Re B (Children)

Re B (Children) [2008] UKHL 35 is a landmark decision of the House of Lords concerning the standard of proof applicable in care proceedings brought under Part IV of the Children Act 1989. The appeal arose from care proceedings in which a local authority sought care orders in respect of four children on the basis that they had suffered, or were likely to suffer, significant harm attributable to the care given by their parents. The central factual dispute concerned whether one of the children had been sexually abused, and if so, by whom.

At first instance, the trial judge found, on the balance of probabilities, that the father had sexually abused one of the children. However, the judge expressed the view that, because the allegation was a serious one, something more than the ordinary balance of probabilities was required before such a finding could be made — reflecting a perception that had gained ground in some judicial and academic commentary that the standard of proof in family proceedings operated on a sliding scale, whereby the more serious the allegation, the more cogent the evidence needed to be before the civil standard could be satisfied.

The Court of Appeal upheld the trial judge's approach, affirming that the inherent improbability of a serious allegation such as sexual abuse meant that the evidence required to satisfy the civil standard was correspondingly heightened. The local authority and the children's guardian appealed to the House of Lords, challenging the correctness of this formulation as a matter of law.

The case therefore arrived before the House of Lords as an occasion to resolve, definitively, a doctrinal uncertainty that had persisted in English family law for many years: whether the balance of probabilities, as the applicable standard of proof in civil proceedings, is a single, uniform standard or whether it operates as a flexible or enhanced standard calibrated to the gravity of the allegation.

The appeal was heard by a panel of five Law Lords — Lord Hoffmann, Lord Hope of Craighead, Baroness Hale of Richmond, Lord Carswell, and Lord Brown of Eaton-under-Heywood — each of whom delivered a speech. The opinions of Baroness Hale and Lord Hoffmann have attracted the greatest academic and judicial attention and are regarded as the leading statements of principle.

The primary issue before the House of Lords was whether the standard of proof applicable to findings of fact in care proceedings under the Children Act 1989 — and specifically to the threshold criteria under section 31 — is a uniform balance of probabilities, or whether it requires a higher degree of probability where the alleged facts are inherently serious, such as allegations of sexual abuse. This required the court to examine the true meaning of the civil standard of proof and to address apparent inconsistencies in earlier authority.

A subsidiary but practically significant issue was whether, where a court cannot identify which of two possible perpetrators caused harm to a child, either or both may be treated as a possible perpetrator for the purposes of subsequent welfare assessments and the risk evaluation that feeds into the making of a care order. This raised questions about the relationship between the fact-finding stage and the welfare stage of care proceedings.

Lord Hoffmann, with characteristic analytical precision, addressed the origins of the confusion in the law. He acknowledged that a line of authority — including passages in Re H and R (Child Sexual Abuse: Standard of Proof) [1996] AC 563 — had been read as supporting the proposition that serious allegations require more cogent evidence before the civil standard is met. He rejected this reading. The civil standard of proof, Lord Hoffmann confirmed, is simply whether the occurrence of the fact in question is more probable than not. It does not admit of degrees: either the court is satisfied on the balance of probabilities or it is not.

Lord Hoffmann drew on the authoritative treatment of the civil standard in In re H and R itself, noting that Lord Nicholls had there stated that the more serious the allegation, the less likely it is to have occurred, and therefore the stronger the evidence needed to prove it. Lord Hoffmann explained that this observation, properly understood, is a statement about the assessment of evidence and the inherent probabilities in ordinary human experience — not a statement that the legal standard itself is elevated. Inherent probability is simply one of the factors a fact-finder takes into account when evaluating whether something more probably occurred than not.

Baroness Hale delivered what has become the most frequently cited speech. She stated with clarity that there is only one civil standard of proof and it is the balance of probabilities. She emphasised that this standard is not subject to any gloss, qualification, or escalation by reference to the gravity of the matter alleged. She observed that the seriousness of an allegation does not change the standard; it may affect the evaluation of the evidence and the weight to be accorded to inherent probabilities, but once the court is satisfied that something is more likely than not to have happened, the threshold is crossed regardless of how serious the finding is.

Baroness Hale reinforced this analysis by reference to the policy underpinning the Children Act 1989. The threshold in section 31(2) requires the court to be satisfied that the child "is suffering, or is likely to suffer, significant harm" attributable to parental care or the child being beyond parental control. This threshold is designed to protect children from harm. To require a higher standard of proof for more serious allegations would paradoxically afford less protection to children at greatest risk: the more serious and egregious the abuse alleged, the harder it would become, on a sliding-scale approach, to satisfy the threshold. This result would be antithetical to the protective purposes of the statute.

The House also addressed the treatment of Re H and R [1996] AC 563 directly. Their Lordships did not overrule that decision but clarified that it had been misread. The majority in Re H and R had held that the standard of proof for the threshold criteria was the balance of probabilities applied with appropriate care. Passages in that decision had been misinterpreted as introducing a heightened standard. The House in Re B firmly corrected that misinterpretation, holding that it was wrong to read Re H and R as authority for a flexible civil standard.

Lord Hope of Craighead agreed with the analysis of Baroness Hale and Lord Hoffmann. He noted that the confusion had arisen partly because courts had conflated two distinct questions: first, the standard of proof (which is fixed and binary — more probable than not), and second, the quality and cogency of evidence required to discharge that standard in any given case (which may vary according to the circumstances, including the inherent probability or improbability of the alleged event). These are conceptually separate exercises, and conflating them had produced a distorted legal standard that was neither principled nor workable.

On the question of unidentified perpetrators, the House of Lords also addressed whether, where it is not possible to determine which of two individuals caused harm, either may be treated as a "possible perpetrator" for the purpose of welfare decisions and risk assessments. The House endorsed the approach that, where the evidence establishes that harm was caused by one of a limited pool of individuals but does not permit identification of which one, a court may proceed on the basis that each member of the pool is a possible perpetrator. This does not involve a finding of proved fact against any individual but is relevant to the welfare analysis and to the court's assessment of future risk.

Baroness Hale was careful to note that the "possible perpetrator" approach does not lower the standard of proof or permit speculative findings. A person may be included in the pool of possible perpetrators only if the evidence establishes a real possibility — not a mere theoretical one — that they caused the harm in question. This preserves the integrity of the fact-finding exercise while allowing the court to take into account genuine uncertainty at the welfare stage, which is itself governed by the paramountcy principle and a holistic assessment under section 1 of the Children Act 1989.

Lord Carswell and Lord Brown of Eaton-under-Heywood each delivered concurring speeches. Lord Carswell noted the practical importance of clarity in this area, given the large number of care proceedings heard each year in which findings of serious harm are contested. He agreed that the attempts by courts to apply a nuanced or graduated civil standard had introduced uncertainty and inconsistency into proceedings that required principled and consistent adjudication. Lord Brown similarly emphasised that the law had to be stated simply and clearly, and that the balance of probabilities — without further qualification — was the correct and complete statement of the standard.

The House also implicitly endorsed the broader principle that the fact-finding stage of care proceedings is governed by ordinary rules of evidence and proof, while the welfare stage that follows a finding that the threshold has been crossed is governed by a different and distinct framework — one that is not constrained by the same evidentiary requirements but is instead focused on the best interests of the child. This two-stage structure, clearly delineated, was confirmed as the correct analytical framework for care proceedings.

The House of Lords allowed the appeal and held, unanimously, that the standard of proof applicable to findings of fact in care proceedings is the simple balance of probabilities — that is, whether the alleged fact is more probable than not. There is no heightened or enhanced civil standard applicable to serious allegations. The seriousness or gravity of an allegation does not alter the standard of proof; it is a matter relevant only to the assessment of evidence and the evaluation of inherent probabilities, which the fact-finder takes into account in the ordinary course of deciding whether the standard has been met.

The House further held that where harm cannot be attributed to a specific individual from a known pool of possible perpetrators, each person in that pool may be treated as a possible perpetrator for the purposes of welfare assessments and decisions about future risk, provided there is a real — not merely theoretical — possibility that they were responsible. This finding does not constitute a proved fact against any individual but is a legitimate and necessary input into the welfare analysis that follows threshold.

The case was remitted for reconsideration in light of the correct legal standard, as the courts below had applied an incorrectly formulated test in making their findings of fact and assessing the threshold criteria.

Re B (Children) [2008] UKHL 35 is the definitive authority on the standard of proof in English family law proceedings. It settled, finally, a controversy that had been generating inconsistent decisions for more than a decade following Re H and R. Its practical importance is immense: every care case in which a local authority makes a disputed allegation of abuse or neglect is governed by the principle articulated here. The uniform balance of probabilities standard is now consistently applied at all levels of the family judiciary, from the Family Court upwards.

The decision has been applied and affirmed in numerous subsequent cases. In Re S-B (Children) [2009] UKSC 17, the newly constituted Supreme Court — the successor to the House of Lords — applied the principles in Re B to refine the law on the identification of perpetrators in the pool, confirming that inclusion in the pool of possible perpetrators requires more than mere suspicion but does not require a finding to the balance of probabilities standard. This sequential refinement demonstrated the enduring influence of Re B as a constitutional statement of evidential principle in family law.

More broadly, Re B reinforces the principle that the civil standard of proof in English law is a single, unified standard. It aligns family proceedings with the general civil law position confirmed in In re B (Allegation of Sexual Abuse: Standard of Proof) and, more generally, with the approach of the House of Lords in the criminal-adjacent civil contexts addressed in Secretary of State for the Home Department v Rehman [2001] UKHL 47 and R (McCann) v Crown Court at Manchester [2002] UKHL 39, in which the courts emphasised that the binary nature of the balance of probabilities does not prevent tribunals from taking account of the inherent improbability of alleged facts when evaluating evidence.

For law students and practitioners, Re B stands as essential reading for at least three reasons: it authoritatively resolves the standard of proof question in child protection law; it offers, particularly in the speech of Baroness Hale, a model of clear judicial reasoning about the relationship between legal standards and evidential assessment; and it illustrates the extent to which the law governing care proceedings is animated by the protective purposes of the Children Act 1989, with the legal standard calibrated to serve — not undermine — the welfare of children who may be at risk of serious harm.