The Child Protection System in England

Written evidence submitted by Julie Haines, Justice for Families

A Discussion of the Threshold Criteria used in Family Law for the Interim Removal of Children, and Possible Remedies.


This is a short submission centring on one of the three areas prescribed by the Education Select Committee for discussion, namely threshold in care proceedings. I am Julie Haines and work as a senior lay advocate for Justice for Families chaired by John Hemming MP.

My back ground is in teaching, performing and conducting. However, in 2005, Worcestershire County Council took an interest in my five children based on highly spurious grounds. The children were registered for 8 months on the child protection register on the strength of incorrect evidence, after which we managed to get them de-registered without making any changes to our family and without a social worker ever seeing our children.

Eventually, we were totally vindicated. Had we dealt with our local authority the way they wanted us to, it is unlikely our children would have made all the achievements they have.

We have a child studying for a degree at Staffordshire University, one studying A-levels, one a boarding chorister singing in a midlands cathedral choir, one who has gained a place at Chetham’s Music School in Manchester to sing in Manchester Cathedral choir and one (eight years of age) who is still emerging in terms of skill.

I joined Justice for Families in 2006 as a volunteer after giving advice on the net for about two years. I have had the benefit and experience since then of working with parents and grandparents in a large number of varied cases, giving me an oversight possibly unique amongst those outside the legal profession. During this time I have been allowed rights of audience in many courts across England and Wales including the High Court and the Court of Appeal, and I have presented cases to most of the Justices and Lord Justices in the Family Division. I have always been complemented on the assistance I have given and the legal arguments I have deployed against the applicant authorities.

Thus, it is from this basis of practical expertise of dealing with many varied situations that I wish to make a contribution for consideration to the Education Select Committee.

Threshold Criteria

Proving the threshold of significant harm to the child has been met is key to Children's Services when considering taking a child into interim care via an Interim Care Order or an Emergency Protection Order. For the courts, looking at whether the threshold is met is fraught with difficulties. For the parent accused of causing harm to the child by the local authority, defending such an action seems an insurmountable task and for the local authorities wishing to prove their case, it seems that the advantage is on their side. Thus, the starting point is statute:

The Children Act 1989 s 31 states:

(1) On the application of any local authority or authorised person, the court may make an order-

(a) placing the child with respect to whom the application is made in the care of a designated local authority; or

(b) putting him under the supervision of a designated local authority

(2) A court may only make a care order or supervision order if it is satisfied-

(a) that the child concerned is suffering, or is likely to suffer, significant harm; and

(b) that the harm, or likelihood of harm, is attributable to-

(i) the care given to the child, or likely to be given to him if the order were not made, not being what it would be reasonable to expect a parent to give to him; or

(ii) the child’s being beyond parental control.

In considering the above it is clear that Parliament has given the courts free reign to define the term "significant harm" within case law authorities and has not deemed it necessary to provide a definitive meaning within the Children Act 1989. There is no check list of harm, no clues as to what the courts could be looking for. This is a problem for parents, as they may feel that they are meeting all their children’s needs and parenting in the way they see others parenting within their own demographic subset..

There are a number of criteria for defining harm to a child, namely: emotional, neglect, sexual and physical. Most of these are fairly self explanatory to the lay person, however some of these as they are currently posed to the parent do not fall within the normal understanding and definitions that a person on the street would recognise. Thus the playing field is certainly not a level one with the balance tipping on the side of the applicant council.

Significant Harm Due to Neglect.

It could be argued that some aspects of neglect do not give rise to the child being put at risk of significant harm such as an untidy or unacceptably dirty home environment or chaotic life style, provided the basic needs of the child are met. Most parents would say that they work on a basis of "my home, my rules".

Another problem for the courts to consider is that of cultural and cultural anthropological issues. For example, parents moving from parts of Africa may not realise that they cannot leave their children home alone, or understand that the law on chastisement in their own country may well be very different from the law here in the United Kingdom. Such cultural differences are never really explained to the parent, nor are they taken in to consideration by the courts who apply the law to the maxim of "our society, our rules".

However the serious neglect of a young child can have negative effects on his or her ability to form attachments and can be linked to major growth impairment and intellectual development. The impact of neglect varies depending on how long a child has been neglected, the child’s age, and the variation of neglectful behaviours the child has been experiencing. Rarely is it the case that sustained neglect over the whole or most of the child's life is an issue. Neglect appears in a child's life when a major family issue has taken hold, such as depression and very commonly post natal depression in mothers. The loss of a job can bring on neglectful attributes in parenting. Bereavement and debt are other circumstances where child care may fall below accepted norms for a time in the child’s life. Yet these are variables that are sadly not taken in to full consideration by the courts and not dealt with on a practical level. It seems the "services" aspect of the social service and children's services departments are missing. People who ask for "services" from their local council end up with their children being permanently taken by the State. I have seen this happen on a number of occasions during the course of my work with Justice for Families.

Significant harm due to Emotional Abuse.

Most people would not realise that a domestic argument between parents or witnessing a parent's distress during a depressive episode are considered to be emotional abuse. If asked the question of what they understand by emotional abuse, then the answer that would likely come back is that they would assume that undermining the child's confidence is emotional abuse.

Emotional abuse is not a reason in itself to take a child from their parents. Yet in Re C & B (Care Order: Future Harm) [2001] 1 FLR 611 this was exactly what happened. At the interim stage there was no threshold to take the child in to care. Yet a thriving 10 month baby ended up residing in local authority care. Hale LJ heavily criticised the local authority to for its attitude, the sense of which is a ‘grab now and ask questions later’ policy.

The other criteria for threshold to take a child in to care, that of risk of sexual harm and physical abuse are fairly self explanatory and will not be discussed here directly.

Interim removal.

There are a number of authorities setting out the remit for the interim removal of children from their families.

In Re O (Supervision Order) [2001] 1 FLR 923 , Hale LJ emphasises that

‘The court should begin with a preference for the less interventionist rather than the more interventionist approach. This should be considered to be in the better interests of the children … unless there are cogent reasons to the contrary …’

In Re C & B (Care Order: Future Harm) [2001] 1 FLR 611 , ( as previously mention above),the applicant authority perceived harm to the child derived from the mother’s personality traits which in stressful situations, including conflict with the father, led to her becoming irrational, aggressive, emotionally demanding and incapable of putting the children’s needs before her own (para 15)). At the time of the interim hearing (resulting in the child’s removal) there was no evidence of physical harm; on the contrary, the evidence was that he was thriving.

In Re G (Care: Challenge to Local Authority’s Decision) [2003] 2 FLR 42 , Munby J held:

‘The fact that a local authority has parental responsibility for children pursuant to s 33(3)(a) of the Children Act 1989 does not entitle it to take decisions about children without reference to, or over the heads of the children’s parents. A local authority, even if clothed with the authority of a care order, is not entitled to make significant changes in the care plan, or to change the arrangements under which the children are living, let alone to remove the children from home if they are living with their parents, without properly involving the parents in the decision-making process and without giving the parents a proper opportunity to make their case before a decision is made. After all, the fact that the local authority also has parental responsibility does not deprive the parents of their parental responsibility.’

If a threshold is being made out for the interim removal of children then the interference with family life needs tto be considered. In Re B (Care: Interference with Family Life) [2003] 2 FLR 813 , Thorpe LJ held:

‘the judge may not make such an order without considering the European Convention for the Protection of Human Rights and must not sanction such an interference with family life unless he is satisfied that that is both necessary and proportionate and that no other less radical form of order would achieve the essential end of promoting the welfare of the children.’

Yet family life in many cases is all but forgotten by the courts citing the "paramountcy principle" of the Children Act 1989 from the welfare checklist. In the balancing act undertaken by the judge, he needs only to cite this principle and state that he has considered the Human Rights Act 1998, though even this is frequently neglected.

Unquantified Risk.

No risk identified to a child should be left as unquantified. To enable this the court should ask itself two things.

1) Having identified a risk, what is the risk to the child?

2) Can the risk be managed, and if so, how?

Case Study 1

This is a case I was directly assisting towards the end of the court of first instance hearings and involves two caring parents and their two boys who shall be referred to as T and K. At the age of approximately nine months K sustained two possible injuries, one to the leg as a metaphyseal fracture and a "boggy swelling" to the side of the head. The possibilities for the pool of perpetrators were the two parents, two day carers and the carers’ two teenage children. The carers and their children were ruled out despite having both boys in their care for considerable amounts of time. No reasons were given for the exclusion of the carers’ children from the pool of perpetrators. Thus, only the parents were considered to be possible perpetrators of one and maybe two injuries to K. Both denied even knowing about it as when the swelling to the head was discovered the child was taken to the hospital. The problem for the court was that the parents were regarded as an "unquantified risk" to both boys. At that point, the possible risk both parents posed should have been quantified. This, for whatever reason, did not happen and the boys were made subject to a placement order in September of last year. They are still in long term foster care after approximately 39 months.

The incongruity of this situation means a life long nightmare for the parents and puts both boys at risk of future significant emotional harm. Assuming that one parent was responsible for the injuries then the innocent parent has lost both of the children. If neither parent was responsible then four innocent people will suffer: both the parents and the children.

The court was not saying that the parents were an "unquantifiable risk" just that they posed an "unquantified risk". In this regard, the threshold for harm could not have been met as the court was dealing with a complete unknown, whereas a threshold document deals with known variables. If family law used a risk based approach such as is used in the insurance industry, a local authority could more accurately assess the probability of harm to a child with a greater confidence level than just having a pool of perpetrators based on the balance of probability.

The likelihood of significant harm occurring to a child could be quantified and managed so that the parent and child can remain together. For this to happen there needs to be a comprehensive probability model developed. For example experts for the local authority can say that the likelihood of harm to a child is ‘minimal’, but that kind of value judgement is meaningless when such a term may mean something different to everyone in the court room.

Additionally the cost to the tax payer in managing such a case so that children can remain with parents is minimal, compared to the costs of long term foster care, which in the first case study will have already exceeded the £150,000 mark.

Case Study 2

I another case I assisted with in Birmingham, the step father was found, in a finding of fact hearing, to have sexually abused his step daughter. Mother did not accept the allegations and supported her husband. They had a boy. The child resided with mother whilst the parents were living apart as part of a working agreement. No harm came to the child, yet when an application for an ICO was made the boy was moved in to care. What tipped the balance for the judge was a psychological report that stated that father was "likely" to sexually abuse his son. Under cross examination by myself the psychologist claimed that since the father had only abused a girl the risk of abuse to the male child was about 25%, and hence not a ‘likelihood’ of harm at all. Nonetheless, the mother lost her younger child as it was deemed by the court that she would fail to protect him.

Case Study 3

In this study, two parents and two children are involved, one with special needs. In October 2011 the county council took the children into care on without any legal authority. The children were returned to their family’s care the next day once it was established that they were taken illegally by the council with the police in attendance at the home.

The parents have always maintained that they cared for the children with no issues of harm, yet the council made an application for an ICO which was granted in January this year. There is no evidence of harm, physical or otherwise, just the merest suggestion that the children aged 12 and 13 years of age may be unhappy about the fact that naturism is practiced within the home. There have been two allegations of sexual abuse perpetrated by the father over a period of five years. Both allegations were investigated by the police and found to have no substance. Therefore, the suggestion by the applicant authority is that the girl aged twelve (nearly thirteen) years of age is at risk. No risk has been demonstrated and yet both children have been removed from the family home. The children are able to speak for themselves, yet the guardian never saw the children and has not sought their views, and the court has refused permission for the older child to appear a witness.

If there was a probability model in place of the kind used in the commercial fields, then in all likelihood the children in these three cases could have remained safe within their own families, as the alleged risk could have been managed.


To conclude, it is very clear to all that more needs to be done to keep children and families together. It is also clear that any perceived risk needs to be quantified and a risk based approach adopted. Once that has been established, then the risk needs to be managed and monitored. Over a period of time to be determined in each case, the management of the perceived risk can be lessened as the longer the time period the less likely the harm is to occur.

I would urge the members of this committee to read the transcript of the lecture given on behalf of NYAS by Mr Justice Ryder in 2007 entitled ‘The Risk Fallacy - A Tale of Two Thresholds‘.

The threshold debate would be further advanced by the application and adjudication of a jury, as in the criminal courts. The criminal standard would not necessarily need to be applied, however the safeguards that a jury can provide would mitigate against the almighty judicial discretion. The exercise of judicial discretion by one person , however experienced cannot indemnify the respondent parent against bias. The use of a jury (and the costs involved) would ensure that the need for litigation against the parent is warranted and that the "on the balance" criteria for judicial decision making against the parent is fair. Obviously there are privacy considerations to factor in, but this as a practical problem could be worked through so as to make the proposition manageable. Juries are already under a duty of confidentiality.

Every member of society has some experience of parenting, whether through the raising of children themselves, or simply though their own experience of childhood. A jury is fully capable of deciding on its own terms whether a situation which is brought before it is abusive or not. Such decisions are better made according to the norms of our own society than by the writers of sociological text books.


Lecture given on behalf of NYAS by Mr Justice Ryder,

Family Division Liaison Judge for the Northern Circuit,

on 8th November 2007


A Tale of Two Thresholds

If a child is removed from a parent on the basis of a finding or allegation that is wrong, a tragedy is caused both to the child and the family. If a child is not so removed and then suffers death or really serious harm, no less a tragedy ensues for all concerned. Consider also the unknown perpetrator cases, such as Lancashire County Council v B [2000] AC 147 at 149 E where it was said at first instance:

"[There is an] obvious dilemma in human terms. If the (threshold) criteria are met and orders are made I am exposing one child to the possibility of removal from parents who are no risk and have done no wrong …. if the applications are dismissed then I will undoubtedly be causing one child to be returned to a parent, or parents, one or both of whom are an obvious and serious unassessed risk".

Despite a number of attempts by the Court of Appeal and the House of Lords to interpret the child protection purpose of the Children Act 1989 in a way that is compatible with the human rights principles that informed its drafting, I am going to suggest that a formula for the determination of the threshold in section 31, the facts in issue and the component elements of welfare in section 1 (3) has eluded their Lordships’ House.

As we approach the introduction of a new and more coherent case management process that is designed to identify the key issues and help to narrow and resolve them within a timetable for the child I believe it is both appropriate and necessary that we also re-examine our risk assessment and decision making process. In particular, I seek to suggest that we need to have a measured and rational debate as to whether in the determination of a likelihood or risk of harm it is always necessary for the court to make findings of fact to the civil standard of proof i.e. on the balance of probabilities.

The origin of the stark dilemma that underpins the jurisprudence of the Children Act 1989 is a legal fiction most recently described by Lord Nicholls of Birkenhead in Re O & N; Re B [2003] 1 FLR 1169 at para [10]

"Courts and tribunals constantly have to decide whether an alleged event occurred. The general rule is that if the likelihood that a past event occurred is proved to the requisite standard the law regards that event as definitely having happened. If not, it is treated as not having happened"

There is no sophisticated rationale for this rule nor is the fiction set in stone. As respects different species of decision the law provides different solutions both as to what the decision maker may take into account and to what standard of likelihood a past event must be proved or a future event forecasted. For example, we protect juries from clearly relevant material that is thought to be more prejudicial than probative and the legal context may permit decision making in accordance with a variety of standards: the Children Act itself is a good example, in the different standards applicable to section 47 investigations, interim care orders and full care orders.

As Lord Nicholls reminds us, the legal context is determined by legal policy, statutory or otherwise. It is accordingly susceptible of change. Furthermore as the policy has been set by the judiciary not Parliament, it is still in the hands of the judiciary to re-consider whether the solution fits the problem.

The problem is how to balance the need to protect families from any disproportionate interference by the state with the imperative to protect children against harm.

Risk management in everyday life usually has regard to the seriousness or severity of the consequence if protective or preventative steps are not taken. Accordingly, such a step is more likely to be taken where the harm or risk of harm would be very serious on the basis that the consequence cannot sensibly or safely be ignored. In social care situations as in other aspects of life, assessments of risk are carried out every day and there are tens of thousands of such assessments each year. With very few exceptions risk assessments do not depend on findings being made by a court, even less so are the assessors expected to come to their conclusions on the basis of their own judgment as to the harm or risk of harm asserted by reference to the civil standard of proof.

The opinion evidence of experts is the consequence of the assessment processes and techniques that they use. It will almost certainly be the case that it is not appropriate to characterise, for example, a paediatric or psychiatric risk assessment as being a conclusion to which the civil standard of proof applies in just the same way that a social care assessment, for example in accordance with the Framework for the Assessment of Children in Need and their Families TSO (2000), is neither based upon nor results in a conclusion on the balance of probabilities: see for example Re S (Sexual Abuse Allegations: Local Authority Response) [2001] EWHC Admin 334, [2001] 2 FLR 776 per Scott Baker J. The task of determining facts to a standard of proof is for the court (a principle that was re-iterated in the House of Lords by Lord Hope of Craighead in Dingley v Chief Constable of Strathclyde Police (2000) 55 BMLR 1 (9 March 2000).

As Lord Nicholls remarked at para [18] of Re O and N local authorities would be prevented from carrying out effective and timely risk assessments if they could act only on the basis of proven facts.

It can be argued that the core assessment and any specialist assessments that inform it which are likewise not dependent on proof of fact (e.g. a doctor’s differential diagnoses and prognoses) are consistent in their method of preparation and analysis with the local authority’s duty to investigate under section 47 of the Act. The trigger under section 47 is having reasonable cause to suspect that a child is suffering or is likely to suffer significant harm and accordingly reasonable cause and an assessment that is not dependent upon proven fact are compatible concepts.

I would suggest, however, that the fact that the test is different in section 31 from that in section 47 is hardly a sufficient rationale for the court in contested cases to have to deconstruct and reconstruct every assessment to examine whether all 22 dimensions and 3 domains of the assessment are well grounded in facts that are susceptible of proof to the civil standard and likewise to unpack every differential diagnosis and prognosis. The fact that we are asked to do so in order to follow the ratio of the majority of the House in Re H (Minors) (Sexual Abuse: Standard of Proof) [1996] AC 565 (otherwise known as Re H & R) is one explanation for the increasing complexity of hearings in even the most simple of disputes.

In this context it is perplexing that the approach laid down in Re H & R to the making of findings about inflicted harm and the establishment of a likelihood of harm tends to work in the opposite way to the general approach to risk management. This is because the approach in Re H & R overtly includes the proposition that the more serious the allegation the less likely it is to have occurred, which necessarily militates against a finding being made to a higher standard and correspondingly leads to the child being less well protected.

That proposition is a component of the overall approach to evidence arising out of Re H & R which it is perhaps wise to recollect. For the purposes of this discussion there are 8 propositions I wish to highlight:

1. The test to be applied in determining the facts in issue is the civil standard, namely the balance of probabilities which means that a court is satisfied an event occurred if the court considers that on the evidence the account of the event was more likely than not

1. The court is not determining whether there is a real possibility that the relevant event occurred but whether it is more likely than not that it did so

1. Where a serious allegation is in issue the standard of proof is not higher but the court will have in mind as a factor that the more serious the allegation the less likely it is that the event occurred and hence the stronger should be the evidence before the court concludes that the allegation is established on the balance of probability: the more improbable the event the stronger must be the evidence that it did occur.

1. It is perhaps of note that Lord Nicholls recognised the essential contradiction within his own analysis when he said:

"In my view, therefore, the context shows that in section 31(2)(a)

likely is being used in the sense of a real possibility, a possibility

that cannot sensibly be ignored having regard to the nature and

gravity of the feared harm in the particular case" (Re H and R at 585F)

That, with respect to the House, is the key to the resolution of the

problem. A key that was reflected in the previous protective

practice of the wardship jurisdiction.

1. The likelihood of future abuse is not to be proved on the balance of probabilities: one cannot say as a fact that something in the future will happen: a real possibility will suffice

1. It is settled law that for a real possibility to be established the harm that is feared need not have happened in the past. However, the court must reach its conclusions on the evidence before it, unresolved judicial doubts and suspicions can no more form the basis of a conclusion that the second threshold condition in section 31 (2) (a) (i.e. likelihood) has been established than they can form the basis of a conclusion that the first has been established

1. If the issue before the court concerns the possibility of something happening in the future: a decision by a court on the likelihood of a future happening must be founded on a basis of present facts and the inferences likely to be drawn therefrom

1. The range of facts that may be relevant to the question of the infliction of harm or its likelihood are infinite e.g. relationships, threats, behaviours and so on. Facts which are minor or even trivial if considered in isolation, taken together may suffice to satisfy the court of the likelihood of future harm.

I ought to add that the opinion of an expert is evidence and an expert can express an opinion on an ultimate issue and that provided the court does not loose sight of the fact that the expert advises and the judge decides, the judge can put such weight on the opinion of an expert as he thinks appropriate. A judge who requests an opinion from an expert does so essentially because that expert possesses a skill and expertise that the court does not have and accordingly the court should give reasons for the rejection of that opinion in any conclusion to that effect.

This is not the place for a digression into the impressive overview of evidential jurisprudence that can be found e.g. in A County Council v K D & L [2005] 1 FLR 851 and Re R (Care: Disclosure: Nature of Proceedings) [2002] 1FLR 755 both per Charles J, but it would probably suffice to say that had one of the legacies of the wardship jurisdiction not been to try and override the law of evidence by reference to the paramountcy principle to achieve a just result (which historically had never been a justification that was permitted to achieve that result) the existing conundrum might never have arisen. Likewise, had the elegant argument put forward by Mr. James Munby QC (as he then was) in Re H & R been heeded, a different result might have obtained.

I have touched on wardship practice which I ought to explain for the majority of the audience who will no longer have had that experience. In the wardship jurisdiction and at least up until 1987 it is arguable that the test for determining whether the jurisdiction should be continued, which was the only jurisidictional threshold for the wide and almost unencumbered discretion to be exercised, was whether there was a real possibility of harm. As Purchas LJ observed in Re F (Minors) [1988] 2 FLR 123 at 128 B – D:

"I do not think that a probability has to be shown but a real possibility. In that way, the interests of the child will be safeguarded".

Whether there was a real possibility of a past event having occurred or a future event taking place had to be based on evidence i.e. it could not be speculation or a fanciful possibility but a real possibility was sufficient. However, in H v H: K v K [1990] Fam 86 the Court of Appeal disapproved of the application of that standard to the proof of past harm. Croom-Johnson LJ described real possibility as a standard of proof for past harm to be a fallacy. He distinguished Re F on its facts and by reference to other contemporaneous authorities that demonstrated that in children cases past harm should be proved to the civil standard. He accepted that on the facts Re F was a case about the future and accordingly that the test was correctly described in that context as a real possibility. Butler Sloss LJ went further, justifying her analysis with the phrase "the court can only act on evidence otherwise the judge would be dispensing palm tree justice". She described the very circumstance later taken up by Lord Nicholls in Re H & R where the ultimate issue cannot be proved to the civil standard but a collection of other facts might give rise to a real possibility of harm in the future and if thereby the child is in a potentially abusing situation the judge will take steps to safeguard the child. She was, however, careful to distinguish between evidence and findings.

That left open the question of whether a real possibility as to the future had to be based on facts found to the civil standard even if the ultimate fact of harm in the past could not be proved or evidenced.

It did not of course take long for the evidential approach in Re H & R to cause difficulties and for its strict application to be re-considered. By 1999 Lancashire CC v B had been heard at first instance and in the Court of Appeal. In order to rid the system of what Lord Nicholls described as the dangerously irresponsible approach ( Lancashire CC v B @ 165G) of cases being dismissed for want of proof of perpetration to the civil standard their Lordships approved and developed 2 related concepts that were described by Walker LJ (as he then was) in the Court of Appeal decision in the same case, namely:

a) that responsibility, culpability or blameworthiness are not relevant to the threshold question and

b) that the causation of harm i.e. the identification of the perpetrator is not a necessary component of the threshold.

By emphasising that the causal connection in the attributability component of the threshold is merely a passive not an active element and that the reasonable standard of parental care necessary for the child is an objective question, Lord Nicholls was able to conclude (@162 C – D) that the absence of a reasonable standard of parental care need not imply that the parents are at fault. Accordingly the construction of ‘care given’ was widened to include any of the shared carers with the inevitable consequence that there may be no more than a possibility that the parents were responsible for inflicting the injury on the child. That has now been extended to the test to be applied to the determination of the size of the pool of possible perpetrators (see North Yorkshire County Council v SA [2003] 2 FLR 849).

In 2003 the conjoined appeals in Re O & N and Re B reached the House. The essential fiction with which we began this discussion was in stark relief. In proceedings where perpetration remained uncertain i.e. there was a real possibility that either parent could have harmed the child but it could not be proved to the civil standard which; how does the court proceed to assess risk of harm for the purposes of the welfare question in section 1(3)e)?

The threshold was satisfied in accordance with Lancashire CC v B and neither parent could be exculpated on cogent evidence proved to the civil standard. There remained a real possibility of perpetration that could not be proved to the civil standard (and therefore just to recollect, that was not a fact that exists in law). Could that be relied upon in the judicial assessment of risk? The answer is provided in the opinion of Lord Nicholls which is, if I may say so, far reaching if taken to its logical conclusion.

In simple terms, the possibility of perpetration is a conclusion on the evidence that is to be taken into consideration when the court considers the welfare question. Lord Nicholls again used strong language in defence of this derogation from the Re H & R approach:

"it would be grotesque if such a case had to proceed at the welfare stage on the footing that, because neither parent, considered individually, has been proved to be the perpetrator, therefore the child is not at risk from either of them".

There was raised before the House the obvious question of logic: if it is right as being in pursuit of the child protection purpose of the legislation to consider unproved perpetrations as a real possibility for the purposes of welfare then why not also carry across equally salient conclusions as to the real possibility of harm? At paragraph [31] Lord Nicholls hints at the argument that it would not be right to exclude from consideration at the welfare hearing all of the circumstances (my emphasis) including the possibility of perpetration. As he said, to do otherwise risks distorting the court’s assessment. And at paragraph [32] he entreats judges at fact finding hearings to give a judgment on the likelihood of perpetration to assist the assessments by the court and professionals alike.

What then is the logical distinction between a distortion caused by a failure to consider the real possibility of harm and a real possibility of perpetration of harm? The idea that perpetration is irrelevant to welfare because blameworthiness is irrelevant to threshold is not of course sustainable. You only have to consider carers who separate so that the court has to choose between an innocent and guilty parent. Likewise the nature and extent of harm itself.

At paragraphs [37] and [38] of the Re O & N Lord Nicholls does not discount consideration of the possibility of harm at the welfare stage. Although a strict application of Re H & R would prevent any consideration of harm that has not been proved, what Lord Nicholls suggests is that the court should proceed on the footing that the unproven allegations are that and no more than that. That gives the evidence an equivocal status. It was said that this approach accords with the earlier decision of the Court of Appeal in Re M & R (Child Abuse: Evidence) [1996] 2 FLR 195 but all that was said in Re M & R was that it would be extraordinary if evidence that was insufficient to satisfy section 31(2)a) should nevertheless be sufficient to satisfy section 1(3)e). And that, with respect, begs the question as Re M & R simply followed Re H & R and Butler-Sloss LJ was careful to restrict her conclusion to the same example, namely:

"if the court concludes that the evidence is insufficient to prove sexual abuse in the past, and if the fact of sexual abuse in the past is the only basis for asserting risk of sexual abuse in the future then it follows that there is nothing (except suspicion or mere doubt) to show a risk of future sexual abuse" [Re M and R @ 203D].

None of this answers the problem which arises out of the legal fiction that the inability to prove a fact to the civil standard means the fact in question does not exist. In two reported cases the High Court has suggested that Lord Nicholl’s analysis as to the application of a real possibility might be extended to include real possibility of harm. The first was A County Council v A Mother, A Father and X, Y and Z (Children) [2005] FLR 129 per Ryder J and the second was the decision of Charles J in Birmingham City Council v H & S [2005] EWHC 2885. Neither case was subsequently reported in respect of the welfare hearings because they resolved and the point was not decided on the facts. Further, although I would strongly oppose the fashionable view that the law of evidence can be overridden by mere reference to paramountcy and on another occasion could give ample justification on the authorities for that view, it was Lord Nicholls who set out the legal policy justification for considering all the circumstances at paragraph [24] of the Re O & N. There he said:

"This has long been axiomatic in this area of the law. The matters the court may take into account are bounded only by the need for them to be relevant, that is, they must be such that, to a greater or lesser extent, they will assist the court in deciding which course is in the child’s best interests. I can see no reason of legal policy why, in principle, any other limitation should be placed on the matters the judge may take into account when making this decision. If authority is needed for this conclusion I need refer only to the wide, all-embracing language of Lord MacDermott in J and Another v C and Others [1970] AC 668, sub nom J v C (1969) FLR Rep 360, at 710 – 711 and 383 respectively."

Let me go back to first principles in order to conclude with a solution. I emphasise there will be other solutions that may be better and that it is in the nature of a lecture that the safety net of applying an idea to the facts in issue is missing. First of all in relation to the legal fiction of proof, one should not forget, most particularly I would suggest in a process that is in part an inquisition not just an Article 6 compliant adversarial trial, that a real possibility on the evidence is something relatively substantial. It is not a mere suspicion or lingering doubt. It is not fanciful. To return to the example of the two perpetrators who have separated after harm is caused: it reflects the fact that there is clear evidence that one or both of them inflicted the serious harm and the court is simply unable to penetrate the fog of accusation and counter accusation to decide who. Not all harms are immediately visible or susceptible of pathognomic diagnosis as distinct from differential diagnosis. To exclude the real possibility of such harm which in reality is what many a differential diagnosis or care assessment concludes is to do a disservice to the child and depart from the intention of Parliament.

Secondly, does that mean that the court should no longer strive to find facts to the civil standard? I would suggest that the many distinguished family judges who have asserted to the contrary are on balance right. It would be just as grotesque for the court to permit innocent parents as carers of children to be left under the disability that such uncertainty causes which in itself causes harm to the child. Furthermore, it is difficult to argue against the proposition that key issues of fact should be determined wherever possible to the civil standard because that provides the protection against arbitrary interference by the state in family life and clarity as to the factual circumstances with which a child thereafter has to live.

Thirdly and where I depart from Re H & R in relation to fact finding is that I would suggest that it is almost inconceivable that in the circumstance that there is insufficient cogent evidence on the ultimate issue to find facts, that it is possible to construct a logically consistent threshold as to the future risk of the same harm based only on satellite facts proved to the civil standard. A judge after all has available to him all of the conventions of evidence: the ability to find secondary facts and the inferences reasonably to be drawn there from (Jones v Great Western Railway Company (1930) 144 LT 194) and to examine the opinion evidence of assessors to see the extent to which the opinion is supported by the evidence (the classic exposition of which is that of Stuart-Smith LJ in Loveday v Renton [1990] 1 Med LR 117 @ 125) but if on completion of that exercise, the primary fact in issue cannot be inferred then without anything more on a Re H & R approach a future event is unlikely to be forecasted and the second protective limb of the threshold is rendered of no effect.

That suggests that to be effective the second limb should be satisfied on a different basis namely the likelihood of harm should be a real possibility based on evidence not proven facts. There would then be no need for any illogical distinction to be drawn between likelihood in section 31(2) and risk in section 1(3)e) and real possibilities could be considered by the court. That would bring social work assessments and expert opinions into line with the court’s assessment process. That also accords with the approach of Kennedy LJ who dissented in the Court of Appeal in Re H & R and Lords Browne-Wilkinson and Lloyd of Beswick who delivered dissenting opinions in the Judicial Committee. For my part I believe they were right. The key to the solution is Lord Lloyd’s description of the fallacy of the majority (@581):

"The likelihood of future harm does not depend on proof that disputed allegations are true. It depends on the evidence. If the evidence in support of the disputed allegations is such as to give rise to a real or substantial risk of significant harm in the future, then the truth of the disputed allegation need not be proved" (and I would add, for the threshold to be satisfied).

We have become transfixed in the headlights of proof to the civil standard. There is nothing wrong with a legal policy of risk being established on the evidence to a different standard. I also agree with the dissenting voices that assessment of the evidence is a one stage process. Many of the problems associated with our fact finding process have been generated out of the erroneous view that the threshold are conditions to be satisfied in the sense of grounds for the care order. If any authority is needed that this is an inappropriate approach it can be found in the judgment of Walker LJ sitting in the Court of Appeal in the Lancashire CC v B @ 149G where he quoted with approval Lord MacKay of Clashfern LC giving the Joe Jackson Memorial Lecture in April 1989 [1989] 139 NLJ 505

"the requirement in the Bill that the court be satisfied that there is significant harm or the likelihood of such harm to the child arising from an absence of reasonable parental care …is NOT a ground or a reason for making a care or supervision order. Those conditions simply set out the minimum circumstances which the Government considers should always be fount to exist before it can ever be justified for a court even to begin to contemplate whether the state should be enabled to intervene compulsorily in family life"

We perhaps should remember that in only 15% of all care proceedings that are issued is there a contested threshold and that in the vast majority of social work cases assessments are relied upon without recourse to the court.

In summary, I am of the view that the following propositions need to be tested on the facts of an appropriate case or cases with a view to the Court of Appeal and their Lordships’ House being asked to re-consider the law:

a) The allegations of fact that give rise to the need to make a welfare assessment i.e. the key issues necessary to satisfy the threshold and inform the plan for the child must be identified in every case.

a) There should be a progression of reasoning as follows: first the identification of the evidence i.e. the factors in favour of and against the competing conclusions; Second an assessment of the weight i.e. the cogency or quality of that evidence and third a conclusion with explicit reasoning.

a) The court should strive to reach a conclusion in respect of each key issue on the balance of probabilities but where it cannot it should describe the evidence it relies upon and the judicial inferences it makes and where it comes to a conclusion that a real possibility exists it should say so and that should form not only part of the second limb of the threshold but also part of the section 1(3) risk assessment.

a) The court process should be a single stage within which two questions are answered, the jurisdictional question and the welfare question: both in the context of an overall rather than a partial or sub-divided review of the evidence and the risk assessment that is required of the court.

There are other good reasons, substantive and procedural for a reconsideration of our process. If out of a review of our approach to evidence we constrain split hearings to single issue cases that can be determined by a judgment on the key issue and the most complex factitious cases where no realistic assessment can take place without the story being re-written by the court then in my view that will be a considerable achievement. In any event I believe a decision as to the need for a split hearing can rarely be taken in early case management. It usually needs the context of assessment evidence that is not generally available until the Issues Resolution Hearing. On an altogether different plane, it is arguable that we need to re-consider Re H & R in the light of the developing Article 2, 3, 6 and 8 jurisprudence so that our decision making process remains Convention compatible.

A more reasoned and global approach to the process of judicial assessment of risk may have the effect of more cases satisfying the threshold for jurisdiction but I would hope that the renewed emphasis of the senior judiciary on whether the harm that is asserted can be regarded as significant and also on the need to show imminent risk of really serious harm before a child is removed under an EPO or ICO will focus minds on the whole process of assessment and protection.

The most recent example of the latter is Re K and H [2007] 1 FLR 2043 CA. A good example of the former is the decision of Hedley J in Re L (Care: Threshold Criteria) [2007] 1 FLR 2050 where he said:

"The current legal starting point was that children were best brought up within natural families: it followed that society must be willing to tolerate very diverse standards of parenting, including the eccentric, the barely adequate and the inconsistent, and that some children would experience disadvantage and harm, while others would flourish in atmospheres of loving security and emotional stability. It was not the provenance of the state to spare children all the consequences of defective parenting: the compulsive powers of the state should only be exercised when the significant harm has been made out".

If ever there was a good place to stop that is probably it.

Prepared 11th June 2012