure 2 :
Figure 2: The overall rates of children becoming 'looked-after' for the first time by deprivation decile (WIMD 2014) 2008-2014