United Kingdom House of Lords Decisions

You are here: BAILII >> Databases >> United Kingdom House of Lords Decisions >> B (Children), Re [2008] UKHL 35 (11 June 2008)
URL: http://www.bailii.org/uk/cases/UKHL/2008/35.html
Cite as: [2008] 4 All ER 1, [2008] UKHL 35, [2009] 1 AC 11, [2009] AC 11, [2008] 2 FLR 141, [2008] Fam Law 837, [2008] Fam Law 619, [2008] 2 FCR 339, [2008] 3 WLR 1

[New search] [Buy ICLR report: [2008] 3 WLR 1] [Buy ICLR report: [2009] 1 AC 11] [Help]

Judgments - In Re B (Children) (Fc) Appellate Committee Lord Hoffmann Lord Scott of Foscote Lord R

HOUSE OF LORDS

SESSION 2007-08

[2008] UKHL 35

on appeal from: [2008] EWCA Civ 282

OPINIONS

OF THE LORDS OF APPEAL

FOR JUDGMENT IN THE CAUSE

In re B (Children) (FC)

Appellate Committee

Lord Hoffmann

Lord Scott of Foscote

Lord Rodger of Earlsferry

Lord Walker of Gestingthorpe

Baroness Hale of Richmond

Counsel

Appellants:

Stephen Cobb QC

Stuart Fuller

(Instructed by Stantons)

Respondents:

Marianne Hildyard QC

Helen Mountfield

Isabelle Watson

(Instructed by Church Bruce)

Interveners

Jo Delahunty QC

Alison Grief

(Instructed by CAFCASS)

Hearing date:

19 AND 20 MAY 2008

ON

WEDNESDAY 11 JUNE 2008

HOUSE OF LORDS

OPINIONS OF THE LORDS OF APPEAL FOR JUDGMENT

IN THE CAUSE

In re B (Children) (FC)

[2008] UKHL 35

LORD HOFFMANN

My Lords,

I have had the advantage of reading in draft the speech to be delivered by my noble and learned friend Baroness Hale of Richmond and I am in complete agreement with her reasoning, analysis of the authorities and conclusions. I add some observations on the standard of proof only to underline, without in any way qualifying, what she has said.

If a legal rule requires a fact to be proved (a "fact in issue"), a judge or jury must decide whether or not it happened. There is no room for a finding that it might have happened. The law operates a binary system in which the only values are 0 and 1. The fact either happened or it did not. If the tribunal is left in doubt, the doubt is resolved by a rule that one party or the other carries the burden of proof. If the party who bears the burden of proof fails to discharge it, a value of 0 is returned and the fact is treated as not having happened. If he does discharge it, a value of 1 is returned and the fact is treated as having happened.

The effect of the decision of the House in Re H (Minors)(Sexual Abuse: Standard of Proof) [1996] AC 563 is that section 31(2)(a) of the Children Act 1989 requires any facts used as the basis of a prediction that a child is "likely to suffer significant harm" to be proved to have happened. Every such fact is to be treated as a fact in issue. The majority of the House rejected the analogy with facts which merely form part of the material from which a fact in issue may be inferred, which need not each be proved to have happened. There is of course no conceptual reason for rejecting this analogy, which in the context of some predictions (such as Lord Browne-Wilkinson's example of air raid warnings) might be prudent and appropriate. But the House decided that it was inappropriate for the purposes of section 31(2)(a). It is this rule which the House reaffirms today.

The question which appears to have given rise to some practical difficulty is the standard of proof in such cases, that is to say, the degree of persuasion which the tribunal must feel before it decides that the fact in issue did happen. Re H (minors) makes it clear that it must apply the ordinary civil standard of proof. It must be satisfied that the occurrence of the fact in question was more likely than not.

Some confusion has however been caused by dicta which suggest that the standard of proof may vary with the gravity of the misconduct alleged or even the seriousness of the consequences for the person concerned. The cases in which such statements have been made fall into three categories. First, there are cases in which the court has for one purpose classified the proceedings as civil (for example, for the purposes of article 6 of the European Convention) but nevertheless thought that, because of the serious consequences of the proceedings, the criminal standard of proof or something like it should be applied. Secondly, there are cases in which it has been observed that when some event is inherently improbable, strong evidence may be needed to persuade a tribunal that it more probably happened than not. Thirdly, there are cases in which judges are simply confused about whether they are talking about the standard of proof or about the role of inherent probabilities in deciding whether the burden of proving a fact to a given standard has been discharged.

A case in the first category was R v Secretary of State for the Home Department, Ex p Khawaja [1984] AC 74 which concerned the summary removal of an immigrant on the ground that he had obtained leave to enter by fraud or deception. These were civil proceedings and Lord Scarman, who dealt with this point most fully, was reluctant to say that the criminal standard of proof should apply: see p. 112. Instead, he said:

"I have come to the conclusion that the choice between the two standards is not one of any great moment. It is largely a matter of words. There is no need to import into this branch of the civil law the formula used for the guidance of juries in criminal cases. The civil standard as interpreted and applied by the civil courts will meet the ends of justice."

He then cited Bater v Bater [1951] P 35, in which the Court of Appeal managed to rule that although it was a misdirection for a judge in matrimonial proceedings to say that the criminal standard of proof applied to allegations of cruelty (Davis v Davis [1950] P 125) it was correct to say that they had to be proved beyond reasonable doubt. Lord Scarman then referred to Hornal v Neuberger Products Ltd [1957] 1 QB 247, which was a case in the second category, and went on at p113:

"My Lords, I would adopt as appropriate to cases of restraint put by the executive upon the liberty of the individual the civil standard flexibly applied in the way set forth in the cases cited… It is not necessary to import into the civil proceedings of judicial review the formula devised by judges for the guidance of juries in criminal cases. Liberty is at stake: that is, as the court recognised in Bater v. Bater [1951] P. 35 and in Hornal v. Neuberger Products Ltd. [1957] 1 Q.B. 247, a grave matter. The reviewing court will therefore require to be satisfied that the facts which are required for the justification of the restraint put upon liberty do exist. The flexibility of the civil standard of proof suffices to ensure that the court will require the high degree of probability which is appropriate to what is at stake."

Another case in the first category is B v Chief Constable of Avon and Somerset Constabulary [2001] 1 WLR 340, which concerned a "sex offender order" under section 2 of the Crime and Disorder Act 1998. Magistrates may make such an order if it is proved that a person is a sex offender and has acted in such a way as to give reasonable cause to believe that an order is necessary to protect the public from serious harm. The order may impose restrictions upon the person's freedom of movement and activity. Lord Bingham CJ said (at pp. 353-354) that the proceedings were civil in domestic law and for the purposes of the Convention but —

"…the civil standard of proof does not invariably mean a bare balance of probability, and does not mean so in the present case. The civil standard is a flexible standard to be applied with greater or lesser strictness according to the seriousness of what has to be proved and the implications of proving those matters: Bater v Bater [1951] P 35, Hornal v Neuberger Products Ltd [1957] 1 QB 247 and R v Secretary of State for the Home Department, ex parte Khawaja [1984] AC 74….In a serious case such as the present, the difference between the two standards is, in truth, largely illusory. I have no doubt that, in deciding whether the condition in section 2(1)(a) is fulfilled, a magistrates' court should apply a civil standard of proof which will for all practical purposes be indistinguishable from the criminal standard."

A similar point arose in R (McCann) v Crown Court at Manchester [2003] 1 AC 787, which concerned an anti-social behaviour order under section 1 of the 1998 Act. The House held that the proceedings were civil for the purposes of article 6 of the Convention. On the standard of proof, however, Lord Steyn said at para 37:

"I agree that, given the seriousness of matters involved, at least some reference to the heightened civil standard would usually be necessary: In re H (Minors)(Sexual Abuse: Standard of Proof) [1996] AC 563, 586D-H, per Lord Nicholls of Birkenhead. For essentially practical reasons, the Recorder of Manchester decided to apply the criminal standard. The Court of Appeal said that would usually be the right course to adopt. Lord Bingham of Cornhill has observed that the heightened civil standard and the criminal standard are virtually indistinguishable. I do not disagree with any of these views. But in my view pragmatism dictates that the task of magistrates should be made more straightforward by ruling that they must in all cases under section 1 apply the criminal standard."

The leading case in the second category was, until Re H (Minors) [1996] AC 563, the decision of the Court of Appeal in Hornal v Neuberger Products Ltd [1957] 1 QB 247. The question there was the appropriate standard of proof of an allegation of fraud in civil proceedings. In a frequently cited passage, Morris LJ said (at p. 266) that it was the normal standard for civil proceedings; proof on a balance of probability. But the gravity of an allegation of fraud was something which should be taken into account in deciding whether the burden had been discharged:

"Though no court and no jury would give less careful attention to issues lacking gravity than to those marked by it, the very elements of gravity become a part of the whole range of circumstances which have to be weighed in the scale when deciding as to the balance of probabilities."

It was this notion of having regard to inherent probabilities which Lord Nicholls of Birkenhead attempted to capture in In re H (Minors)(Sexual Abuse: Standard of Proof) [1996] AC 563, 586D-H:

"The balance of probability standard means that a court is satisfied an event occurred if the court considers that, on the evidence, the occurrence of the event was more likely than not. When assessing the probabilities the court will have in mind as a factor, to whatever extent is appropriate in the particular case, 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. Fraud is usually less likely than negligence. Deliberate physical injury is usually less likely than accidental physical injury. A step-father is usually less likely to have repeatedly raped and had non-consensual oral sex with his under age stepdaughter than on some occasion to have lost his temper and slapped her. Built into the preponderance of probability standard is a generous degree of flexibility in respect of the seriousness of the allegation.

Although the result is much the same, this does not mean that where a serious allegation is in issue the standard of proof required is higher. It means only that the inherent probability or improbability of an event is itself a matter to be taken into account when weighing the probabilities and deciding whether, on balance, the event occurred. The more improbable the event, the stronger must be the evidence that it did occur before, on the balance of probability, its occurrence will be established."

The degree of confusion which is possible on this issue is exemplified by the fact that despite the painstaking clarity with which Lord Nicholls explained that having regard to inherent probabilities did not mean that "where a serious allegation is in issue the standard of proof required is higher", Lord Steyn in R (McCann) v Crown Court at Manchester [2003] 1 AC 787, 812 cited this very passage as authority for the existence of a "heightened civil standard". This appears to have resulted in submissions that the Family Division should also apply a "heightened civil standard", equivalent to the criminal standard ("in serious cases such as the present, the difference between the two standards is, in truth, largely illusory", per Lord Bingham CJ in B v Chief Constable of Avon and Somerset Constabulary [2001] 1 WLR 340, 354), in local authority applications for care orders. Dame Elizabeth Butler-Sloss P restored clarity and certainty in re U (A Child) (Department for Education and Skills intervening) [2005] Fam 134,143-144:

"We understand that in many applications for care orders counsel are now submitting that the correct approach to the standard of proof is to treat the distinction between criminal and civil standards as 'largely illusory'. In our judgment this approach is mistaken. The standard of proof to be applied in Children Act 1989 cases is the balance of probabilities and the approach to these difficult cases was laid down by Lord Nicholls in In re H (Minors) (Sexual Abuse: Standard of Proof) [1996] AC 563. That test has not been varied nor adjusted by the dicta of Lord Bingham of Cornhill CJ or Lord Steyn who were considering applications made under a different statute. There would appear to be no good reason to leap across a division, on the one hand, between crime and preventative measures taken to restrain defendants for the benefit of the community and, on the other hand, wholly different considerations of child protection and child welfare nor to apply the reasoning in McCann's case [2003] 1 AC 787 to public, or indeed to private, law cases concerning children. The strict rules of evidence applicable in a criminal trial which is adversarial in nature is to be contrasted with the partly inquisitorial approach of the court dealing with children cases in which the rules of evidence are considerably relaxed. In our judgment therefore…the principles set out by Lord Nicholls should continue to be followed by the judiciary trying family cases and by magistrates sitting in the family proceedings courts."

My Lords, I would invite your Lordships fully to approve these observations. I think that the time has come to say, once and for all, that there is only one civil standard of proof and that is proof that the fact in issue more probably occurred than not. I do not intend to disapprove any of the cases in what I have called the first category, but I agree with the observation of Lord Steyn in McCann's case (at 812) that clarity would be greatly enhanced if the courts said simply that although the proceedings were civil, the nature of the particular issue involved made it appropriate to apply the criminal standard.

Finally, I should say something about the notion of inherent probabilities. Lord Nicholls said, in the passage I have already quoted, that —

"the court will have in mind as a factor, to whatever extent is appropriate in the particular case, 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."

I wish to lay some stress upon the words I have italicised. Lord Nicholls was not laying down any rule of law. There is only one rule of law, namely that the occurrence of the fact in issue must be proved to have been more probable than not. Common sense, not law, requires that in deciding this question, regard should be had, to whatever extent appropriate, to inherent probabilities. If a child alleges sexual abuse by a parent, it is common sense to start with the assumption that most parents do not abuse their children. But this assumption may be swiftly dispelled by other compelling evidence of the relationship between parent and child or parent and other children. It would be absurd to suggest that the tribunal must in all cases assume that serious conduct is unlikely to have occurred. In many cases, the other evidence will show that it was all too likely. If, for example, it is clear that a child was assaulted by one or other of two people, it would make no sense to start one's reasoning by saying that assaulting children is a serious matter and therefore neither of them is likely to have done so. The fact is that one of them did and the question for the tribunal is simply whether it is more probable that one rather than the other was the perpetrator.

For the reasons given by Baroness Hale of Richmond, I would make the order she proposes.

LORD SCOTT OF FOSCOTE

My Lords,

I have had the advantage of reading in draft the opinion prepared by my noble and learned friend Baroness Hale of Richmond. Having done so I find myself wholly convinced by her analysis of the legal issues and the conclusions she has reached, and convinced also that any attempt by me to add anything of my own would no more than muddy the waters that she has left so admirably limpid. For the reasons she has given I too would dismiss this appeal.

LORD RODGER OF EARLSFERRY

My Lords,

I have had the privilege of considering the speeches of my noble and learned friends, Lord Hoffmann and Baroness Hale of Richmond, in draft. For the reasons which they give, I too would dismiss the appeal and make the order which Baroness Hale proposes.

LORD WALKER OF GESTINGTHORPE

My Lords,

I have had the privilege of reading in draft the opinions of my noble and learned friends Lord Hoffmann and Baroness Hale of Richmond. I am in complete agreement with them, and I would make the order that Baroness Hale proposes.

BARONESS HALE OF RICHMOND

My Lords,

Taking a child away from her family is a momentous step, not only for her, but for her whole family, and for the local authority which does so. In a totalitarian society, uniformity and conformity are valued. Hence the totalitarian state tries to separate the child from her family and mould her to its own design. Families in all their subversive variety are the breeding ground of diversity and individuality. In a free and democratic society we value diversity and individuality. Hence the family is given special protection in all the modern human rights instruments including the European Convention on Human Rights (art 8), the International Covenant on Civil and Political Rights (art 23) and throughout the United Nations Convention on the Rights of the Child. As Justice McReynolds famously said in Pierce v Society of Sisters 268 US 510 (1925), at 535, "The child is not the mere creature of the State".

That is why the Review of Child Care Law (Department of Health and Social Security, 1985)) and the white paper, The Law on Child Care and Family Services (Cm 62, 1987), which led up to the Children Act 1989, rejected the suggestion that a child could be taken from her family whenever it would be better for her than not doing so. As the Review put it, "Only where their children are put at unacceptable risk should it be possible compulsorily to intervene. Once such a risk of harm has been shown, however, [the child's] interests must clearly predominate" (para 2.13).

The principle of "unacceptable risk of harm" is easy enough to state but difficult to put into statutory language. The draft Children Bill annexed to the Law Commission's Report on its Review of Child Law, Guardianship and Custody (1988, Law Com No 172, BAILII: [1988] EWLC 172 ) required that "the child concerned has suffered significant harm, or that there is a real risk of his suffering such harm" (clause 12(2)(a)). This was refined in the Bill presented to Parliament and eventually emerged in the so-called "threshold criteria" in section 31(2) of the Children Act 1989:

"A court may only make a care order or a 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."

This case is about the meaning of the words "is likely to suffer significant harm". How is the court to be satisfied of such a likelihood? This is a prediction from existing facts, often from a multitude of such facts, about what has happened in the past, about the characters and personalities of the people involved, about the things which they have said and done, and so on. But do those facts have to be proved in the usual way, on the balance of probabilities? Or is it sufficient that there is a "real possibility" that they took place, even if the judge is unable to say that it is more likely than not that they did?

That issue was authoritatively determined in favour of the former solution by a majority of this House in In re H (Minors) (Sexual Abuse: Standard of Proof) [1996] AC 563. It was reaffirmed in two later decisions: see Lancashire County Council v B [2000] 2 AC 147 and In re O (minors) (Care Proceedings: Preliminary Hearing) [2004] 1 AC 523. The latter also supported, albeit obiter, the Court of Appeal decision in In re M and R (Child Abuse: Evidence) [1996] 2 FLR 195. There, the Court of Appeal had applied the same approach to the words "and harm which he has suffered or is at risk of suffering", a factor which, under section 1(3)(e) of the 1989 Act the court has to take into account when deciding what order, if any, will be best for the child.

In this case, the children's guardian, with the support of the local authority and the children's mother, seeks to over-turn the decision in In re H and over-rule In re M and R, in favour of a "real possibility" test. The children's father, with the support of the interveners, the Children and Family Court Advisory and Support Service (Cafcass), seeks to uphold the present law.

The factual background

As the case is still continuing, the least said about the factual background the better. It concerns the future of two children, a nine year old girl, NB, and a six year old boy, AB. Their parents are Mr and Mrs B, who began their relationship in 1996 and married in 1999. Mrs B has two children by her previous marriage, a 16 year old girl, R, and a 17 year old boy, S. They all lived together until April 2006, when Mr B left the family home, although he later visited from time to time.

Social services and the police became involved with the family shortly afterwards. N's school became concerned about her possibly sexualised behaviour. In July 2006, Mrs B undoubtedly assaulted S in the street. Following that S left the family home and has not returned. Mrs B threatened to allege that S had sexually assaulted N if he persisted in an assault charge against her. Mrs B and R then both made false allegations of sexual abuse and assault against S. The police and social services began their inquiries.

Those inquiries were still continuing on 30 October 2006, when Mr B applied, with the support of the local authority, for residence orders in respect of N and A. Instead, the district judge made interim care orders in respect of R as well as N and A, on the basis of a plan to remove them all from Mrs B and place them with Mr B at his parents' home. However, while they were being removed from home that evening, R alleged that Mr B had sexually abused her and had also assaulted her and S with a belt. R was placed with foster carers and has since returned to her mother's care. N and A were placed with Mr B's parents and Mr B moved out. In September 2007 they were moved to foster carers, where they remain.

The care proceedings were transferred to the High Court. A fact finding hearing took place before Charles J over 29 days in June and July 2007. As is now the practice, the local authority set out a statement of the basis upon which it was alleged that the section 31(2) criteria were met. The parents made a number of concessions but there remained a number of matters in dispute, amongst them R's allegations of sexual abuse against Mr B. No-one is now persisting in the suggestion that S had sexually abused N or anyone else.

On 19 October 2007, Charles J handed down a judgment comprising 347 substantive paragraphs and 8 schedules, totalling 159 pages in all. He made detailed findings of physical and emotional abuse supporting his overall conclusion that:

"The combination of the physical and verbal violence, abuse and relationships within the household and the antagonism both between members of the household, and them and persons outside it, had the result that the children were being brought up in a fraught household in which they had no realistic prospect of being able to develop as young children, and then as teenagers, into adults against a background of emotional and physical stability or in which balanced and reasonable approaches were taken to the events of every day life." (para 48)

Important components in that conclusion were his findings about Mrs B's aggressive and bullying behaviour in her dealings with the children's schools and others and the existence of an "allegation culture" within the family for which Mrs B was primarily responsible:

" . . . the family have an allegation culture in which all four of Mrs B, Mr B, S and R have made and supported allegations that are simply untrue, or allegations that are so exaggerated and misrepresented that they become untrue, to promote a campaign against others or to get back at them." (para 133)

However, despite an elaborate and meticulous analysis of all the evidence, the learned judge was unable to make a finding about the alleged sexual abuse of R by Mr B. Instead he concluded that:

"(i) I cannot make a properly founded and reasoned conclusion that it is more likely than not that R was sexually abused by Mr B as she alleges or substantially as she alleges, and thus that she is telling the truth,

(ii) I cannot make a properly founded and reasoned conclusion that it is more likely than not that R was not sexually abused by Mr B, and thus that Mr B is telling the truth,

(iii) my answer to the question which of the above two possibilities (and thus which of Mr B and R is telling the truth) is more likely, would be a guess because I cannot even answer that question by attributing and giving weight to the competing arguments on a properly founded and reasoned basis, and

(iv) on an approach founded on evidence and reasoning, and not on suspicion and/or concern, I am unable to conclude that there is no real possibility that Mr B sexually abused R as she asserts or substantially as she asserts and I have therefore concluded that there is a real possibility that he did." (para 339)

My Lords, if the judiciary in this country regularly found themselves in this state of mind, our civil and family justice systems would rapidly grind to a halt. In this country we do not require documentary proof. We rely heavily on oral evidence, especially from those who were present when the alleged events took place. Day after day, up and down the country, on issues large and small, judges are making up their minds whom to believe. They are guided by many things, including the inherent probabilities, any contemporaneous documentation or records, any circumstantial evidence tending to support one account rather than the other, and their overall impression of the characters and motivations of the witnesses. The task is a difficult one. It must be performed without prejudice and preconceived ideas. But it is the task which we are paid to perform to the best of our ability.

In our legal system, if a judge finds it more likely than not that something did take place, then it is treated as having taken place. If he finds it more likely than not that it did not take place, then it is treated as not having taken place. He is not allowed to sit on the fence. He has to find for one side or the other. Sometimes the burden of proof will come to his rescue: the party with the burden of showing that something took place will not have satisfied him that it did. But generally speaking a judge is able to make up his mind where the truth lies without needing to rely upon the burden of proof.

The judge's findings in this case were expressed in such a way as squarely to raise the issue of principle. Is it possible to be satisfied that a child is likely to suffer a particular kind of harm in the future when the basis for suggesting this is that there is a "real possibility" that another child has suffered the same kind of harm in the past? There are, of course, many degrees of possibility - from a fifty/fifty chance that it happened down to an infinitesimal chance that it did. In this case, the judge seems to have concluded that there was a "real possibility" because he could not conclude that there was none.

Having set the case up in such a way as to raise the issue of principle, the judge further elaborated upon the problem as he saw it in Schedule A to his principal judgment, which was handed down with a further judgment containing a draft letter of instruction to the experts who were to conduct assessments of the parents and the family for the next stage in the proceedings on 11 December 2007. He had indicated that he considered the case suitable for a "leapfrog" appeal to this House but Mr B did not agree to this. Accordingly Charles J gave leave to appeal to the Court of Appeal. The Court of Appeal, being bound on the authorities to dismiss the guardian's appeal, gave leave to appeal to this House.

The authorities

In In re H and others (Minors) (Sexual Abuse: Standard of Proof) [1996] AC 563, the facts were unusually simple. The mother had four daughters, two by her husband and two by R, the man with whom she was now living. The eldest daughter alleged that R had been sexually abusing her for some years. The local authority brought care proceedings in respect of the three younger children, relying solely on these allegations as proof of the likelihood of similar harm to them. The judge was not satisfied that the allegations were true, although there was a real possibility that they were. He dismissed the local authority's applications.

The local authority's appeal to this House established three quite separate propositions. The first, on which all five of their lordships were agreed, was that the words "is likely to suffer significant harm" did not mean that such harm had to be more likely than not to happen in the future: it was enough if its happening was 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" (Lord Nicholls of Birkenhead at p 585).

The second, on which all were also agreed, was that the standard of proof of facts in issue was the balance of probabilities; but there was a difference between the ways in which that standard was expressed, on the one hand by Lord Nicholls with the majority, and on the other hand by Lord Lloyd of Berwick. That issue has not only led to considerable confusion in the past but is also the source of some of the perceived difficulties with the present law. Miss Jo Delahunty QC, on behalf of Cafcass, has made an eloquent plea to us to clarify it for the future.

The third issue, on which Lord Nicholls, with whom Lord Goff and Lord Mustill agreed, took one view and Lord Browne Wilkinson and Lord Lloyd took another, was the issue which is raised in this case. Lord Nicholls' conclusion was that a conclusion as to future risk had to be based on facts:

"In my view these unresolved judicial doubts and suspicions can no more form the basis of a conclusion that the second threshold condition in section 31(2)(a) has been established than they can form the basis that the first has been established." (p 589E)

He gave several reasons for arriving at that conclusion. He pointed to the distinction between interlocutory relief, which can be granted on the basis of a good arguable case without resolving disputed questions of fact, and a trial, in which "the court normally has to resolve disputed issues of relevant fact before it can reach its conclusion on the issue it has to decide". (p 589G) "A decision by a court on the likelihood of a future happening must be founded on a basis of present facts and the inferences fairly to be drawn therefrom." (p 590A) For the first limb of section 31(2)(a), "There must be facts, proved to the court's satisfaction if disputed, on which the court can properly conclude that the child is suffering harm". (p 590C) "Similarly, with the second limb: there must be facts from which the court can properly conclude there is a real possibility that the child will suffer harm in the future." (p 590C)

He found several indications in the Act that this, the ordinary approach, was to be applied. First, when dealing with investigations and child assessment orders, the Act uses the term "reasonable cause to suspect" (ss 47(1)(b) and 43(1)(a)), and when dealing with emergency protection orders, police protection and interim care orders, it uses the term "reasonable cause to believe" (ss 44(1)(a), 46(1) and 38(2)), that the child is suffering or likely to suffer significant harm. This is the sensible approach to child protection before the stage of a final order is reached. But the language of section 31(2) is different: the court must be "satisfied . . . that the child . . . is suffering, or is likely to suffer, significant harm; . . ." "This is the language of proof, not suspicion, however reasonably based." (p 590H)

Second, the second threshold condition, likelihood of harm, is "cheek by jowl" with the first, that the child is suffering harm. If the evidence of maltreatment is not sufficient to establish that the child is suffering harm, "It would be odd if, in respect of the selfsame non-proven allegations, the self-same insufficient evidence could nonetheless be regarded as a sufficient factual basis for satisfying the court there is a real possibility of harm to the child in the future". (p 591B)

Third, if this were the case, it "would effectively reverse the burden of proof in an important respect". Once apparently credible evidence of misconduct had been given, those against whom the allegations were made would have to disprove them. "Otherwise it would be open to a court to hold that, although the misconduct has not been proved, it has not been disproved and there is a real possibility that the misconduct did occur. . . . I do not believe Parliament intended that section 31(2) should work in this way." (p591D)

This is, of course, exactly what the judge found in our case and exactly what the children's guardian says should be enough to cross the threshold. One reason for this is the alleged inconsistency between the approach taken in Re H and the approach taken by the House in the two later cases. In each of them, it was clear that the children involved had suffered significant harm but it was not clear who had been the perpetrator. The point arose in particularly striking circumstances in Lancashire County Council v B [2000] 2 AC 147. A seven month old baby, A, lived with her parents but was cared for by a child minder during working hours. She suffered very significant harm as a result of violent shaking on at least two occasions. The local authority brought care proceedings, not only in respect of child A, but also in respect of the childminder's son, B, who was a month older than the injured child. The possible perpetrators, as found by the judge, were A's mother and father or the childminder, B's mother (he was able to exonerate B's father). But the injuries could have happened in either household; he could not say in which or who was the likely perpetrator.

The judge dismissed the applications relating to each child. He could not be satisfied that harm suffered by child A was attributable to a lack of reasonable care on the part of the parent against whom the order was sought, nor could he be satisfied of the likelihood of future harm to child B attributable to a lack of reasonable care by his mother. The Court of Appeal held that the judge was plainly right in relation to child B. There was no evidence that he had been harmed in any way. It had not been proved that his mother was responsible for A's injuries and that was the only basis for suggesting that there was any risk of harm to B in the future. In re H applied. Child A, on the other hand, had undoubtedly been harmed. This was not an accident. It was attributable to a lack of proper care. It was not necessary, in order to establish the criterion in section 31(2)(b)(i) (see para 22 above), to show who was responsible for that lack. The local authority's appeal was allowed in relation to her. The parents' appeal to the House of Lords was dismissed.

The question was whether the "care given to the child", which had been found to be deficient, meant the care given to the child by her parents or other primary carers, as argued on behalf of the parents, or whether it meant the care given by anyone who plays a part in the care arrangements for the child, as argued by the local authority and the child's guardian. Lord Nicholls found both extremes produced unacceptable results. He favoured a middle course:

"The phrase 'care given to the child' refers primarily to the care given to the child by a parent or parents or other primary carers. That is the norm. . . Different considerations from the norm apply in a case of shared caring where the care given by one or other of the carers is proved to have been deficient, with the child suffering harm in consequence, but the court is unable to identify which of the carers provided the deficient care. In such a case the phrase 'care given to the child' is apt to embrace not merely the care given by the parents or other primary carers; it is apt to embrace the care given by any of the carers." (p 166 C-D)

He recognised that this construction meant that the conditions might be satisfied when there was no more than a possibility that the parents were responsible and that parents who might be wholly innocent would face the possibility of losing their child. But to hold otherwise

"would mean that the child's future health, or even her life, would have to be hazarded on the chance that, after all, the non-parental carer rather than one of the parents inflicted the injuries. Self-evidently, to proceed in such a way when a child is proved to have suffered serious injury on more than one occasion could be dangerously irresponsible." (p 165G)

The more common version of this dilemma, however, is not where a child's care is shared between two households, but where it is shared between two parents. If the child suffers harm, and the judge cannot decide which parent was responsible, the threshold criteria are met. But how is the court to approach the next stage in the proceedings, the stage of deciding what order, if any, will be in the best interests of the child? In In re O (Minors)(Care: Preliminary Hearing); In re B (A Minor) [2003] UKHL 18, [2004] 1 AC 523, the Court of Appeal in one case had held that the judge had to proceed on the basis that the child had not been harmed by the mother and that she did not present a risk of harm to that or another child; in the other, a differently constituted Court of Appeal had held that as the mother had not been exonerated, the judge could not disregard the risk that she might present.

Lord Nicholls (with whom Lord Hoffmann, Lord Millett, Lord Scott and Lord Walker simply agreed) preferred the latter view:

"Quite simply, 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. This would be grotesque because it would mean the court would proceed on the footing that neither parent represents a risk even though one or other of them was the perpetrator of the harm in question. . . .

. . . The preferable interpretation of the legislation is that in such cases the court is able to proceed at the welfare stage on the footing that each of the possible perpetrators is, indeed, just that: a possible perpetrator." (paras 27, 28)

However, no doubt was cast on the conclusions reached in In re H. Lord Nicholls went on to consider how unproven allegations of harm should be treated at the "welfare stage" in care proceedings. Once the threshold in section 31(2) has been crossed, the court is required to apply the welfare principle in section 1(1) of the 1989 Act:

"When a court determines any question with respect to -

(a) the upbringing of a child; or

(b) the administration of a child's property or the application of any income arising from it,

the child's welfare shall be the court's paramount consideration."

The court is also required to have regard, in particular, to the "checklist" of factors set out in section 1(3) of the Act. These include, along with such obvious matters as the wishes and feelings of the child, and the capacities of the adults around her to meet her needs, in section 1(3)(e), "any harm which he has suffered or is at risk of suffering".

Is section 1(3)(e) to be interpreted and applied differently from section 31(2)(a)? Can a court conclude that there is a risk of the child suffering a particular kind of harm even though the allegations said to give rise to such a risk cannot be proved? The question did not arise directly in In re O, but Lord Nicholls said this:

"On balance, I consider that to have regard at the welfare stage to allegations of harm rejected at the threshold stage would have the effect of depriving the child and the family of the protection intended to be afforded by the threshold criteria. Accordingly, at the welfare stage in this type of case the court should proceed on the footing that the unproven allegations are no more than that." (para 38).

This accorded with the approach of the Court of Appeal in In re M and R (Minors) (Abuse: Expert Evidence) [1996] 4 All ER 239. There, the judge was satisfied that the threshold was crossed on the basis that the children had suffered emotional abuse and were likely to do so in future. He was not, however, satisfied that sexual abuse had also occurred, although there was a real possibility that it had. The local authority contended that he should have taken this into account under section 1(3)(e) when assessing the welfare of the children. Butler-Sloss LJ, giving the judgment of the court, rejected their submissions at p246H to 247D:

"They amount to the assertion that under section 1 the welfare of the child dictates that the court should act on suspicion or doubts, rather than facts. To our mind the welfare of the child dictates the exact opposite. . .

If there is a dispute as to whether the child has suffered or is at risk of suffering harm, the task of the judge, when considering whether to make any order, whether it be a care or supervision order under section 31 or a section 8 order (residence, contact and other orders with respect to children), must be to resolve that dispute. . . The question is how such a dispute is to be resolved.

To our minds there can be only one answer to this question, namely the same answer as given by the majority in In re H. The court must reach a conclusion based on facts, not on suspicion or mere doubts. ...

[Counsel's] point was that if there is a real possibility of harm in the past, then it must follow (if nothing is done) that there is a risk of harm in the future. To our minds, however, this proposition contains a non sequitur. The fact that there might have been harm in the past does not establish the risk of harm in the future. The very highest it can be put is that what might possibly have happened in the past means that there may possibly be a risk of the same thing happening in the future. Section 1(3)(e), however, does not deal with what might possibly have happened or what future risk there may possibly be. It speaks in terms of what has happened or what is at risk of happening. Thus, what the court must do (when the matter is in issue) is to decide whether the evidence establishes the harm or the risk of harm"

In re M and R was a care case but, in a private law dispute between mother and father over contact, the Court of Appeal had a month earlier (and in the knowledge the judgment in In re M and R was pending) pointed to the judge's "fundamental error" in exercising his discretion as to the future on the basis of a finding that there was "a substantial risk" that abuse had occurred in the past. In Re P (Sexual Abuse: Standard of Proof) [1996] 2 FLR 333, 343 Wall J commented that:

"It has also had the effect, in the instant case, of producing the worst of all worlds. The father remains under a cloud. Abuse is not proved on the balance of probabilities, but he remains effectively branded an abuser: as the judge himself said, 'at the very lowest he will remain under suspicion until his daughters are old enough to be able to cope with any risk of abuse themselves'. Furthermore, the mother's beliefs are reinforced. It thus becomes impossible for the parties and the children to put the issue of sexual abuse behind them. The end result is highly unsatisfactory."

The children's guardian also invites us to overrule the Court of Appeal's decision in In re M and R, so that, the threshold having been crossed on other grounds, the judge will be able to take into account at the welfare stage the unproven allegations of sexual abuse against Mr B.

This appeal

Mr Stephen Cobb QC, on behalf of the children's guardian who represents N and A, invites us to depart from In re H and to overrule In re M and R principally on the ground that, in combination with Lancashire County Council v B and In re O, they produce illogical results. If a parent can be deprived of her child, or a child deprived of her mother, on the basis of a real possibility that she may have been the perpetrator of the harm suffered by the child, why should not such a real possibility that harm has been suffered in the past be the basis of a finding that it is likely that such harm will be suffered in the future? The artificiality of proceeding on the basis that such harm did not happen at all, when there is a real possibility that it did, is just as irresponsible and dangerous as proceeding on the basis that neither parent was the perpetrator, rejected by this House in In re O.

My Lords, I would unhesitatingly decline that invitation. The reasons given by Lord Nicholls for adopting the approach which he did in Re H remain thoroughly convincing. The threshold is there to protect both the children and their parents from unjustified intervention in their lives. It would provide no protection at all if it could be established on the basis of unsubstantiated suspicions: that is, where a judge cannot say that there is no real possibility that abuse took place, so concludes that there is a real possibility that it did. In other words, the alleged perpetrator would have to prove that it did not. Mr Cobb accepts that it must be proved on the balance of probabilities that a child "is suffering" significant harm. But nevertheless he argues that those same allegations, which could not be proved for that purpose, could be the basis of a finding of likelihood of future harm. If that were so, there would have been no need for the first limb of section 31(2)(a) at all. Parliament must be presumed to have inserted it for a purpose. Furthermore, the Act draws a clear distinction between the threshold to be crossed before the court may make a final care or supervision order and the threshold for making preliminary and interim orders. If Parliament had intended that a mere suspicion that a child had suffered harm could form the basis for making a final order, it would have used the same terminology of "reasonable grounds to suspect" or "reasonable grounds to believe" as it uses elsewhere in the Act. Instead, as Butler-Sloss LJ pointed out in In re M and R, it speaks of what the child is suffering or is likely to suffer.

My Lords, it is rare for the facts to be as stark as they were in In re H. There are usually many facts from which an inference that a child is likely to suffer harm in the future can be drawn. In In re H it appears that there was nothing, other than the allegations of sexual abuse made by the oldest child, from which it would have been possible to draw the inference that the other children were likely to suffer such harm. In the air-raid example used by Lord Browne Wilkinson in In re H, it would have been difficult to conclude that an air raid was likely from nothing more than that five unidentified planes had been seen in the skies overhead. But the country was at war. Air raids were frequent. The fact that there had been raids in the past would make it possible to draw the inference that there would be raids in the future. The only question was when.

But in a case such as this, as indicated by Butler-Sloss LJ in In re M and R, the "risk" is not an actual risk to the child but a risk that the judge has got it wrong. We are all fallible human beings, very capable of getting things wrong. But until it has been shown that we have, it has not been shown that the child is in fact at any risk at all.

It is also important to keep separate the roles of the courts and the local authorities in the protection of children from harm. Where a local authority have reasonable cause to suspect that a child in their area is suffering or likely to suffer significant harm, they must make the inquiries necessary to enable them to decide whether they should take any action to protect the child and if so what (1989 Act, s 47(1)). This is done by way of a core assessment, conducted in accordance with the Framework for the Assessment of Children in Need and Their Families (Department of Health and others, 2000). This is "an in-depth assessment which addresses the central or most important aspects of the needs of a child and the capacity of his or her parents or caregivers to respond appropriately to those needs within the wider family and community context" (para 3.11). As such, it will clearly range far wider than the threshold criteria. It will form "a central part of the evidence supporting any application that the local authority may make for a care or supervision order" (Department for Children, Schools and Families, Children Act 1989, Guidance and Regulations, Volume 1, Court Orders, 2008, para 3.17). It will also form the basis for the plan for the future care of the child which the local authority must put before the court under section 31A of the 1989 Act (ibid, para 3.18).

The local authority make the application for a care or supervision order under section 31(1) and the local authority will be responsible for carrying out any order which the court may make. The task of the court is to hear the evidence put forward on behalf of all the parties to the case and to decide, first, whether the threshold criteria are met and, second, what order if any will be best for the child. While the local authority may well take preliminary or preventive action based upon reasonable suspicions or beliefs, it is the court's task when authorising permanent intervention in the legal relationship between parent and child to decide whether those suspicions are well-founded. As the Review of Child Care Law (1985, para 2.20) put it,

"One of our guiding principles has been that the court should be able to determine major issues such as the transfer of parental rights and duties where there is or may be a dispute between parents and local authorities, while the management of the case should be the responsibility of the local authority".

To allow the courts to make decisions about the allocation of parental responsibility for children on the basis of unproven allegations and unsubstantiated suspicions would be to deny them their essential role in protecting both children and their families from the intervention of the state, however well intentioned that intervention may be. It is to confuse the role of the local authority, in assessing and managing risk, in planning for the child, and deciding what action to initiate, with the role of the court in deciding where the truth lies and what the legal consequences should be. I do not under-estimate the difficulty of deciding where the truth lies but that is what the courts are for.

I am fortified in that conclusion by two things. The first is that Cafcass does not support the guardian's stance in this appeal. Cafcass is the body responsible for safeguarding the interests of children in the family courts. It appoints the individual guardians. If Cafcass thought that the decision in In re H was causing serious difficulties and jeopardising the welfare of our most vulnerable children, no doubt it would have said so. The second is that Parliament has recently had the opportunity of reviewing the 1989 Act in the light of the inquiry into the tragic death of Victoria Climbié (The Victoria Climbié Inquiry, Report of an Inquiry by Lord Laming, 2003, Cm 5730). Children's services have been thoroughly reorganised by the Children Act 2004, but no change has been made to the Act's substantive provisions.

The decisions in In re H, Lancashire County Council v B, and In re O fit together as a coherent whole. The court must first be satisfied that the harm or likelihood of harm exists. Once that is established, as it was in both the Lancashire and Re O cases, the court has to decide what outcome will be best for the child. It is very much easier to decide upon a solution if the relative responsibility of the child's carers for the harm which she or another child has suffered can also be established. But the court cannot shut its eyes to the undoubted harm which has been suffered simply because it does not know who was responsible. The real answers to the dilemma posed by those cases lie elsewhere - first, in a proper approach to the standard of proof, and second, in ensuring that the same judge hears the whole case. Split hearings are one thing; split judging is quite another.

The standard of proof

All of their Lordships in In re H were clear that there was one standard of proof, the balance of probabilities. But Lord Nicholls went on to say this at p586:

"The balance of probability standard means that a court is satisfied an event occurred if the court considers that, on the evidence, the occurrence of the event was more likely than not. When assessing the probabilities the court will have in mind as a factor, to whatever extent is appropriate in the particular case, 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. Fraud is less likely than negligence. Deliberate physical injury is usually less likely than accidental physical injury. A step-father is usually less likely to have repeatedly raped and had non-consensual oral sex with his under age stepdaughter than on some occasion to have lost his temper and slapped her. Built into the preponderance of probability standard is a generous degree of flexibility in respect of the seriousness of the allegation.

Although the result is much the same, this does not mean that where a serious allegation is in issue the standard of proof required is higher. It means only that the inherent probability or improbability of an event is itself a matter to be taken into account when weighing the probabilities and deciding whether, on balance, the event occurred. The more improbable the event, the stronger must be the evidence that it did occur…. Ungoed-Thomas J expressed this neatly in In re Dellow's Will Trusts [1964] 1 WLR 451, 455: 'The more serious the allegation the more cogent is the evidence required to overcome the unlikelihood of what is alleged and thus to prove it.'" emphasis supplied)

If he had stopped there, perhaps there would have been no difficulty, provided that lawyers and courts paid attention to the whole passage, including the words which I have italicised, rather than extracting a single phrase. But he went on:

"This substantially accords with the approach adopted in authorities such as the well known judgment of Morris LJ in Hornal v Neuberger Products Ltd [1957] 1 QB 247, 266. This approach also provides a means by which the balance of probability standard can accommodate one's instinctive feeling that even in civil proceedings a court should be more sure before finding serious allegations proved than when deciding less serious or trivial matters."

"More sure" may be read as suggesting a higher standard than the simple preponderance of probabilities.

Lord Lloyd, at pp 577-578 on the other hand, took a more straightforward line:

"In my view the standard of proof under [section 31(2)] ought to be the simple balance of probability however serious the allegations involved. . . . mainly because section 31(2) provides only the threshold criteria for making a care order. . . if the threshold criteria are not met, the local authority can do nothing, however grave the anticipated injury to the child, or however serious the apprehended consequences. This seems to me to be a strong argument in favour of making the threshold lower rather than higher. It would be a bizarre result if the more serious the anticipated injury, whether physical or sexual, the more difficult it became for the local authority to satisfy the initial burden of proof, and thereby ultimately, if the welfare test is satisfied, secure protection for the child. . .

There remains the question whether anything should be said about the cogency of the evidence needed to 'tip the balance'. For my part I do not find those words helpful, since they are little more than a statement of the obvious; and there is a danger that the repeated use of the words will harden into a formula which, like other formulas (especially those based on a metaphor) may lead to misunderstanding."

My Lords, Lord Lloyd's prediction proved only too correct. Lord Nicholls' nuanced explanation left room for the nostrum, "the more serious the allegation, the more cogent the evidence needed to prove it", to take hold and be repeated time and time again in fact-finding hearings in care proceedings (see, for example, the argument of counsel for the local authority in Re U (A Child) (Department for Education and Skills intervening) [2004] EWCA Civ 567, [2005] Fam 134, at p 137. It is time for us to loosen its grip and give it its quietus.

Indeed, later events made matters worse. In B v Chief Constable of the Avon and Somerset Constabulary [2001] 1 WLR 340, the issue was the standard of proof to be applied when finding the facts needed to make a sex offender order under section 2 of the Crime and Disorder Act 1998. The Court of Appeal held that these were civil proceedings, but Lord Bingham of Cornhill CJ said this about the standard of proof:

"30. It should, however, be clearly recognised, as the justices did expressly recognise, that the civil standard of proof does not invariably mean a bare balance of probability, and does not so mean in the present case. The civil standard is a flexible standard to be applied with greater or lesser strictness according to the seriousness of what has to be proved and the implications of proving those matters (see Bater v Bater [1951] P 35, Hornal v Neuberger Products Ltd [1957] 1 QB 247, and R v Secretary of State for the Home Department, ex parte Khawaja [1984] AC 74).

31. In a serious case such as the present the difference between the two standards is, in truth, largely illusory. . . ."

In re H was neither referred to nor cited in that case, but of course the link could be made through the references to Hornal v Neuberger Products. However, In re H was cited in R (McCann and others) v Crown Court at Manchester [2002] UKHL 39, [2003] 1 AC 787. One issue was the standard of proof in finding the facts needed to make an anti-social behaviour order under section 1 of the Crime and Disorder Act 1998. Lord Steyn said this:

"37. Having concluded that the relevant proceedings are civil, in principle it follows that the standard of proof ordinarily applicable in civil proceedings, namely the balance of probabilities, should apply. However, I agree that, given the seriousness of matters involved, at least some reference to the heightened civil standard would usually be necessary (see Re H(minors)(sexual abuse: standard of proof) [1996] AC 563, 586D-H per Lord Nicholls of Birkenhead). . . . Lord Bingham of Cornhill has observed that the heightened civil standard and the criminal standard are virtually indistinguishable. I do not disagree with any of those views."

The House went on to hold that in anti-social behaviour order proceedings the court should apply the criminal standard of proof.

The link with In re H having been made, it is not surprising that judges should think that the same "heightened" standard should also apply in care proceedings. In In re ET (Serious Injuries: Standard of Proof) [2003] 2 FLR 1205, Bodey J directed himself that, in applying the civil standard and "the Re H cogency test", he would have well in mind the dicta in B v Chief Constable of Avon and R (McCann) v Chief Constable of Manchester. So, if he found any facts it would be "on the basis that, in this very serious case, the difference between the civil and criminal standards of proof is 'largely illusory'" (para 6) And who, in the light of the passages quoted above, can blame him?

Fortunately for care proceedings, the status quo was restored by the Court of Appeal in In re U (A Child) (Department for Education and Skills intervening); In re B (A Child) (Department for Education and Skills Intervening) [2004] EWCA Civ 567, [2005] Fam 134. The issue was the approach to be taken to medical evidence in care proceedings following the decision of the Court of Appeal (Criminal Division) in R v Cannings [2004] 1 WLR 2607. Dame Elizabeth Butler-Sloss P, giving the judgment of the court, said this:

"13. We understand that in many applications for care orders counsel are now submitting that the correct approach to the standard of proof is to treat the distinction between criminal and civil standards as 'largely illusory'. In our judgment this approach is mistaken. The standard of proof to be applied in Children Act 1989 cases is the balance of probabilities and the approach to these difficult cases was laid down by Lord Nicholls in In re H (Minors)(Sexual Abuse: Standard of Proof) [1996] AC 563. That test has not been varied or adjusted by the dicta of Lord Bingham of Cornhill CJ or Lord Steyn who were considering applications made under a different statute. There would appear to be no good reason to leap across a division, on the one hand, between crime and preventive measures taken to restrain defendants for the benefit of the community and, on the other hand, wholly different considerations of child protection and child welfare . . . "

My Lords, I entirely agree. There are some proceedings, though civil in form, whose nature is such that it is appropriate to apply the criminal standard of proof. Divorce proceedings in the olden days of the matrimonial "offence" may have been another example (see Bater v Bater [1951] P 35). But care proceedings are not of that nature. They are not there to punish or to deter anyone. The consequences of breaking a care order are not penal. Care proceedings are there to protect a child from harm. The consequences for the child of getting it wrong are equally serious either way.

My Lords, for that reason I would go further and announce loud and clear that the standard of proof in finding the facts necessary to establish the threshold under section 31(2) or the welfare considerations in section 1 of the 1989 Act is the simple balance of probabilities, neither more nor less. Neither the seriousness of the allegation nor the seriousness of the consequences should make any difference to the standard of proof to be applied in determining the facts. The inherent probabilities are simply something to be taken into account, where relevant, in deciding where the truth lies.

As to the seriousness of the consequences, they are serious either way. A child may find her relationship with her family seriously disrupted; or she may find herself still at risk of suffering serious harm. A parent may find his relationship with his child seriously disrupted; or he may find himself still at liberty to maltreat this or other children in the future.

As to the seriousness of the allegation, there is no logical or necessary connection between seriousness and probability. Some seriously harmful behaviour, such as murder, is sufficiently rare to be inherently improbable in most circumstances. Even then there are circumstances, such as a body with its throat cut and no weapon to hand, where it is not at all improbable. Other seriously harmful behaviour, such as alcohol or drug abuse, is regrettably all too common and not at all improbable. Nor are serious allegations made in a vacuum. Consider the famous example of the animal seen in Regent's Park. If it is seen outside the zoo on a stretch of greensward regularly used for walking dogs, then of course it is more likely to be a dog than a lion. If it is seen in the zoo next to the lions' enclosure when the door is open, then it may well be more likely to be a lion than a dog.

In the context of care proceedings, this point applies with particular force to the identification of the perpetrator. It may be unlikely that any person looking after a baby would take him by the wrist and swing him against the wall, causing multiple fractures and other injuries. But once the evidence is clear that that is indeed what has happened to the child, it ceases to be improbable. Some-one looking after the child at the relevant time must have done it. The inherent improbability of the event has no relevance to deciding who that was. The simple balance of probabilities test should be applied.

Split hearings

Care proceedings are not a two stage process. The court does have two questions to ask. Has the threshold been crossed? If so, what will be best for the child? But there are many cases in which a court has two or more questions to ask in the course of a single hearing. The same factual issues are often relevant to each question. Or some factual disputes may be relevant to the threshold while others are relevant to the welfare checklist: it may be clear, for example, that a child has suffered an injury while in the care of the mother, but whether the father or step-father has a drink problem and has been beating the mother up is extremely relevant to the long term welfare of the child.

The purpose of splitting the hearing is not to split the two questions which the court must answer. It is to separate out those factual issues which are capable of swift resolution so that the welfare professionals have a firm foundation of fact upon which to base their assessments of family relationships and parenting ability: see In Re S (Care Proceedings: Split Hearing) [1996] 2 FLR 773. A fact finding hearing is merely one of the case management possibilities contemplated by the new Public Law Outline. It is not a necessary pre-condition for the core professional assessment, which the Public Law Outline now expects should normally be done before the proceedings even begin (Judiciary of England and Wales and Ministry of Justice, The Public Law Outline, Guide to Case Management in Public Law Proceedings, April 2008, President's Practice Direction, para 9.2, pre-proceedings checklist and Flowchart). There is no point in splitting the issues if the facts cannot be determined relatively quickly, still less if it is unlikely to result in clear cut findings to help the professionals in their work.

But the finding of those facts is merely part of the whole process of trying the case. It is not a separate exercise. And once it is done the case is part heard. The trial should not resume before a different judge, any more than any other part heard case should do so. In the particular context of care proceedings, where the character and personalities of the parties are important components in any decision, it makes no sense at all for one judge to spend days listening to them give evidence on one issue and for another judge to send more days listening to them give evidence on another. This is not only a wasteful duplication of effort. Much useful information is likely to fall between the gaps. How can a judge who has not heard the parents give their evidence about how the child's injuries occurred begin to assess the risk of letting them care for the child again? The experts may make their assessments, but in the end it is for the judge to make the decision on all the evidence before him. How can he properly do that when he has heard only half of it?

Human rights

Children have both the right to life under article 2, and the right not to be subjected to torture or inhuman or degrading treatment or punishment under article 3, of the European Convention on Human Rights. States are required to take measures to protect them, principally in form of effective deterrence through the criminal law (A v United Kingdom (1999) 27 EHRR 611 ) but also through taking steps to remove them from an abusive situation about which the authorities knew or ought to have known (Z v United Kingdom (2002) 34 EHRR 3). But there is nothing in those cases to suggest that the authorities are failing in their duty to protect children from inhuman or degrading treatment because they are unable permanently to remove children from their families on the basis of unproven allegations. Deterrence through the criminal law depends upon proof to the criminal standard. The duty to take positive protective steps depends upon the authorities having reasonable notice of the risk (see Osman v United Kingdom (1998) 29 EHRR 245). In Osman the police knew of the threats to kill; in Z the local authority were well aware of the long history of neglect and abuse. For a risk to be "real" it has to be founded on real facts not unproven speculations.

Children also have the right to respect for their family lives under article 8 of the Convention. This is, of course, a qualified right. Interference by the authorities is justified if it is "necessary in a democratic society" in order to protect the child's own rights, which in this context include the right to be protected from harm. But there has to be a "pressing social need" for the interference, the reasons for it have to be "relevant and sufficient", and the interference itself has to be proportionate to the need: see, for example, K & T v Finland (2001) 31 EHRR 18, Scozzari & Giunta v Italy (2002) 35 EHRR 12, Kutzner v Germany (2002) 35 EHRR 25. It is difficult to see how the reasons for taking a child away from her family for the indefinite future can be "relevant and sufficient" if they rely upon unproven allegations as the only basis for inferring that the child is at risk of harm.

For my part, therefore, I see no reason to revise the existing law in the light of the Human Rights Act 1998.

Conclusion

I would therefore dismiss this appeal. However, two consequential points arise. Part of the judge's order in preparation for the next part of the hearing was a draft letter of instruction to the experts. This was designed to point up the problems with the present law as the judge saw them. The parties are all content with the much simpler draft proposed by Cafcass for cases such as this:

"The court has considered the allegations of [type of harm] made against [name(s) of alleged abuser(s)] and has concluded that the court is not satisfied that they are more likely than not to be true.

In those circumstances the fact that those allegations were made remains part of the factual matrix of the family history and the ramifications of their having been made may well be relevant to your assessment. However, given that the court was not satisfied that the allegations were true, they cannot form the basis for asserting that there is a current risk of the same type of harm occurring in the future."

This would certainly seem a more appropriate form of words than that chosen by the judge for use in the present case and may well be suitable for use generally.

The second point is that, although not invited to do so by any of the parties, the judge recused himself from the case. His concern was that he had deliberately instigated this test to the existing law and others might perceive that he would find it difficult put out of his mind his view that the "real possibility" ought to be taken into account at the welfare stage. However, all judges are from time to time required to apply legal principles with which they have intellectual difficulty. The problem which the judge saw in this case will arise in any other care case in which allegations are made but not found on the balance of probabilities to be true. If the judge is not fitted to try this case, it might be said that he is not fit to try any case in which the same problem could arise, and that would be absurd. For all the reasons given earlier, the same judge should hear the whole case. Indeed, this case is a good illustration, for any subsequent judge might well have difficulty in extracting the really important findings from such a long and complicated judgment on the factual issues.

I would therefore send the case back for the experts to be instructed and the judge to complete his hearing of the case in the light of the judgments in this House. As with so many family cases, it is likely that things have moved on since these proceedings were begun. The problems which loomed so large in the past may have receded while others have reared their heads. In family life, as in family proceedings, nothing stands completely still.