England and Wales Court of Appeal (Civil Division) Decisions

You are here: BAILII >> Databases >> England and Wales Court of Appeal (Civil Division) Decisions >> J (Children), Re [2012] EWCA Civ 380 (03 April 2012)
URL: http://www.bailii.org/ew/cases/EWCA/Civ/2012/380.html
Cite as: [2012] EWCA Civ 380, [2012] 3 WLR 952, [2012] 2 FCR 1, [2012] 2 FLR 842, [2012] Fam Law 790, [2012] WLR(D) 111

[New search] [Printable RTF version] [Buy ICLR report: [2013] 1 AC 680] [Buy ICLR report: [2012] 3 WLR 952] [View ICLR summary: [2012] WLR(D) 111] [Help]

		Neutral Citation Number: [2012] EWCA Civ 380
		Case No: B4/2011/3230

IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM MIDDLESBROUGH DISTRICT REGISTRY
Her Honour Judge Hallam
TE11C00099

Royal Courts of Justice
Strand, London, WC2A 2LL

03/04/2012

B e f o r e :

THE LORD CHIEF JUSTICE OF ENGLAND AND WALES
THE MASTER OF THE ROLLS
and
LORD JUSTICE MCFARLANE
____________________

Re: J (Children)

____________________

Mr S Cobb QC and Mr J Gray (instructed by Stockton on Tees Borough Council) for the Appellant
Mr P Storey QC and Mr M Todd (instructed by William Bache and Co) for the First Respondent
Ms Judith Rowe QC and Mr B Boucher-Giles (instructed by Leigh Turton Dixon Solicitors) for the Second Respondent
Hearing date : 1st February 2012
____________________

HTML VERSION OF JUDGMENT
____________________

Lord Justice McFarlane :

On the morning of 29th March 2004 a three week old baby, T-L, was found to be dead in the bed of her parents in hostel accommodation where they were staying in South Wales. Post-mortem examination showed that during her short life this baby had sustained some seventeen fractures to her ribs a week or more prior to her death, together with bruising to her face, a shoulder and arm. She was also suffering from serious, untreated nappy rash. The cause of death was asphyxia by obstruction of her airways, the precise mechanism being unascertained.

JJ was T-L's mother and SW her father. T-L was their first child. Following the birth of their second child, S, on 13 August 2005, care proceedings were immediately instigated and, in due course, a fact finding hearing was conducted by HHJ Masterman focussing on the injuries and death of T-L when in her parents' care.

On 24 May 2006 HHJ Masterman gave judgment in the course of which he made the following findings:

a) The week old multiple rib fractures were caused as a result of someone gripping and squeezing the baby's rib cage with a force which would have been way beyond any normal or even rough handling and were the result of a deliberate act of physical abuse;

b) As a result of the fractures, T-L would have been in very obvious pain with every breath, cough, sneeze or passing of a motion. The nappy rash would have been inflamed by the passing of urine;

c) The four rib fractures that were sustained around the time of death probably preceded death and were unlikely to be a consequence of attempts at resuscitation;

d) The bruises to the face and upper body were non-accidental injuries;

e) Such explanations as were offered by the parents were not accepted;

f) The parents had deliberately kept T-L away from appointments with health professionals so that the bruise on her jaw would not attract unwelcome attention;

g) T-L's death was as a result of asphyxia by obstruction of her airways, however the judge was not able to determine whether this arose by one parent accidentally overlaying during the night on the baby whilst she was in the parental bed, or as a result of deliberate suffocation.

HHJ Masterman concluded his judgment in these terms:

"If one parent is innocent of actually abusing T-L, then he or she knows it was the other. If he or she did not know or understand fully at the time, he or she has come to know the truth. Despite this knowledge each maintains their innocence and protests the innocence of the other, only conceding the other's guilt when forced by logic in cross examination to do so. The couple remain together as a couple and still plan to marry. To put this starkly, one parent is prepared to marry a person they know to have inflicted deliberate and serious harm on their child and is thereafter, presumably, prepared to allow that person to be in charge of S in the future. There is no doubt in my mind that SW and JJ are colluding with one another to hide the truth from the rest of the world. One is protecting the other or they are both protecting each other. They are putting the continuance of their relationship before the welfare of S.

I express it in this way because there is no definite evidence to link one or other parent to these injuries…T-L's injuries could have been inflicted by either, or both, of them. Singling out a likely perpetrator does not help this couple because it must be debatable as to which is worse, to inflict this injury or to protect the person responsible. On these findings it is very difficult to see how either parent, let alone both together, could be safely entrusted in the future with the care of S. Sadly, unless and until they are prepared to tell the truth there seems little prospect of this situation changing."

In consequence of Judge Masterman's findings, in July 2006, young S was made the subject of a care order and an order authorising the Local Authority to place him for adoption.

About a year later JJ and SW separated and JJ moved to the North of England. In March 2008 she met and commenced a relationship with a new partner, DJ. DJ had recently separated from the mother of his two children, H, born 20th June 2005, and therefore now aged six and a half, and T, born 17^th August 2006 and therefore five and a half years old. H and T's mother had left the family home with the result that DJ had sole care of the children. As their relationship developed, JJ moved in to that family home with those two children in the early part of 2008.

In the Autumn of 2008 JJ once again became pregnant and in due course her third child, I, was born on 19^th July 2009. It was originally assumed that the father of I was her new partner, DJ. However DNA testing has now confirmed that I's father is in fact JJ's estranged former partner, SW. It is apparently the case that JJ and SW rekindled their relationship during a period measured in months at the end of 2008 when JJ had returned to South Wales. Be that as it may, by the time of I's birth JJ was back living with DJ and his two children and I has been brought up as part of that family since her birth.

To conclude the relevant dates in the recent family history, before turning to the progress of these proceedings, JJ and DJ married in June 2011 and their first child, R, was born on 1^st December 2011.

Whilst the circumstances surrounding T-L's short life and death were well known to the Social Services in South Wales, their counterparts in the North East of England, where JJ had formed her new family with DJ, were unaware of these matters until December 2010. Once they had considered material disclosed to them from South Wales, the local Social Services instigated a child protection plan under which they required JJ to vacate the family home. This plan was put into effect on 3^rd March 2011 and that date has been accepted as the "relevant date" for the purposes of determining the threshold criteria in Children Act l989 s 31, with respect to the care proceedings that were issued in relation to H, T and I on 30^th April 2011.

It follows from the chronology that JJ had been living in a family with H and T for some three years prior to the "relevant date" and her new baby, I, had been a full part of that family for 21 months prior to the "relevant date".

It is apparent that the only facts leading to the Local Authority issuing an application under s 31 with respect to these three children arose out of the findings of fact made in 2006 by HHJ Masterman in Wales.

A preliminary issue was raised as to whether those findings, which related to a family unit comprised of JJ and SW, could support a determination under CA 1989, s 31 that the threshold criteria were satisfied in relation to a different family unit in which SW played no part and the parental figures were JJ and DJ.

The preliminary issue was heard by Her Honour Judge Hallam who concluded, in a judgment dated 25^th November 2011, that the established facts regarding physical injury could not support a threshold finding in relation to the new family, with the consequence that the local authority's s 31 application failed and the care proceedings were dismissed. The protective measures therefore ceased and JJ returned home to live with her husband, DJ, and the three (now four) children.

In the present appeal the Local Authority seeks to overturn the decision of HHJ Hallam.

Before considering the circumstances of this case and the decision of HHJ Hallam further, it may be helpful to set the scene by describing the issue of principle which arises from the facts of this case.

The point in this appeal: past "possible perpetrators" in a new family

Put shortly the point raised by this appeal is as follows:

a) Where a previous court has found that there is a 'real possibility' that one or other or both of two or more carers have perpetrated significant harm on a child in his/her care;

b) Is that 'finding' a 'finding of fact' that may be relied upon in subsequent proceedings relating only to one of the potential perpetrators in support of a conclusion that there is a 'real possibility' or likelihood of a subsequent child in a new family unit of which he/she is part suffering significant harm or is it a 'finding' that must be totally ignored in the subsequent proceedings?

There have been authoritative statements in judgments of the Court of Appeal and the Supreme Court which assert that the previous adverse 'finding' must be totally ignored in any subsequent proceedings involving a new family unit, on the basis that a 'finding' that there is a 'real possibility' is not a 'finding of fact' at all as, by definition, it falls short of a finding on the balance of probabilities. Despite these apparently clear authoritative statements, the appellant argues that such statements are, or may well be, at odds with the fully reasoned analysis of the approach to be taken to the statutory threshold criteria in Children Act 1989, s 31 as set out by Lord Nicholls of Birkenhead in a trilogy of cases in the House of Lords.

The approach to 'possible perpetrators':

Where a court is in the position faced by Judge Masterman in the present case of finding that significant harm has been occasioned to a child, but being unable to identify on the balance of probabilities which of a number of individuals perpetrated the harm, the most the court can do is to identify a pool of possible perpetrators. As will be seen, the case law establishes that an individual will be considered as a "possible perpetrator" where the evidence establishes that there is "a real possibility" that that is indeed the case. As the concept of the "pool of possible perpetrators" only arises where the evidence is insufficient to identify one or other possible perpetrator as being "the" perpetrator on the balance of probabilities, a name goes into the pool of possible perpetrators only where the evidence falls short of the balance of probabilities but is sufficient to establish "a real possibility" that a particular individual caused a particular injury.

The threshold criteria in CA 1989, s 31 are as follows:

"31 Care and Supervision

(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 a case such as the present, where there is no evidence that any of the three subject children are suffering significant harm, the local authority must rely upon the second limb of the threshold criteria, namely that the child concerned is "likely to suffer significant harm". Again as will be seen shortly, the case law establishes that the future prediction that a particular child is likely to suffer significant harm, must be based upon the establishment of past facts, each of which has been proved to the requisite standard of proof, namely the balance of probabilities. The question raised in these proceedings is, where a finding that one parent in the new family in the past has been the subject of a finding that there is a real possibility that they perpetrated significant harm on another child, whether that finding is sufficient to support a determination that a child or children in their present care is "likely to suffer significant harm".

The case law

The key authorities charting the path from the passing of CA 1989, s 31 to the accepted interpretation of its main provisions by the House of Lords and Supreme Court are well known; reference to them in this judgment is necessary to set the context within which the discrete issue in the present case falls to be considered.

As indicated at paragraph 16 above, the central question to be considered when looking at these authorities is whether or not the trilogy of decisions in which Lord Nicholls gave the leading judgment (Re H (Minors) (Sexual Abuse: Standard of Proof) [1996] AC 563; Lancashire County Council v B [2000] 2 AC 147; and Re O and N (Minors) (Care Preliminary Hearing); Re B (A Minor) [2003] UKHL 18; [2004] 1 AC 523) in fact supports the shortly stated but nonetheless authoritative statements made in the three cases which compelled the learned judge in the present case to hold that she had no option but to ignore the earlier adverse finding. Those three cases are: Re B and W (Threshold Criteria) [1999] 2 FLR 833 (that is the Lancashire County Council case in the Court of Appeal); Re S-B (Children) (Care Proceedings: Standard of Proof) [2009] UKSC 17; [2010] 1 AC 678; and Re F (Interim Care Order) [2011] EWCA Civ 258. A subsidiary, but not unimportant question, is whether the jurisprudence described by Lord Nicholls draws a distinction between those cases where there is no finding at all, on the balance of probabilities, that a child has suffered any past harm, and those cases where past harm has plainly occurred, but it is not possible to identify an individual perpetrator.

In 1995 the House of Lords in Re H (Minors) (Sexual Abuse: Standard of Proof) [1996] AC 563 established that the phrase 'likely to suffer' significant harm in CA 1989, s 31(2)(a) was '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'. In the leading speech Lord Nicholls of Birkenhead stressed that, whilst the task of assessing likelihood involved looking to the future, the process must be based on established fact:

"Whether at [the relevant date] the child was suffering significant harm is an issue to be decided by the court on the basis of the facts admitted or proved before it. The balance of probability standard applies to proof of the facts.

The same approach applies to the second limb of s 31(2)(a). This is concerned with evaluating the risk of something happening in the future; aye or no, is there a real possibility that the child will suffer significant harm? Having heard and considered the evidence, and decided any disputed questions of relevant fact upon the balance of probability, the court must reach a decision on how highly it evaluates the risk of significant harm befalling the child, always remembering upon whom the burden of proof rests."

Later, under a heading 'a conclusion based on facts, Lord Nicholls said:

"But the rejection of a disputed allegation as not proved on the balance of probability leaves scope for the possibility that the non-proven allegation may be true after all. There remains room for the judges to have doubts and suspicions on this score. This is the area of controversy.

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

Having set out his reasons and observed that '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', Lord Nicholls concluded:

"Here [in CA 1989, s 31(2)(a)], as much as anywhere else, the court's conclusion must be founded on a factual basis. … 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."

In Re H the primary allegation relied upon by the local authority was that the mother's partner had sexually abused one of the children living in the family. The allegation was not proved to the requisite standard and, their Lordships held, it was not therefore open to the judge to rely on that unproved allegation to support a finding on the second limb of the threshold.

Re H was followed and developed in the Court of Appeal in Re M and R (Child Abuse: Evidence) [1996] 2 FLR 195. The case concerned the approach to be adopted in care proceedings where the s 31 threshold criteria are met on one, comparatively minor, basis but the court has found that serious allegations (in that case sexual abuse) were not proved: could the court nonetheless have regard to those unproven allegations at the welfare stage of the case when having regard, as required by CA 1989, s 1(3)(e), to 'any harm which [the child] has suffered or is at risk of suffering'? In line with Re H, the Court of Appeal firmly concluded that allegations which had failed to be proved could not form the basis of establishing a risk of future harm.

The approach to be taken where there are a number of potential perpetrators, none of whom can be positively found on the balance of probability to be the actual perpetrator was considered by the House of Lords in Lancashire County Council v B [2000] 2 AC 147. The case arose from the established fact that a seven month old child had been violently shaken on at least two occasions. The child had spent a deal of time in the care of a childminder, as well as in the care of her parents. The local authority not only applied for a care order with respect to the injured child ('A') but also for a similar order with respect to the child minder's child ('B'). The judge found that A's injuries had been caused by the mother and/or the father or by the childminder, but was unable to go further and identify any one of these three as a perpetrator on the balance of probabilities. As a result of his inability positively to identify any perpetrator to the civil standard, the judge concluded that both sets of care proceedings should be dismissed. The local authority appealed. The Court of Appeal allowed the appeal with respect to the parents' child, A, but dismissed the appeal with respect to the childminder's child, B. A's parents appealed to the House of Lords. There was no appeal by the local authority with respect to B, consequently the childminder's case was not before the House of Lords.

The case focussed upon the 'attributable' element within s 31(2)(b)(i), namely that 'the harm, or likelihood of harm, is attributable to 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'.

Their Lordships dismissed the parents' appeal on the basis that, whilst normally the 'care given to the child' would refer to care by his parents or primary carers, where a child, who was cared for by a wider variety of carers, has been harmed, the threshold criteria will be satisfied even if the court is unable to identify which of the carers had provided deficient care.

Lord Nicholls, again giving the leading speech, said:

"In the present case the child is proved to have sustained significant harm at the hands of one of both of her parents or at the hands of a daytime carer. But, according to this argument, if the court is unable to identify which of the child's carers was responsible for inflicting the injuries, the child remains outside the threshold prescribed by Parliament as the threshold which must be crossed before the court can proceed to consider whether it is in the best interests of the child to make a care order or supervision order. The child must, for the time being, remain unprotected, since CA 1989, s 31 and its associated emergency and interim provisions now provide the only court mechanism available to a local authority to protect a child from risk of further harm.'

I cannot believe Parliament intended that the attributable condition in s 31(2)(b) should operate in this way. Such an interpretation 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."

In relation to the present appeal, it is important to note that Lord Nicholls expressly endorsed an approach which necessarily contemplated that the threshold would be crossed, with the risk of a care order separating parent and child, where there had been no positive finding against that parent, who may indeed be 'wholly innocent'. Such an outcome was to be contemplated and was justified to avoid 'the prospect that an unidentified, and unidentifiable, carer may inflict further injury on a child he or she has already severely damaged'. For clarity I will set out the relevant extract of Lord Nicholls words:

"I recognise that the effect of this construction is that the attributable condition may be satisfied when there is no more than a possibility that the parents were responsible for inflicting the injuries which the child has undoubtedly suffered. That is a consequence which flows from giving the phrase, in the limited circumstances mentioned above, the wider meaning those circumstances require. I appreciate also that in such circumstances, when the court proceeds to the next stage and considers whether to exercise its discretionary power to make a care order or supervision order, the judge may be faced with a particularly difficult problem. The judge will not know which individual was responsible for inflicting the injuries. The child may suffer harm if left in a situation of risk with his parents. The child may also suffer harm if removed from parental care where, if the truth were known, the parents present no risk. Above all, I recognise that this interpretation of the attributable condition means that parents who may be wholly innocent, and whose care may not have fallen below that of a reasonable parent, will face the possibility of losing their child, with all the pain and distress this involves. That is a possibility, once the threshold conditions are satisfied, although by no means a certainty. It by no means follows that because the threshold conditions are satisfied the court will go on to make a care order. And it goes without saying that when considering how to exercise their discretionary powers in this type of case judges will keep firmly in mind that the parents have not been shown to be responsible for the child's injuries.

I recognise all these difficulties. This is indeed a most unfortunate situation for everyone involved: the child, the parents, the child-minder, the local authority and the court. But, so far as the threshold conditions are concerned, the factor which seems to me to outweigh all others is the prospect that an unidentified, and unidentifiable, carer may inflict further injury on a child he or she has already severely damaged."

In Lancashire County Council v B, in contrast to Re H, the past harm had been established. The parents, both of whom (with the child minder) were in the pool of possible perpetrators were still together. The Court of Appeal held that the threshold criteria were satisfied in relation to the parents' child, A, on the first limb as that child had suffered harm. In the case of the child minder's child, B, the focus was the second limb looking at likely future harm. As it had not been proved, on the balance of probabilities, that the child minder had caused the injury to child A, it was not possible to find that her own child was likely to suffer harm in her care (Court of Appeal decision: Re B and W (Threshold Criteria) [1999] 2 FLR 833). There was no appeal by the local authority in respect of the Court of Appeal decision relating to child B; accordingly, as Lord Nicholls observed, the correctness of that decision was not a matter before their lordships.

In the context of the present case, it is to be noted that the ratio in the Court of Appeal determination with respect to child B is, as will be seen, the only clear ratio in a previously decided case (prior to Re F (Interim Care Order) [2011] EWCA Civ 258) in the higher courts on the point that arises in the present appeal. I will therefore set out the entirety of the Court of Appeal judgment on the point found in the judgment of the court given by Robert Walker LJ:

"In our judgment the judge was clearly right in the conclusion which he reached in relation to child B (although we too reach that conclusion with no enthusiasm). There is no allegation or evidence that B has been harmed in any way. In relation to B the first relevant threshold condition is risk of future harm, which can be established only on the basis of proven facts, not just suspicion. It has not been proved to the requisite standard of proof that [the child minder] was the perpetrator of A's injuries. Any notion that [A's mother, father and the child minder] should for the future be regarded as a group can be dismissed without any need for close examination, because it is quite clear from what counsel told us that (whatever the future holds for A and B) [the child minder] will not in future participate in any way in A's care, and [A's parents] will not participate in any way in B's care." [emphasis added]

In 2003 two separate cases heard together by the House of Lords (Re O and N (Minors) (Care Preliminary Hearing); Re B (A Minor) [2003] UKHL 18; [2004] 1 AC 523) focussed on the approach a court should take at the 'welfare' stage of a child care case where the s 31 threshold criteria have been satisfied, but where the court has been unable to identify which, out of a number of potential perpetrators, was responsible for past injury. Their lordships held that in such a case it was appropriate for the judge when determining whether or not to make a care order to proceed on the footing that each carer was indeed a potential perpetrator of past injury having regard to whatever extent is appropriate to the facts that had previously been found.

Once again it fell to Lord Nicholls to deliver the leading speech. During the course of describing the conclusion that I have summarised in the previous paragraph, he said [at paragraphs 26 to 28]:

"26 The first area concerns cases of the type involved in the present appeals, where the judge finds a child has suffered significant physical harm at the hands of his parents but is unable to say which. I stress one feature of this type of case. These are cases where it has been proved, to the requisite standard of proof, that the child is suffering significant harm or is likely to do so.

27 Here, as a matter of legal policy, the position seems to me straightforward. 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.

28 That would be a self-defeating interpretation of the legislation. It would mean that, in "uncertain perpetrator" cases, the court decides that the threshold criteria are satisfied but then lacks the ability to proceed in a sensible way in the best interests of the child. 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. As Hale LJ said in Re G (Care Proceedings: Split Trials) [2001] 1 FLR 872 , 882, para 44: "the fact that a judge cannot always decide means that when one gets to the later hearing, the later hearing has to proceed on the basis that each is a possible perpetrator ..." This approach accords with the basic principle that in considering the requirements of the child's welfare the court will have regard to all the circumstances of the case."

It is of particular note that the section of Lord Nicholls judgment from which these quoted passages are extracted is headed 'The Welfare Stage: "uncertain perpetrator" cases'. The judgment plainly draws a distinction between such cases and those covered by the succeeding heading 'The Welfare Stage: unproved allegations of harm'. In this latter section (paragraphs 37 to 41), Lord Nicholls is dealing with cases where the threshold is crossed on the basis that some adverse facts are proved, but others, for example physical harm, are not: can the court at the welfare stage rely upon the unproven physical harm allegations? Lord Nicholls response, based upon Re M and R, is that that is not permissible. The important point to be taken from the contrast drawn by Lord Nicholls in these two sections of his judgment is that he plainly does not regard the 'uncertain perpetrator' cases as being ones of unproved allegations of harm.

The distinction in Re O and N between 'uncertain perpetrator' cases and those where no allegations had been proved, expressly endorsed the approach taken by Hale LJ in Re G (Care Proceedings: Split Trials) [2001] 1 FLR 872, whereby (notwithstanding Re M and R) the Court of Appeal held that at the welfare stage in an uncertain perpetrator case the single parent, who had separated from her co-uncertain perpetrator, had to be assessed on the basis that she posed a risk to the child.

Also in 2003 the Court of Appeal (in North Yorkshire County Council v SA [2003] EWCA Civ 839; [2003] 2 FLR 849) considered the test to be applied when determining whether or not a person was a 'possible perpetrator' and held that where there is insufficient evidence positively to establish the identity of the perpetrator of injuries using the balance of probabilities, the test to be applied is 'is there a likelihood or real possibility that A or B or C was the perpetrator or a perpetrator of the inflicted injuries'.

In 2008, in Re B (Children) (Care Proceedings: Standard of Proof) [2008] UKHL 35; [2009] 1 AC 11, the House of Lords clarified the law relating to the standard of proof in children cases by stating very plainly that the standard relevant to facts under-pinning the s 31 threshold criteria or welfare considerations under CA 1989, s 1 was neither more nor less that the simple balance of probabilities.

It is to be noted that none of the cases considered by the House of Lords directly focussed on the second limb of the threshold in cases where an uncertain perpetrator is a member of a different household to that in which past injuries took place. Prior to 2009, the only authoritative determination in that regard was that of the Court of Appeal in the Lancashire case (see paragraph 33 above).

In 2009 the Supreme Court allowed an appeal in the case of Re S-B (Children) (Care Proceedings: Standard of Proof) [2009] UKSC 17; [2010] 1 AC 678. A panel of seven Supreme Court Justices heard the appeal and the judgment of Baroness Hale of Richmond SCJ is the judgment of the court. The judgment in Re S-B and the subsequent Court of Appeal decision in Re F (Interim Care Order) [2011] EWCA Civ 258 are at the core of the present appeal and I will therefore describe them in some detail.

Re S-B concerned two children, 'Jason' and 'William'. At the age of four weeks Jason was found to have bruising to his arms and face. He was removed to foster care and there was subsequently a fact-finding hearing to determine the causation of his injuries and, if possible, the perpetrator of them. During the course of the proceedings regarding Jason, William was born. By that time the parents had separated and the father had withdrawn from any involvement with the family, the professionals and the proceedings. William was removed from his mother's care shortly after birth and placed with his brother in foster care. William had not suffered any injury and the case in relation to him with respect to the s 31 threshold criteria was based on the second limb relating to future harm.

The bruises to Jason were not old and were not of different ages. They would not have been caused by both parents and the non-abusing parent would not have been likely to have been aware of them. It was likely to have been a one-off assault by one parent and was not therefore a case of both parents being to some extent blameworthy whatever their precise role was, rather, as Baroness Hale colloquially put it, it was a pure 'whodunit'.

The trial judge found that there was a high index of suspicion in relation to the father as the perpetrator of the injuries, but concluded that the mother could not be ruled out and, at the request of counsel, indicated that she thought that there was a 60% likelihood that the father had caused the injuries. She subsequently made care orders and placement for adoption orders in relation to both children.

Baroness Hale identified the case of Re S-B as being:

"… about the proper approach to deciding who has been responsible for harming a child in proceedings taken to protect that child, and others in the family, from harm. It raises profound issues: on the one hand, children need to be protected from harm; but on the other hand, both they and their families need to be protected from the injustice and potential damage to their whole futures done by removing children from a parent who is not, in fact, responsible for causing any harm at all." (Paragraph 2).

The trial judge in Re SB approached the standard of proof in these terms: 'When I apply the appropriate standard of proof, it has to be based on evidence of reliability and cogency equivalent to the gravity of the allegations' (which she considered to be 'very serious indeed'). The judge, in maintaining the mother in the pool of perpetrators, had found the s 31 threshold criteria satisfied in relation to both children on the basis that, to use Baroness Hale's words (paragraph 31), 'the fact that there was a real possibility that [the mother] had caused the injuries to Jason meant that there was a real possibility that she would injure William'.

The conclusion of the Supreme Court in Re S-B (paragraph 48) was that the judge had misdirected herself as to the standard of proof at the fact-finding stage. Their lordships confirmed that the simple balance of probability test, following the House of Lords decision in Re B (above), should be applied in finding that a person was the perpetrator of an injury and confirmed the approach, where the evidence falls short of that standard, in North Yorkshire County Council v SA (above) to the effect that an individual will be found to be a possible perpetrator if the evidence establishes 'a real possibility' that they caused the injury. The case was therefore remitted for a complete rehearing before a different judge applying the correct standard of proof.

Baroness Hale's judgment in Re S-B only touches upon the issue in the present appeal in one paragraph, paragraph 49, which is at the very end of the judgment:

"49 There is a further reason to remit the case. The judge found the threshold crossed in relation to William on the basis that there was a real possibility that the mother had injured Jason. That, as already explained, is not a permissible approach to a finding of likelihood of future harm. It was established in In re H [1996] AC 563 and confirmed in In re O [2004] 1 AC 523 , that a prediction of future harm has to be based upon findings of actual fact made on the balance of probabilities. It is only once those facts have been found that the degree of likelihood of future events becomes the "real possibility" test adopted in In re H. It might have been open to the judge to find the threshold crossed in relation to William on a different basis, but she did not do so."

That paragraph has been the focus of much debate during the hearing of this appeal and I will return to consider its import in more detail in due course.

The second recent authority which has been at the forefront of our consideration is Re F (Interim Care Order) [2011] EWCA Civ 258. Re F involved a father in respect of whom there had been a previous finding of fact relating to his eldest child who had suffered two leg fractures. In care proceedings relating to that child the finding as to the perpetrator of the fractures was that both of that child's parents were in the pool of possible perpetrators. Time then moved on, the father separated from that child's mother and he became a parent for a second time when a baby was born to his new partner. The local authority issued care proceedings with respect to the new baby on the basis that the previous finding of the father being a possible perpetrator was sufficient to cross the threshold criteria on the basis that the new baby was likely to suffer significant harm. The trial judge dismissed the care proceedings on the basis that the previous finding was insufficient to support a 'likelihood' threshold finding in relation to the new baby. The trial judge did, however, grant the local authority permission to appeal to the Court of Appeal.

Prior to the commencement of the appeal hearing the local authority accepted that the approach to the likelihood threshold had been restated by the Supreme Court at paragraph 49 of Re S-B and that, as the Court of Appeal was bound by that decision, the appeal was doomed to failure. The local authority did, however, seek permission to take the matter to the Supreme Court. The Court of Appeal dismissed the appeal and refused permission to appeal to the Supreme Court.

The leading judgment in Re F was given by Wilson LJ, as he then was. The appeal had been conceded at the outset and was dismissed by Wilson LJ in his second paragraph. The remaining 17 paragraphs of the ex tempore judgment relate to the decision to refuse permission to appeal to the Supreme Court. I regard the following as the central paragraphs of Wilson LJ's judgment:

"12. There is no doubt that the reference in section 31(2)(a) to the child who is likely to suffer significant harm is a reference to a child in respect of whom there is at least a real possibility that he will suffer significant harm. But it has been established in law for 15 years that, to quote Lord Nicholls in In Re H (Minors) (Sexual Abuse: Standard of Proof) [1996] AC 563 at 590C,"there must be facts from which the court can properly conclude that there is a real possibility that the child will suffer harm in the future". That proposition has been carried seamlessly through more recent decisions of the House of Lords and of the Supreme Court. It was, for example, firmly restated by Baroness Hale in In Re B (Children) (Care Proceedings: Standard of Proof) [2008] UKHL 35, [2009] 1 AC 11, at [22] and [23]' [emphasis as in original].

…

Mr Heaton wishes to argue to the Supreme Court that, where in relation to one child there has been a finding of non-accidental injury and the only uncertainty relates to the identity of its perpetrator, there should be a relaxation of the principle which requires the real possibility of future harm to a second child to be founded only on a further proven fact in relation to the identity of the perpetrator. But that the principle operates with full force even in that area has been demonstrated first by the decision of this court in Lancashire CC v B [2000] 2 AC 147, at 155F, being a conclusion which was not disturbed on the further appeal to the House of Lords. Moreover the principle was in particular applied to our very situation by the Supreme Court as recently as in In Re S-B (Children) (Care Proceedings: Standard of Proof) [2009] UKSC 17, [2010] 1 AC 678, at [49]. It is strictly true that the reasoning in that paragraph is obiter: for the court directed a fresh fact-finding enquiry on the basis that the trial judge had applied too high a standard of proof. But the court there proceeded as follows:

"There is a further reason to remit the case. The judge found the threshold crossed in relation to William on the basis that there was a real possibility that the mother had injured Jason. That, as already explained, is not a permissible approach to a finding of likelihood of future harm ... a prediction of future harm has to be based on findings of actual fact made on the balance of probabilities. It is only once those facts have been found that the degree of likelihood of future events becomes the 'real possibility' test..."

In my view the strict status of that passage as obiter carries very little significance in circumstances in which it is all of a piece with a number of earlier, yet also recent, decisions of the House of Lords.

Mr Heaton contends, and Mr Hayden concedes, that the reasoning so clearly set out in In Re S-B, at [49], has caused great consternation among local authorities, among other professionals who work in the area of public law and among academic commentators. Mr Heaton's aspiration is to persuade the Supreme Court to modify its demand for proven factual foundation in uncertain perpetrator cases. Take, says he to us this morning, a case of two parents who are consigned to a pool of possible perpetrators of non-accidental injuries to their child; and who then separate; and who each, with other partners, produce a further child, who together become the subject of conjoined care proceedings. Are both those applications for care orders required to be dismissed even though before the court is, on any view, a perpetrator of injuries to that older child? No doubt there are hard and worrying cases. But the requirement of proven factual foundation is a bulwark against the state's removal of a child from his family, which I consider very precious. I also applaud the Supreme Court's regular acknowledgement of the fact that, although it can depart from its previous decisions, the exercise of departure is highly unsettling for the law and should be undertaken only with great caution.

For those reasons I would not wish us to foist upon the Supreme Court a full appeal in circumstances in which it had not itself had the opportunity to consider whether to accept it."

In addition, by the time the matter was before the Court of Appeal in Re F, it was asserted that the s 31 threshold may well be satisfied following the father and his new partner admitting that they had cared for the younger child whilst under the influence of drugs. On that basis, Wilson LJ considered that the case might be an inapt vehicle for the contentions that the local authority wished to place before the Supreme Court.

Rimer and Pill LJJ both agreed that the appeal should be dismissed and that permission to appeal to the Supreme Court should be declined.

In the present appeal, it has been submitted that Wilson LJ may have erroneously stated two matters in the course of his judgment in Re F namely:

a) That the Court of Appeal's decision regarding the child minder's child in the Lancashire case was 'not disturbed' on the further appeal to the House of Lords, whereas that child's case had not in fact been before the House of Lords; and

b) That Baroness Hale's judgment at paragraph 49 of Re S-B is all of a piece with a number of earlier, yet also recent, decisions of the House of Lords.

The question of whether the submission at (b) above is made out is obviously at the forefront of the present appeal.

The present case

Against the background that I have now described, HHJ Hallam was required to determine whether or not, on the basis of the findings of fact made in 2006 by HHJ Masterman, the threshold criteria in CA 1989, s 31(2) were satisfied with respect to H, T and I.

In a clear and helpfully structured judgment the judge summarised the background material and quoted HHJ Masterman's conclusion to the effect that the pool of perpetrators was comprised only of JJ and SW and that it was not possible on the evidence to single one or other of them out as being the perpetrator of any of the physical injuries suffered by T-L.

At the hearing before HHJ Hallam the Local Authority conceded that, although there may have been other concerns about the family, the only matters that could meet the threshold criteria at the relevant time were the findings as to physical injuries sustained by T-L. The judge expressly recorded that the Local Authority "do not seek to bring failure to protect into the equation". The judge summarised the issue in these terms:

"Thus the issue I have to determine is whether JJ's inclusion in a pool of perpetrators in earlier proceedings involving a different child and a different relationship, can form the basis of the threshold in relation to a subsequent child in later proceedings."

The judge reviewed the relevant case law and made the following conclusions:

a) "Thus I conclude…that in order for the court to properly find that there is a real possibility that the child will suffer harm in the future there must be facts proved to the requisite standard i.e. on the balance of probabilities. [Paragraph 19];

b) I draw the conclusion that [HHJ Masterman] was unable to say on the balance of probabilities that JJ or SW caused the injuries. The only conclusion I can draw from his judgment is that he was satisfied that there was a real possibility that JJ caused the injuries and that is why he included her in the pool of perpetrators. [Paragraph 25];

c) In my judgment the seriousness of the findings made by HHJ Masterman cannot be a factor sufficient to alter the approach of this court as determined by the guidance and decisions of the higher courts. [Paragraph 31];

d) The likelihood of significant harm in s 31(2) can only be proved by reference to past facts which are proved on the balance of probabilities. In this case the only facts available to the Local Authority have not been proved to that standard. The fact that the mother injured T-L has only been proved on the basis of "a real possibility" by virtue of the fact that HHJ Masterman found only that she was within the pool of perpetrators rather than finding that she was "the" perpetrator. [Paragraph 36];

e) The previous findings are the only facts relied upon by the Local Authority. Thus the threshold criteria cannot be met in any other way and I have to conclude that the threshold are not met in this case." [Paragraph 37]

In the concluding paragraph of her judgment HHJ Hallam made the following observation:

"I am aware that the present law does cause consternation for Local Authorities, professionals involved in the protection of children and academic commentators. However it is quite apparent that the higher courts have considered those concerns and taken them into account in reaching their decisions. That is clear from Wilson LJ's judgment in the case of Re F. That decision was made in recent times and I am not able to depart from it."

The appeal

Before this court the Local Authority seeks to appeal the decision of HHJ Hallam. The grounds of appeal may be summarised as follows:

a) Error in concluding that, in the absence of a determination of which of JJ and/or SW caused significant harm to the child T-L, there were no facts upon which the court could conclude that there was a real possibility that the present children would suffer significant harm while in the care of JJ;

b) Failure to have sufficient regard to the nature and gravity of the harm suffered by T-L;

c) Error of law in rejecting the authority of Re O and N [2003] UKHL 18 as being irrelevant to her consideration;

d) Erroneously failing to apply the principle that the likelihood of significant harm to an uninjured sibling is not eliminated merely because that child is in the care of only one of the parents adjudged to be in the pool of perpetrators of significant harm to her or her sibling;

e) Wrongly concluding that she was bound by the decision of the Court of Appeal in Re F (A Child) [2011] EWCA Civ 258;

f) Erred by declining to hear and consider evidence and, in particular, having regard to the potential for JJ to expose the child I to the joint care of herself and SW.

In relation to the latter ground, (f) above, it was accepted before us that the judge was not invited to consider any evidence outside the basic findings made by HHJ Masterman. Consequently that ground was not pursued before this court.

The appeal is opposed by both JJ and DJ. Lawyers acting on behalf of the Children's Guardian have throughout, quite properly, adopted a neutral stance on this issue and have played no part in the appeal proceedings.

In developing his argument in support of the appeal Stephen Cobb QC, leading Justin Gray, focussed in upon paragraph 49 in the judgment of Baroness Hale in Re SB together with the later decision of Wilson LJ in Re F and in particular at paragraph 11 of that judgment. The Local Authority's submission is that those two comparatively short judicial assertions as to the applicable law are unsupported by any earlier determinations of the House of Lords or the Supreme Court and are, indeed, at odds with the House of Lords' decision in Re O and N and the first instance judgment of Wall LJ in Re CB and JB (which was later expressly adopted and endorsed by Lord Nicholls in the course of his judgment in Re O and N).

This court was told that these two judicial pronouncements were causing difficulty in interpretation in courts of first instance and (as demonstrated in this case, and in the case of Re F itself) were a cause of real concern to social services departments and others charged with the protection of children.

In relation to matters of law Mr Cobb attributes the greatest prominence in presenting his argument to the House of Lords' decision in Re O and N. He draws particular attention to paragraph 20 of the speech of Lord Nicholls of Birkenhead which states:

"[20] Section 31 and its associated emergency and interim provisions comprise the only court mechanism available to a Local Authority to protect a child from risk. The interpretation of the "attributable" condition adopted by the House is necessary to avoid the unacceptable consequence that, otherwise, if the court cannot identify which of the child's carers was responsible for inflicting the injuries the child will remain wholly unprotected. As Wall J observed in Re B (minors) (Care proceedings: practice) [1999] 1 WLR 238, that would render the statutory provisions ineffective to deal with a commonplace aspect of child protection. The interpretation adopted by the House avoids this result while, at the same time, encroaching to the minimum extent necessary on the principle underpinning s 31(2)."

Mr Cobb argues that those words amount to an express endorsement by Lord Nicholls, with whom the other members of the House agreed, of Wall J's obiter observations in Re B (otherwise referred to as Re CB and JB). It is therefore helpful at this stage to look at those observations.

In Re CB and JB, Wall J was required to consider two twins, one of whom had been injured in the care of her parents and the other had not. It was not possible to identify, as between the parents, which was the perpetrator. Having found the 'is suffering' threshold satisfied in relation to the injured twin, CB, Wall J went on to reject a submission that, because he could not identify the perpetrator there was no established fact proved on balance against either parent and therefore the likelihood threshold could not be established in relation to JB. Wall J said:

"Furthermore, in my judgment, there must be a likelihood of JB suffering significant harm, if his twin sister has actually suffered significant harm in the form of life-threatening, non-accidental injuries at the hand of one or both of her parents. Lord Nicholls of Birkenhead in Re H (Minors) (Sexual Abuse: Standard of Proof) [1996] AC 563 made it clear that "likely" in the context of section 31(2) does not mean more probable than not: it is 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." CB has suffered life-threatening non-accidental injuries at the hands of one or both of her parents; there must therefore, in my judgment, be a possibility which cannot sensibly be ignored that if JB were left in the care of his parents — or either of them — he too will suffer significant harm. [emphasis added]"

Accordingly, in my judgment, a finding of fact that a child in CB's position has been non-accidentally injured by one or both of her parents whilst she was in their joint care is sufficient to satisfy the threshold criteria under section 31(2) of the Children Act 1989 in relation to both children, notwithstanding the fact that only one has suffered non-accidental injury and that on the available evidence the court cannot be satisfied on the balance of probabilities that it was one parent rather than the other who inflicted those injuries. To hold otherwise would in my judgment not only be illogical, but would render the statutory provisions ineffective to deal with a commonplace aspect of child protection."

In the Lancashire case in the Court of Appeal, Robert Walker LJ makes express reference to Re CB and JB without voicing any apparent disapproval. Instead Robert Walker LJ agreed with the first instance judge in Lancashire in holding that Re CB and JB was distinguishable and did not apply to the child minder's child because the care of the injured child A was not confined to one household.

Re CB and JB would therefore seem to maintain its authority. A position which is enhanced by express approval from Lord Nicholls in the House of Lords in Re O and N (see paragraph 65 above).

The case for Re O and N focussed particularly on the approach of the court at the "welfare stage" in cases where the identity of the perpetrator of earlier injury remaining "uncertain". In that regard I would draw attention to paragraphs 27 and 28 of his speech Lord Nicholls set out at paragraph 36 above.

Miss Judith Rowe QC for DJ, correctly points out that Lord Nicholls' speech at that point relates to the welfare stage and not the earlier determination of threshold, which is of course the point now before us. Despite that potential impediment in directly applying Lord Nicholls' judgment to the present case, Mr Cobb nevertheless forcefully argues that it describes the correct approach to the determination of threshold in subsequent proceedings where one of the two potential perpetrator parents has joined a new family unit. His submission is that, just as the approach to the welfare stage in the original proceedings is on the basis that each of the original two parents is a possible perpetrator, so too should the likelihood of significant harm be judged on the same basis.

Given the prominence of Baroness Hale in developing the law in this respect, Mr Cobb also draws attention to Lord Nicholls' express endorsement of Hale LJ's, as she then was, contribution in Re G.

In Re G (Care proceedings: split trials) [2001] 1 FLR 872, paragraph 44 of Hale LJ's judgment in full reads as follows:

"[44] Some cases are very difficult; and they can be made more difficult by the nature of the task confronting the fact finding court. Even in a case like this, where it is clear that a child has suffered non-accidental injury at the hands of one or both parents, it may be impossible for the judge to decide which one, in the light of the guidance given by the House of Lords in the well known case of Re H and others (minors) (sexual abuse: standard of proof) [1996] AC 563, although for my part I would regard the threshold of incredulity that such things can ever take place as being much lower once it is clear that the thing has indeed taken place. But the fact that a judge cannot always decide means that when one gets to the later hearing, the later hearing has to proceed on the basis that each is a possible perpetrator, even though that has not been proven. That is an ironic result of the decision in Re H, because the main thrust of that case was to decide that children should not be subject to care orders on the basis that something may have happened rather than on the basis that it did happen. Perhaps at some time in the future somebody will consider that. But this makes it all the more important that the judge who heard the evidence about those events also hears the evidence about what should now be done, particularly as that later evidence is more likely to include evidence of character and propensity, which may well not have been put before the fact finding hearing for very good reasons."

Mr Cobb's attractive presentation of his case leads the judicial listener through the words of Lord Nicholls endorsing, as he does, the earlier utterances of Wall J and Hale LJ. The Local Authority's ultimate submission is that the court should adopt the approach of Lord Nicholls in Re O and N to the determination of the "likely to suffer" threshold criteria in a later case, with the result that an uncertain perpetrator from an earlier determination is regarded as equally likely to have been the cause of past significant harm, and that that established "fact" is a sufficient vehicle to satisfy the threshold in the later case.

Attractive though that argument is, Mr Paul Storey QC on behalf of JJ and Miss Rowe QC on behalf of DJ argue that it is untenable. They submit that to compare the stage of proceedings being described by Lord Nicholls in Re O and N with the stage of proceedings being undertaken by HHJ Hallam in the present case is not to compare like with like. In the former the threshold was crossed and there was no strict evidential yardstick to be deployed by the court in exercising its jurisdiction in affording the child's welfare paramount consideration in deciding what, if any, order to make in the concluding "welfare stage" of the proceedings. Judge Hallam, on the other hand, was required to operate within a strict evidential context, only relying on past facts which had been established on the balance of probabilities.

The opposing counsel submit that Mr Cobb's argument falls down because a finding that someone is in the pool of perpetrators is no more than a finding on the basis that there is "a real possibility" that they perpetrated past injuries, a finding which by definition falls short of the balance of probabilities. To use that finding as the basis for holding that there is a further "a real possibility" (that is a likelihood) of that person causing significant harm in the future, is to find the threshold crossed on the basis of two successive findings of "real possibility" and no finding at all of fact on the balance of probabilities in relation to the one, now separated, parent. Mr Storey submits that such an outcome, on the present law, would be impermissible and that parents and children would risk permanent separation by the State in circumstances where absolutely nothing adverse about the parental care had been established on the balance of probabilities.

Discussion: The procedure adopted in the lower court

Before turning to the legal argument which is the substance of the appeal as it has been presented to this court, I wish to make a number of observations upon the manner in which the case proceeded before the lower court.

It is apparent from her judgment, that the case before HHJ Hallam was put starkly on the basis of the joint pool of perpetrators responsible for the physical injuries to T-L. Despite being unable to identify the perpetrator of those physical injuries to the required standard, HHJ Masterman was, however, able to make specific findings against JJ amounting to failure to protect T-L, taking steps (in particular deliberately avoiding seeking medical attention) which will have exacerbated T-L's suffering and, finally, joining in a concerted cover up with SW to prevent the court and authorities understanding what had happened to T-L. We were told that the Local Authority had chosen not to rely upon any of those discrete findings on the basis that individually or cumulatively they would not now support a threshold finding in relation to the current children. The case was therefore presented to the learned judge on the single basis that JJ was one of the two possible perpetrators of physical injury to T-L.

Pausing there, I would question the desirability of artificially limiting the judicial consideration to just one, albeit important, aspect of the case. A judge hearing a fresh s 31 application, some years later, about a new family unit which involves a parent about whom adverse findings have previously been made in another family context, should be exposed to the full detail of the available evidence and be permitted to come to her own overview and determination taking into account all of the material insofar as she considers it to be relevant and giving it such weight as she may see fit at the time of her determination. Artificially to limit the judicial exercise in a manner which invites the court to ignore part of the evidence in the case, might well set up the legal point for determination in a clinically clear and legally accessible manner, but it cannot, in my view, represent a proper exercise of the judicial task. In determining whether the threshold criteria are satisfied in relation to each of these three children as at 3^rd March 2011 a judge must be under a duty to acquaint herself with all of the available evidence and then bring it to bear on the ultimate question of whether, in the context of this case, each or any of these three children can be said to be "likely to suffer significant harm" attributable to failures in parental care likely to be given to him as at that date.

I would go further and criticise the decision to tee up the preliminary issue for determination within an otherwise entirely empty evidential context. By the relevant date seven years had passed between the death of T-L and the commencement of child protection procedures in relation to the new family unit. JJ was seventeen when T-L died and she is now twenty five. Much has no doubt happened in her life in the intervening period, some of it has continued to involve SW. On the positive side it is apparently the case that she had lived with these three children as DJ's partner for well over two years prior to the implementation of protective measures and without apparently causing any degree of concern to child protection professionals.

When a local authority issues a s 31 application seeking the court's determination that the threshold criteria is crossed at the present time in relation to a child or children, the court must be under a duty to ensure that it has before it not only evidence of what may have happened in one of the parent's lives years before, but also some account of the events in that parent's life during the following years and the current circumstances in the family unit which is now being brought before the court within the proceedings. The extent, quality and character of the evidence that a judge may need will of course vary from case to case. But in the present case it would seem to me that, in addition to social work evidence as to what, if any, social work and other professional contact there has been with the mother and/or the new family over the intervening period, together with a basic social work assessment of the children's current circumstances within the home, there should be an expectation for the mother to file evidence which should include an up to date statement of what she now says with respect to her care of T-L and the matters that had been the subject of HHJ Masterman's findings.

A judge in care proceedings at the threshold stage has the important responsibility of determining whether or not he or she is satisfied that, at the relevant date, that is therefore in the current period of time, the threshold criteria are or are not satisfied with respect to a particular child. Artificially to limit the judicial exercise to the consideration of facts relating to a period seven years earlier and, further, to limit consideration to only some of those facts, seems to me to fall well short of the required evaluation of the circumstances of the new child before the court in the current time period.

Despite making these negative observations about the process adopted in this case, I have nothing but professional sympathy for HHJ Hallam who was presented with the task of determining the preliminary issue in accordance with ground rules chosen by the parties. The Local Authority chose not to file any additional evidence. A direction for the mother to file a statement was made but produced no more than a position statement. The Local Authority expressly elected not to rely upon any of the negative findings made by HHJ Masterman other than those relating to the perpetrators of physical injury. Having been presented with the case within this restricted compass, HHJ Hallam was further handicapped by the fact that she was not HHJ Masterman, who had plainly regarded the aspects of the case that fell short of the direct physical injury of T-L as being of an equal standard of concern with the infliction of those injuries.

It is sometimes convenient, efficient and realistic for local authorities to make concessions as to aspects of the evidence. Indeed, such no doubt is the stuff of every day experience in courts conducting care proceedings. There is, however, a danger of such matters being elevated to the status, as would be the case in civil proceedings, of points of pleading, thereby removing from the judge's consideration evidence which may, despite the view of the local authority, be of some importance.

The observations that I have made thus far have led me to consider whether the erroneous course, as I see it to be, followed in the lower court, should now be rectified by allowing the appeal in order that a more comprehensive judicial evaluation of all of the relevant evidence, including the mother's account, could be placed before the judge who is to determine the threshold question. However, whilst this case may well be some distance from the paradigm 'one off' case of the childminder in Lancashire, I take the view that such is the present influence of the decisions in Re S-B and Re F that the central point of law that has been raised in this appeal would be very likely to ensure that the matter came back on appeal at the culmination of any re-hearing. I therefore propose to determine the appeal on the basis upon which it was argued and to put no reliance upon the erroneous process, as I find it to be, in determining the whole question of threshold on a narrowly defined legal issue arising from part of a seven year old fact finding determination.

Discussion: The need to establish proven facts

The stricture that the s 31 threshold criteria may only be satisfied on the basis of previously established fact proved on the balance of probability is a fundamental principle that has been described, accepted and endorsed in many of the decisions to which reference has already been made. Wilson LJ rightly described it as a 'bulwark against the state's removal of a child from his family'. It is a stricture which must be at the forefront of any consideration of the issues in this appeal.

There is, however, a need to tease out from the previously decided cases just what is meant by a previously established fact that has been proved on the balance of probability for the purposes of CA 1989, s 31. Is there, or should there be, a difference in the approach of the court between, on the one hand, a case where absolutely no adverse findings of fact have been made, for example where allegations of sexual abuse have failed to be established on the balance of probability, and, on the other hand, a case where findings of past serious child abuse have been made, albeit that it is not possible to prove on balance which of the only two possible perpetrators inflicted that harm? Should the judgments in the previous authorities be read as applying to both of these two categories of case, or only to the former?

The two principal authorities for the statement of the bulwark principle are Re H [1996] AC 563, in the context of s 31, and Re M and R (Child Abuse: Evidence) [1996] 2 FLR 195, in the context of 'any harm which [the child] has suffered or is at risk of suffering' under CA 1989, s 1(3)(e). It is of note that in both of those cases the allegations, which in each case related to sexual abuse, were found not to be proved on the balance of probability. It was therefore held that there was no factual basis either for satisfying the threshold criteria or for finding a 'risk' of future harm under s 1(3)(e).

In Lancashire CC v B [2000] 2 AC 147 serious physical harm was found proved and both parents remained in the pool of possible perpetrators; the s 31 threshold was established on the basis of 'is suffering' with regard to their child, A. It is, however, this authority which is the only one of the group which was required to engage with a likelihood of significant harm in relation to the child-minder's child based on a past finding of serious physical abuse to a different child. As recorded above, the Court of Appeal dismissed the local authority's appeal in relation to the child minder's child and the matter was not considered by the House of Lords.

Standing back from these cases, there is, I would suggest, a qualitative difference, in terms of the need for child protection, between a situation where absolutely no adverse findings have been made as to past child abuse, and a situation where serious findings of child abuse have been made in relation to a child in the joint care of her parents either one or both of whom were the perpetrator(s).

The point is starkly illustrated by extrapolating, not too fancifully, from the facts of the present case. It is now established that young I is the child of both SW and JJ who were the only two possible perpetrators of the gross injuries to their first child T-L. If JJ were now to leave DJ, taking I with her and set up home once again with SW, the law would hold that the previous finding of fact would support a finding that the s 31 threshold were established with respect to I in that household, just as it was some years ago with respect to S. How, one might reasonably ask, can it be proportionate or indeed logical from the perspective of protecting children, to hold that if SW is not in fact part of the equation and JJ has the sole care of I (or jointly with her new partner DJ) the past finding of serious harm to a child in her care has absolutely no legal significance and she is treated in the same manner as any other parent against whom nothing adverse has been found?

How do the observations made in the paragraphs 88 to 93 above sit within the previous authorities? I would suggest that the 'bulwark' that has rightly been erected by the House of Lords relates to the proof of past or present harm in s 31(1)(2)(a) but not, necessarily, to the attribution of harm in s 31(2)(b)(i). The focus of Re H was upon the establishment of the likelihood threshold and the House of Lords held that that could only be achieved on the basis of the proven fact of past or existing abuse. It is of note that in Re O and N Lord Nicholls expressly ties the House's consideration in Re H to the matters 'the court may take into account in assessing whether the child is likely to suffer significant harm' (paragraph 16) and pointed to that approach to s 31(2)(a) as providing that 'protection Parliament intends the threshold criteria shall provide against arbitrary intervention by public authorities' (paragraph 17).

In contrast the House of Lords in the Lancashire case considered the attributability provision in s 31(1)(2)(b) and held that it may be satisfied where it is not possible to prove to the civil standard that one carer, as opposed to another, was at fault. Lord Nicholls expressly recognised that his interpretation meant that 'parents who may be wholly innocent, and whose care may not have fallen below that of a reasonable parent, will face the possibility of losing their child.' The Lancashire case held, therefore, that the s 31 threshold criteria were met on the basis of the proven fact of abuse, but also on the basis of a finding falling short of proof of the perpetrator. On that basis the 'bulwark' finding is the former one establishing that there has been abusive parenting, and not necessarily the latter, relating to perpetrator. These two House of Lords authorities do not indicate that the approach to s 31(1)(2)(b) would or should be materially different if the two potential perpetrators were to separate.

In making those observations I readily accept that Robert Walker LJ took a contrary course in the Court of Appeal in the Lancashire case and that this court is bound to follow that decision. It is with great respect, however, that I suggest that Wilson LJ may have inadvertently overstated the weight of that decision by stating in Re F (paragraph 14) that the Court of Appeal decision was 'a conclusion which was not disturbed on the further appeal to the House of Lords'. The decision was not challenged in the House of Lords and, as Lord Nicholls stated, the correctness of that decision was not a matter before their lordships.

Whereas the ratio of Re O and N relates to the welfare stage and not to establishing the threshold criteria, it is relevant to note Lord Nicholls' approach to the evaluation of future risk in relation to the child on one of the cases, Y. By the time of the hearing the relationship between Y's mother and the other potential perpetrator (her previous partner) had ended. Whilst of the two, the partner was the more probable culprit, Lord Nicholls expressly drew attention to Thorpe LJ's words in the Court of Appeal by placing them in italics:

'But the important factor that the judge must bring into the foundation for the disposal hearing is that he cannot disregard the risk that the mother presents as a primary carer for either Y or a future child'.

Lord Nicholls held that the mother's appeal in Y's case should be dismissed saying 'in assessing risk the judge at the disposal hearing should have regard, among other matters, to the facts that … the mother might have been a perpetrator' (emphasis added). Those words, coupled with Lord Nicholls clear endorsement of Thorpe LJ's reference to 'a future child', indicate that once the core fact of past abuse is established (as per Re H and Re M and R) the court may act on a finding of the real possibility that a parent is the perpetrator at the welfare stage.

It is illuminating to consider the position as it now is under the criminal law. Following the implementation of the Domestic Violence Crime and Victims Act 2004, s 5 in March 2005, a criminal court may now convict a parent who causes or allows the death of a child. The relevant terms of s 5 are:

"(1) A person ("D") is guilty of an offence if—

(a) a child or vulnerable adult ("V") dies as a result of the unlawful act of a person who—

(i) was a member of the same household as V, and

(ii) had frequent contact with him,

(b) D was such a person at the time of that act,

(c) at that time there was a significant risk of serious physical harm being caused to V by the unlawful act of such a person, and

(d) either D was the person whose act caused V's death or—

(i) D was, or ought to have been, aware of the risk mentioned in paragraph (c),

(ii) D failed to take such steps as he could reasonably have been expected to take to protect V from the risk, and

(iii) the act occurred in circumstances of the kind that D foresaw or ought to have foreseen.

(2) The prosecution does not have to prove whether it is the first alternative in subsection (1)(d) or the second (sub-paragraphs (i) to (iii)) that applies."

The death of T-L pre-dates the implementation of DVCVA 2004, s 5 and in any event it would seem unlikely that s 5 would have applied in relation to her death. The evidence may have been insufficient to attribute her death to the 'result of an unlawful act'. The reason for referring to the 2004 Act is to point up the fact that there will now be other cases where parental involvement in a case similar to the present one where the death of a child is attributable to an unlawful act, but it is not possible to identify the perpetrator, but where the non-perpetrating parent's actions are culpable in the terms of s 5(1)(d), the criminal law may mark that behaviour with a conviction and a sentence of up to 14 years imprisonment (DVCVA 2004, s 5(7)). Debate is currently taking place in Parliament to extend s 5 to cover cases of serious injury falling short of death.

An offence under DVCVA 2004, s 5 is now listed in Schedule 1 of the Children and Young Persons Act 1933. A parent convicted of a s 5 offence would therefore become 'a Schedule 1 offender' or (to use the modern parlance) 'a person posing a risk to children' (Home Office Circular 16 of 2005). The current, revised, edition of Working Together to Safeguard Children (2010), Chapter 12 gives guidance, which must be followed by local authorities, on the approach to be taken to those convicted of a Schedule 1 offence. The existence of the conviction should trigger an assessment to determine whether the individual continues to present a risk of harm to children (paragraph 12.7).

The development of the criminal law leads to contemplation of a situation where a parent may have served a substantial prison sentence for culpable involvement in the death of a child, where it is not possible to establish which of two or more carers actually caused the death, yet (on the basis described by Re S-B and Re F) the s 31 threshold criteria would not be met if that same parent on release from prison were to set up home with a new partner and a different child. Whilst no express reference is made in Re S-B or Re F to this situation, it does not seem that the mere fact of there being a criminal conviction based on precisely the same facts would alter the position in relation to the future likelihood of harm under the threshold as described in those two cases.

This apparent potential difference in the approach of the criminal law and the law relating to child protection is, of course, in part to be explained by the fact that the criminal offence is looking backwards and attributing culpability for past behaviour, whereas the s 31 threshold in the context of different children in a new family is looking forward and assessing the likelihood of future harm. But I would suggest the difference is also in part due to the fact that the criminal offence moves beyond the bald attribution of culpability for the actual harm to the child and attaches equal culpability to a parent who, even if they did not commit the harmful act, ought to have known of the risk of the type of harm that occurred, or failed to protect the child from the risk of that harm (s 5(1)(d)). The current state of child protection law under CA 1989, s 31 as described in Re S-B and in Re F seems limited to identifying the perpetrator of the physical harm and no more; it does not regard a s 5(1)(d) class of finding to be 'a finding of fact' for the purpose of establishing future harm.

Equally, the criminal law, through s 5(1)(d), puts both the perpetrator of the actual harm and the other parent who has failed to protect the child in precisely the same position. Both may be found guilty of the same offence and it matters not that the jury may be unable to be satisfied as to the identity of the actual perpetrator. 'Joint and several liability' is a phrase from ordinary civil law which is not altogether inapt to describe this situation.

There must be a danger that focus on whether or not the past perpetrator of the actual harm is or is not identified on the balance of probabilities when looking at the potential for future harm to another child, may lead a court to ignore other culpable aspects of past behaviour. In this category would go matters relating to failure to protect, however in any given case such matters may not carry substantial weight in establishing likelihood of future harm where the parent in question is no longer in a relationship where the need to protect arises.

Of more concern is the situation that arises where the court is unable to identify the perpetrator precisely because the two parents are deliberately colluding with each other after the event in order to prevent the authorities and the court from gaining insight into the identity of the perpetrator. In short, where the parents' very behaviour is directly responsible for the fact that the court is unable to identify a perpetrator. This case provides clear evidence of past behaviour of this type. HHJ Masterman found that both parents were lying to the court in order to cover up the identity of the perpetrator. I have already quoted his conclusion in full, but the following extracts highlight the present point:

"If one parent is innocent of actually abusing T-L, then he or she knows it was the other. … There is no doubt in my mind that SW and JJ are colluding with one another to hide the truth from the rest of the world. One is protecting the other or they are both protecting each other. … it must be debatable which is worse, to inflict this injury or to protect the person responsible. On these findings it is very difficult to see how either parent …. could be safely entrusted in the future with the care of S. Sadly unless and until they are prepared to tell the truth there seems little prospect of this situation changing." [emphasis added]

On the basis of the submissions of Mr Storey and Miss Rowe, relying upon Re S-B and Re F, these very significant findings, which are findings of fact against the mother JJ in this case, must be ignored in evaluating the likelihood of future harm to a subsequent child in her care. Thus it is said that, despite these being findings of fact, if the perpetrator is not identified then there is no relevant finding of fact upon which the s 31 likelihood threshold could be satisfied.

With respect, I do not consider the very narrow evaluation of this issue which it is said the short judicial statements in Re S-B and Re F require is supportable on the basis of the earlier authorities, the principal one in this context being the Court of Appeal judgment of Robert Walker LJ in Lancashire County Council v B which allowed the appeal in relation to the child minder's child, B. In my view the Lancashire case can be readily distinguished from the present case. The only culpable matter that could have been established against the child minder was that she had caused the injuries to child A. The evidence was insufficient to establish, on the balance of probability, that she was the perpetrator. There was no suggestion that she would in the future be part of a group of carers with the parents (about whom, collectively, there was a finding). There was therefore no proven fact upon which the risk of future harm could be established.

The Lancashire case was truly a one point case. There were no other adverse findings made against the child minder. The present case is, in my view, very different. Here there are a number of adverse findings of fact that have been made against JJ, albeit that they fall short of identifying the perpetrator. In short terms they are:

a) Gross and substantial collusion expressly designed to prevent the court identifying the perpetrator;

b) Failure to protect T-L;

c) Deliberately keeping T-L away from health professionals in order to avoid the detection of injury.

The Lancashire case in the Court of Appeal is undoubtedly authority for the proposition that where there is absolutely no adverse finding of fact against a parent, the likelihood threshold cannot be established. It is not, however, authority for the obverse proposition that you can only establish the likelihood threshold if you can identify the perpetrator irrespective of any other adverse findings that have been made.

Likewise Re S-B is not authority for the proposition 'if you cannot identify the past perpetrator, you cannot establish future likelihood'. In that case, which was of a one-off ('whodunit') injury, there was no question of failure to protect and no finding of collusion. Thus when, in paragraph 49, Baroness Hale describes as it as impermissible to hold that the likelihood threshold in relation to the non-injured child, William, was established on the basis only that there was 'a real possibility' that the mother had caused the injury, she does so in the context of there being no other adverse findings against that mother.

In the same manner, in Re F, apart from the implication that there was a finding of failure to protect the child in the previous proceedings from one fracture, there were no adverse findings against the father other than that he was, with the mother, in the pool of possible perpetrators for two fractures on the basis of that being 'a real possibility'.

The difference that I am seeking to describe is the difference between the position of (a) the child minder in Lancashire and (b) a DVCVA 2004, s 5(1)(d) parent in the criminal context. In the former, it is a one point case, if it is not proved that the individual perpetrated the past injury there is no adverse finding, in the latter case, irrespective of who caused the injury, there is a finding of highly culpable behaviour which may include the generation of the situation that leaves the court unable to identify the perpetrator.

At paragraphs 79 to 87 I criticised the narrow and legalistic approach adopted before HHJ Hallam in the present case. In my view, where there are significant adverse findings made on the balance of probability against a parent in previous proceedings, irrespective of the previous court having been unable to identify a perpetrator, a court seized of an application under s 31 has a duty to look at all of the available material including those adverse findings. Each case will no doubt differ from each other both on the question of whether in the new circumstances the s 31 threshold is actually established and on the question, if it is, whether the welfare of the child requires protection under a s 31 order. These are matters for courts to evaluate on a case by case basis and are not the stuff of a blanket policy based on the single point that, irrespective of what else has been found, if the past perpetrator has not been proved the threshold cannot be crossed.

On the above basis, if the matters that I have just described were the only issues in the case, I would have allowed the appeal and remitted this case for it to be determined afresh with the court taking into account all relevant evidence, including the adverse findings that were made against JJ (see paragraph 109 above) together with the other issues that I have identified at paragraph 83.

Re S-B and Re F

Before turning to the all important question of the approach to be taken to the judgment of the Supreme Court at paragraph 49 of Re S-B and the judgment of Wilson LJ in Re F, it is useful to take stock of the conclusions rehearsed earlier in this judgment in relation to the case law as it was prior to these two cases:

a) The leading decisions which establish that the threshold criteria relating to likely future harm or 'risk of suffering' harm in s 1(3)(e) must be based on fact(s) established on the balance of probabilities are both cases where absolutely no adverse findings of past harm had been made (Re H) or not harm of the relevant category (Re M and R);

b) Re H was expressly focused upon the likelihood provision in s 31(2)(a) and not the attributability provision in s 31(2)(b)(i) (see Re H and Lord Nicholls in Re O and N);

c) Where past harm has been proved, the attributability provision will be established in relation to a couple who are the only potential perpetrators irrespective of the fact that the court is not able to prove on balance which of them was the actual perpetrator (Lancashire);

d) In a case such as (c), the evaluation of risk at the welfare stage will be undertaken on the basis that a parent is no more than a potential perpetrator, notwithstanding that that approach may lead to a wholly innocent parent losing their child (Lancashire);

e) The approach in (d) to the welfare stage regards the finding of 'potential perpetrator' as a finding of 'fact' for the purposes of s 1(3)(e), and not therefore an unproved allegation, and is compatible with the 'bulwark' decision of Re M and R (Re O and N);

f) At Court of Appeal level, the decision in the Lancashire case to the effect that, where there are no other adverse findings made with respect to a parent, a finding that that parent is merely a potential perpetrator of past proven harm, is insufficient to meet the s 31 threshold, is binding upon this court;

g) The Court of Appeal decision in the Lancashire case applies to one issue uncertain perpetrator cases, where there is no other adverse finding against the parent.

Against that background how is paragraph 49 of the Supreme Court's judgment in Re S-B to be interpreted?

In addressing that question, the position is complicated in that it became apparent during this hearing that there are a number of competing interpretations of paragraph 49, and in particular what Baroness Hale was referring to in the phrase 'as already explained'.

One suggested interpretation was that the reference is to paragraph 23, which is in these terms:

"However, it is worth noting that the Court of Appeal [in the Lancashire case] had confirmed that the criteria were not satisfied in respect of the childminder's child, B, because he had not been harmed at all. The only basis for suggesting that there was any likelihood of harm to him was the possibility that his mother had harmed the other child and that had not been proved: In Re H [1996] AC 563 applied. The local authority did not appeal against this."

Another suggestion proffered was that 'as already explained' simply refers back to the generality of the Supreme Court's criticism of the approach taken by the trial judge to the standard of proof and the use of a 60/40 apportionment of likelihood.

A further, more general interpretation of the paragraph, was provided by Hedley J in the case of K County Council v W, D, L and A, C and M [2010] EWHC 3342 at paragraphs 19 and 20, having quoted paragraph 49 of Re S-B:

"19. … It is argued that that is not consistent with the conclusion in Re B, O and N and that in this paragraph the learned justice is requiring proof of identity of perpetrator."

20. That outcome would be surprising in the context of the judgment as a whole. It would seem to import the very consequence (described as 'grotesque') that Lord Nicholls was at pains to avoid. I have understood these words simply as a reminder that in care cases the threshold must be established by proof of facts or that in a private law case (where no threshold is required) the evaluation of risk in s 1(3)(e) of the Act is founded on proven facts. In my judgment where, as here, the threshold would have been established, that is a sufficient factual basis on which to undertake a risk assessment albeit on the basis of 'uncertain perpetrator'. It cannot be right that in a case where a parent was one of only two who inflicted non-accidental injury to a baby that fact is irrelevant to any future assessment of risk because the court cannot go the final step and decide as between the two."