Written evidence submitted by the Social Justice Research Group, University of the West of England, Bristol [ACS 028]

Contents

Executive Summary

Introduction

Local Authority budgets

Problems: the current approach

Figure 1: The theory of child protection circuit

Figure 2: Referral to substantiation ratio

Figure 3: Longitudinal increase in referrals and assessments by types of suspected abuse

Context: the neo-liberalist economic framework

Our research and solutions

Conclusions

Appendix 1: Our programme of work

2018-2020: ‘The Risk of Risk’

2017-2020: 'The Care Cases Crisis'

2014 – 2016: ‘Rethinking child protection strategy’

2012 – 2014: 'Safer Children?’

Executive Summary

The Social Justice Research Group, University of the West of England, Bristol (UWE) is undertaking a comprehensive programme of work in relation to Children’s Social Care (CSC) service provision, budget management and possibilities for new technologies and innovations.  The work is funded by grants from the Economic and Social Research Council and the Nuffield Foundation.[1] 

Our findings help to answer the questions as follows:

1.       Is the funding for local authorities’ children’s services sufficient to enable local authorities to fulfil their statutory duties?  

In short, yes, but not in the current policy framework.  There are viable alterations to the framework that will alleviate the perceived crisis in CSC funding.  Our research suggests a means to achieve this. 

2.       What are the financial challenges for local authorities’ in providing non-statutory services? 

a)      The biggest financial challenge is the requirement to ensure that statutory services are delivered on budget, to time, and to an acceptable standard.  In this regard non-statutory services are a ‘poor relation’ and may therefore be required to work with precarious and uncertain budgets from year to year.

b)      The decision to provide a non-statutory service may be multi-factorial and the cost-benefit analysis for each service is likely to be complex.  In the current climate of ‘innovation’ there is no shortage of potential services, so the challenge will be to ensure that funded non-statutory services prove they provide return on investment by an objectively robust measure, that they are demonstrably improving lives and not (as an unintended by-product) causing harm.

3.       How can funding for children’s services be made more sustainable in the short and long term? 

As noted above in 1. Our research programme is providing answers to this complex question.  A full explanation is outside the scope of this report but available on request.  In summary, the question of short-term sustainability poses a more challenging question than long-term issues. The removal of services, particularly those provided for profit will ensure a more stable position whilst long-term changes to the policy framework are implemented.  In the long-term, if the recommendations from our research are taken up, we will provide a review and re-triaging of how statutory services are delivered.  This will release the budget pressure. 

4.       How can innovative approaches to the design and delivery of children’s services support financial sustainability?

We agree with the concept of an ‘innovative approach’ but advise that the use of the term ‘innovation’ in CSC has more than one meaning and operates at numerous points in the primary, secondary, tertiary and quarternary stages of CSC service delivery.[2]  Rather than continue to fund increasing numbers of piecemeal individual, profit-making ‘innovations’ we recommend the use of a wholesale innovative redesign of CSC and their budgets with provision of ‘gold standard’, research-backed innovation, together with an ethical charter.  Our research provides a suggested new policy framework, with focus on financial and welfare outcome sustainability within the current legislation.  We suggest policy and practice re-organisation within the available space and budget, together with provision of the national and local level ‘dashboard’ we are funded to scope in order to monitor critical points in CSC that indicate adverse budget, welfare or legal outcomes. 

Introduction

Children’s social care services (CSC) are reportedly undergoing a dual crisis:

1. There are increasing numbers of children referred into, and through, the system but shrinking social care budgets;[3]
2. There are increasing numbers of children coming into state care, indicating that ‘early intervention’ is not having the effect envisaged since 2004 with the ‘Every Child Matters’ agenda.[4]

Given these issues, the questions you pose are important, particularly so in the current political and economic backdrop of uncertainty in the UK.  The balance between statutory and non-statutory service delivery has become a fine balancing act, besieged with a barrage of media, public and professional criticism postulating that austerity has caused a lack of funding particularly in relation to provision of support services.[5]  There is a linked perception that children and families are needlessly escalating through the stages of the child protection system,[6] causing a crisis of increasing numbers of s.31 care order applications.   It follows that the solution is to increase funding to head off this increase, but this demands an increased overall budget, and fails to address whether there is a flaw in the current approach that will not be alleviated (and may be exacerbated) by a budget increase.

There are several assumptions inherent in the current narrative which merit close attention before accepted as fact:

1. That child abuse is almost inevitably a continuum escalating from unmet need;[7]
2. That there is a direct causative correlation between a reduction in support services and an increase in care order applications that cannot be explained by other variables;
3. That an increase in funding will make things better, rather than escalate the scale of the problem.

Whilst we agree that funding is a key element to effective delivery of statutory and non-statutory children’s social care functions, our findings, gained through a thorough review of the whole of the CSC system[8] do not support an increase in funding as a panacea to deeper, systemic problems. It will not provide a solution to systemic issues which are costly in terms of budgets deficits, welfare failures and failures of social justice.  Our concern with an increase in budget in the current local authority delivery framework is that we will simply see the same problems of high numbers of false positives and false negatives, amplified.

Local Authority budgets

Each of the 152 Local Authorities in England sets their own CSC budget from their overall devolved budget.  They are free to deliver children’s services using their own judgement within the confines of fulfilling their statutory duties and adhering to the statutory guidance[9]

The Local Government Association (LGA), which represents more than 370 councils in England and Wales suggests that in 2017/18 councils surpassed their CSC budgets by £816m.[10] They forecast a deficit of £2billion 2019/20 for CSC budgets.[11]  We assume this is calculated by a linear progression analysis which presumes all other variables (including the framework) are unchanged.  On this assumption, the forecast highlights the status quo is unsustainable.  There are only two possible solutions:

1. Increase the budget and risk exacerbation of the problems within the current framework and a spiralling demand for increased budgets; or
2. Change selected variables to reduce spend and enable a re-triaging of the overall money available, with a ring-fenced budget around statutory service delivery.

Notably, there is no routine provision of detailed financial information across all Local Authorities (LAs) in respect of CSC budgets to provide a detailed audit and projection.  This is a deficit that Government may wish to address.  The Local Government Association (LGA) project a significant deficit going forward but do not offer a solution.  Their report is largely uncritical of the current underlying policy and approach.  As a consequence its forecasts are alarming.

Problems: the current approach

The problems in the current approach can be summarised as:

1. An underlying series of assumptions that underpin the statutory guidance and local frameworks of delivery;
2. A lack of engagement with the longitudinal data showing that this approach leads to an inevitable increase in service demand, and worse outcomes for the most vulnerable children;
3. The decision in policy to conflate ss.17 and 47 and to treat early intervention and the prediction of children requiring those services as ‘heading off escalation’;
4. The lack of central Government control over LA spend on unregulated profit-making innovations.

We have undertaken longitudinal data analysis, together with policy analysis to investigate these issues and have found that there are underpinning assumptions within the current policy framework which self-justifies the current approach (hence its representation here as a ‘circuit’).  The longitudinal data shows that this approach is not resolving the current problems.  On examination, our research showed that each assumption is based on scant hard evidence but perpetuates a significant budget deficit.

Figure 1: The theory of child protection circuit[12]

This underlying ideology in the circuit is evident in the statutory guidance Working Together to Safeguard Children.[13] The guidance has become progressively entrenched in the ideological approach without critical appraisal.  Our findings show that it is adherence to this underlying ideology which is at the heart of the current budget deficit.  The longitudinal data shows increasing numbers of children referred into the system, but a reduced child abuse detection ratio from 24.1% to 7.4%.  This is deeply concerning against a backdrop of increasing number of children at each stage of the system and reports of a consequential CSC budget crisis.

Figure 2: Longitudinal fall in referral to substantiation ratio[14]

Figure 3: Longitudinal increase in referrals and assessments by types of suspected abuse[15]

The answer to the current budget question becomes clear once longitudinal data analysis of outcomes since the Children Act 1989[16] is placed in juxta-position to the theory.  The data demonstrates that this approach progressively fails to address the social issues intended by the legal framework. 

These issues are to:

1. Provide direct support to CiN and their families; and as a separate provision
2. To intervene to protect children suffering, or at risk of suffering significant harm.

We suggest that the mixing of CSC provision of services under s.17 and child protection action under s.47 in ideology, practice and budgets is based on flawed ideology that has led to increases in demand without reduction in child harm.   The postulation that child abuse arises from unmet need is not an inevitability.  This approach misses serious cases of abuse and draws many children into unnecessary secondary-layer interventions.  The resultant data shows the unsustainability of this approach.  Evidence shows that despite a massive fiscal, policy and ideological drive towards early intervention, the result is the relentless creation of new clients (service users) who require services (privately provided) to apparently head off abuse. The data, however, does not support this approach as successful. Coupled with the drive towards predictive technologies, the amount of wasted resources is significant.  In particular the increase in the use of (mainly) ‘black box’ algorithmic predictive tools to identify apparently risky families is little more than a profit-making enterprise based on population-level epidemiological methodology with an extraordinarily high error rate in identifying child abuse.[17] This is inappropriate for use as a ‘diagnostic tool’ at individual level, raising accuracy and privacy questions in addition to questions of lack of robust UK-wide cost/benefit analysis. Our ongoing work evaluates the statistical accuracy and ethical parameters of such systems, with a view to development of a non-profit making system with appropriate safeguards and controls.

We have found that the progressive conflation of ss.17 (support for children in need and their families) and s.47 (child protection) is causing multiple problems in budgets and delivery of practice.  The statutory duties need to be discharged and to do so there must be sufficient budget.  However, the demand level for assessment is at a high (albeit relatively stable since 2016) level[18] and is driven by several factors which, taken together, can be described as an encouragement to refer children who may be socially disadvantaged as well as those where there is a ‘reasonable suspicion of significant harm’.  This conflation causes a triage problem at the outset which the Munro Review and Report’s[19] uptake evident in changes to Working Together to Safeguard Children in 2015 has exacerbated.  The removal of initial and core assessment and replacement with ‘continuous assessment’, frequently delivered by use of ‘tools’ such as ‘Signs of Safety’ is shown in research to be lacking in evidence of significant return on investment or effectiveness.[20] 

In summary, we support the use of new technologies and innovations, and our full research findings focus on where and when these are helpful.  However, uncritical uptake of innovations and technologies that purport to solve complex social problems should be treated with caution and evaluated by experts before use and implemented within an ethical framework.

Context: the neo-liberalist economic framework

We have already touched on the issue of private, profit-making providers of services and innovations.  This phenomenon should be placed in the context of the UK’s approach to state provision of services.  England’s welfare systems have been adapting and changing to accommodate neoliberal policy implementation.[21]  Neoliberalist approaches seek to reduce the size and influence of government, favouring deregulation and private enterprise.[22] In relation to CSC, this is evident in shrinking ‘in house’ services and a corresponding increase of outsourced, profit making, service provision.  It is well established that philanthropic bodies and the third sector should be engaged in supporting society’s vulnerable members, but this is extended in the UK to a wholesale policy of outsourcing and largely unregulated provision of CSC services.  Many are marketed as providing solutions but there is a lack of evidence of their fiscal or welfare benefits.[23]

The adoption of neoliberalism within England’s government policies has elicited criticism with preliminary research observing adverse welfare and wellbeing outcomes for both the family and vulnerable family members.[24]  The UK’s austerity measures have been internationally condemned[25] with some public cuts found to be unlawful.[26] As noted above, we suggest the problem is more accurately identified as the mixing of ss.17 and 47 budgets and practice which is encouraged by the neoliberalist agenda.  This is evident in the unregulated individual local authority purchase of innovations that claim to reduce child abuse, but in relation create a never-ending stream of clients to justify its ongoing uptake.

The UK’s neoliberalism is predicated on the belief that outsourcing services into a free market structure allows the state to incur less responsibility, and thus its financial requirements would diminish.  In turn this should result in lower taxation hence simulating economic growth.[27] We question, however, whether an unregulated (financially or ethically) market structure is appropriate within the context of child welfare. Although free markets should promote competition and best value for money we have observed the following adverse behaviours:

• Providers actively collaborate to fix fees, providing no financial competition within the market.  When further budgets become available, the provider fees increase meaning that overall there is no increase in service provision. An increased CSC budget in this climate is likely to improve profit rather than improve social justice.
• Prices can be ‘too good to be true’ when providers offer below odds prices for long-term contracts.  The provider risks failing to provide the contract and/or dissolves.  The government can ill-afford a ‘Carillion’ in social care to happen
• There is generally poor financial and ethical regulation of service providers, and little auditing and quality assessment of services being delivered.

Our research findings do not preclude the use of out-sourced services or innovations within the broader neo-liberalist approach evident in the UK.  We do, however, offer a more precise policy and ethical framework which is designed to work around the reality of budget constraints and address the risks posed by these adverse behaviours.

Our research and solutions

As noted above, it is outside the scope of this report to provide the full analysis and results of our programme of research, together with proposals for change.  This is available on request as it is by necessity complex.  The following is a brief overview of the scope of our multi-disciplinary programme of work for context.

We focus on data analysis and theoretical law, policy and economic evaluation of safeguarding systems in the UK, with focus on the legal and conceptual space between consensual and non-consensual interventions. Our current work spans five complementary activity streams that may be of use to Government:

1. Creation of frameworks, policies, systems and materials for redesign of CSC;
2. The use of artificial intelligence, big data, risk and prediction in social care;
3. The development of multi-jurisdictional measurement methodologies and ‘dashboard’ development for use in social care monitoring;
4. The identification of inequalities arising from individual and group vulnerabilities in the context of social justice; 
5. Theoretical and empirical work concerning the consequences and unintended consequences of State powers, duties and interventions.

Conclusions

Our findings show that the current policy framework surrounding CSC budgets is not working well. 

We flag that continuing with the current framework either with the same, or an increased budget is likely to exacerbate the current crisis in CSC rather than alleviate it. 

We advise a revised policy framework taking into account the issues our detailed full research findings have identified which explain why various aspects of CSC are not working as intended. 

As an action for the Committee we suggest you contact us to invite us to submit full evidence to inform your review and possibilities for action.

Appendix 1: Our programme of work

2018-2020: ‘The Risk of Risk’[28]

We are currently undertaking detailed work on an analysis of risk prediction in CSC, an area of growing interest in the UK since the late 1980s. Academic studies on risk prediction focus on variable correlation, rather than causation, by using epidemiological methods.[29] There is no ‘gold standard’ methodology or minimum acceptable study size. This ad hoc development has resulted in commercial technology companies marketing algorithmic software packages, developed from the correlated risk characteristics gleaned from small numbers of the epidemiological studies.  It is acknowledged that risk prediction studies in child protection are of varying sizes and methodologies, yielding a high number of false positives and a smaller, number of false negatives casting doubt on their practical use.[30]

A further layer of inaccuracy occurs when the risk factors are simply combined in an algorithm, providing a cumulative ‘score’ from which decisions regarding children and families are made.[31] Each local authority commissions their own electronic risk assessment system from commercial providers, resulting in a rise of unreliable, unregulated systems, sold at a significant profit. This represents a significant contribution from CSC budgets at a time when they are limited and stretched, and a lack of regulation regarding the methodological robustness and reliability of such systems. Of significance is that many of the algorithms that sit behind the systems are a ‘black box’; agencies that fund and deploy these cannot assess the quality of these algorithms, which are often made obscure due to commercial sensitivities. The team’s data from previous funded work identified an unacceptably high error rate from such algorithms being used and has evidence of consequential serious problems in the child protection system, obtained from a detailed longitudinal data analysis.  We observe an unacceptable error rate (in excess of 97%) on the data from which these algorithms have been based. The use of these algorithms has fuelled a high number of unnecessary investigations into apparently ‘risky’ families (9% of all families annually) who, once identified, are put through a costly and traumatic process of suspicion and investigation. Of the accused families, 88% are found, following investigation, not to be abusing their children.  We also have data suggesting that around 17% of seriously abused children are missed in the current framework.

2017-2020: 'The Care Cases Crisis'[32]

We are also undertaking a programme of work using Ministry of Justice and CAFCASS data to investigate the ‘Care Cases Crisis’. The child protection system is complicated and multi-stage. Since the Children Act 1989 introduced the modern framework for safeguarding and child protection there have been sustained increases in children referred into the system with the intention that early intervention will support families in addressing problems early on.

However, across the UK, the child rights model of protection is adopted as opposed to a model focussed on promoting family welfare. This model places the state under a duty to investigate suspected ‘significant harm, or the risk of it’ as well as to assess whether a referred child and his or her family is ‘in need’.

The state thus has two roles: that of providing family support under s.17 and the duty to investigate and act to protect children under s.47. The question of how best to deliver these two potentially conflicting roles is complex: since the Children Act 1989 there has been a 311% increase in the number of children referred into the system and more recently there has also been concerning increases in the number of children subject to a s.31 application. The increases are occurring despite the strategies of early intervention intended to reduce children progressing to this stage of the system. The resultant strain on social work and social care resources has left the system with a crisis causing tremendous strain on the family courts.

Our research is focussed on finding solutions to the rising numbers entering each stage of the system without compromising child welfare.

2014 – 2016: ‘Rethinking child protection strategy’[33]

This project provided a detailed, longitudinal consideration of the number of children and families referred and assessed, together with the outcomes.  This project identified foundational problems with the way in which the child protection and safeguarding system operates.  We concluded the problems originated with the policy framework.  We investigated Serious Case Reviews and their use in risk prediction, together with the legal, policy and practice framework.  Our concluding statistics showed that the efficiency rate of the child protection system in England has fallen from 24.1% to 7.4% in terms of abuse detection since the implementation of the Children Act 1989.  We found this to be coupled with steadily increasing costs to the system.  Overall, we concluded the system requires policy and practice review.

2012 – 2014: 'Safer Children?’[34]

This empirical project investigated the use of profit making child protection risk prediction schools for schools.  The mixed methods analysis considered the accuracy of the tools, together with qualitative and quantitative analysis of their outcomes and teacher’s views about whether they felt confident using them.

December 2018