This is a guest post by Saara Parijaat. Saara is an undergraduate law student at the London School of Economics and hopes to pursue a career at the Bar.
Re E (Special Guardianship; Pubic Law Proceedings) [2026] EWCA Civ 1167 is a recently published Court of Appeal judgment which clarifies the approach the court should adopt when making a Special Guardianship Order (SGO) at the end of care proceedings. Practically, the outcome of the case means that a court cannot determine whether to make an SGO simply by asking what outcome best serves the child’s welfare. Where an SGO is proposed as the permanent outcome of care proceedings, the court should typically ask whether the legal threshold for state intervention has been crossed by accounting for whether the child is suffering, or is likely to suffer, significant harm. It must then consider whether separating the child from their family in this way is a necessary and proportionate interference with family life.
What Is a Special Guardianship Order?
In simple terms, an SGO is a court order which gives a person responsibility for a child on a long-term basis without legally severing the child’s relationship with their birth parents. The special guardian shares parental responsibility with the child’s parents but, in most matters, can exercise it to the exclusion of the parents. This gives the special guardian a much stronger legal position than an ordinary carer and is intended to provide the child with a stable and permanent home. Unlike adoption, however, an SGO does not bring the legal relationship between the child and their birth family to an end.
Nevertheless, as the judgment of the Court of Appeal in Re E emphasizes, an SGO is a “very significant order” for the child and for their relationship with their family. For a large part, this is because SGOs are underpinned by the concept of permanence. This means that the order is intended to provide a long-term plan for the child’s upbringing, giving them stability and a sense of belonging. For that reason, an SGO can have profound consequences for family life even though it falls short of adoption. This significance is important to understanding Re E, because it explains why the Court of Appeal rejected the idea that an SGO at the end of care proceedings could be approached through a simple analysis about which of the available options would best serve the child’s welfare.
Background to Re E
The case concerns a six-year-old child, E, who had already been the subject of earlier care proceedings. The Local Authority first became involved before her birth, and concerns continued after she was born, principally in relation to her mother’s mental health. Although the mother engaged with treatment and was at times described as stable by her GP, parenting and independent social work assessments continued to raise concerns about her insight into the risks her mental ill-health might pose to E and about her ability to care for her safely.
The first care proceedings ended in 2021 with an SGO in favour of E’s maternal grandmother, with the intention that E, her mother and grandmother would live together. E remained in their care until March 2024, when the Local Authority received a referral containing a number of serious allegations. Most of those allegations were later disputed and not pursued, but social workers who visited the home became concerned about the mother’s presentation and about E’s behaviour and account of events. As a consequence, E was removed and placed in foster care under an interim care order in a new set of care proceedings.
By the time of the final hearing, E had been living with her current foster carer for almost 18 months while continuing regular contact with her mother and grandmother. The Local Authority sought an SGO in favour of the foster carer, while the mother and grandmother sought E’s return to one or both of them. The Cafcass guardian supported the Local Authority’s plan.
The decision in the lower court
The judge in East London Family Court approached the decision whether to make an SGO by treating E’s welfare as the paramount consideration. In his view, determining what was best for E required a balancing exercise in which the advantages and disadvantages of each realistic option available to the court were weighed against one another. He structured that exercise around the Welfare Checklist in s 1(3) of the Children Act 1989, which directs the court to consider matters such as the child’s needs, the likely effect of any change in circumstances, the harm the child has suffered or may suffer, and the ability of those caring for the child to meet those needs.
Applying that framework, the judge placed significant weight on the stability E had found with her foster carer and the risk of emotional harm if she were moved back into an arrangement with her mother and maternal grandmother. Particularly, he relied on concerns about the mother’s mental health and the relationship between the mother and maternal grandmother, including the possibility of future relapse. The mother’s lawyer in the Court of Appeal argued that, while the judge referred to a continuing risk of relapse, he did not analyse the likelihood or degree of that risk in light of more recent evidence about the mother’s stability.
Ultimately, the judge concluded that E’s welfare was best served by remaining with her foster carer under an SGO. He considered that the risks associated with placing E back with her mother and grandmother, and the possibility of that placement later breaking down were too great, even if the family were supported by a supervision order. The mother and grandmother appealed this decision, and the case proceeded to the Court of Appeal.
The flaw in the lower court’s reasoning
The judge had erroneously treated the case as turning solely on E’s welfare. He considered his task to be deciding which of the available options would best serve her interests and, on that basis, did not separately address either the criteria in s 31(2) of the Children Act 1989 (the ‘threshold criteria’) or whether the proposed SGO was a necessary and proportionate interference with family life. When asked to clarify his reasoning, the judge explained that he had not considered the s 31 threshold because he did not think it necessary or appropriate to make a care or supervision order under s 31, although he had considered harm and the likelihood of harm in his analysis.
The Court of Appeal in Re E held that this approach was wrong. Although the s 31(2) threshold does not, as a matter of law, directly apply where the court ultimately makes an SGO rather than a care or supervision order, the court should typically determine whether that threshold has been crossed before deciding whether to make an SGO at the end of care proceedings. That was particularly important here because one of the realistic alternatives was for E to return to her mother and/or grandmother under a supervision order. By deciding that there was no need to consider threshold because he was not making an order under s 31, the judge had effectively put the matter the wrong way round: he could not properly rule out a supervision order without first asking whether the statutory threshold for making one was satisfied.
More broadly, the Court of Appeal rejected the idea that the judge’s task was simply to undertake a welfare balancing exercise. He was also required to determine whether placing E permanently with the foster carer under an SGO was justified and proportionate. Even if he was not strictly required to determine the s 31(2) threshold because he did not ultimately make a care or supervision order, expressly addressing it would have provided an important analytical framework through which to assess the various available options. In that sense, threshold could be considered a part of the reasoning needed to decide whether such a significant intervention in E’s family life was warranted.
The appeal by the mother and grandmother was upheld and the case was sent back to the Family Court for re-hearing.
The Court of Appeal’s clarified approach
Re E clarified that Special Guardianship cases should, typically, be approached by first considering whether the threshold criteria are established. Although the statutory provision containing the criteria, s 31(2), does not, as a matter of strict law, apply to the making of an SGO, the purpose of the threshold criteria remains highly relevant. The criteria seek to balance two equally serious concerns: the state intervening to remove a child from their family without sufficient justification, and the state failing to safeguard a child from harm within that family. The Children Act 1989 addresses that tension by setting a threshold in s 31(2) which must be crossed before the court can consider what order should be made to protect the child. Once that threshold is crossed, the court is required to treat the child’s welfare as its paramount consideration, having regard to the Welfare Checklist that the first-instance judge mistook to be his starting point.
The Court of Appeal held that both the context of the proceedings and the intrusive nature of such an order mean that the court should adopt broadly the same approach as it would when considering a care order, where the threshold criteria do formally apply. This is because, even with SGOs, it is not enough that a child might be better off living in another family; the threshold exists precisely because state intervention requires more than a conclusion that another arrangement would better promote the child’s welfare.
There is also an important connection between threshold and proportionality. The proportionality requirement also reflects the court’s obligations under Article 8 of the European Convention on Human Rights under which any interference with family life must be proportionate to the needs of the situation. In practical terms, this reinforces the longstanding principle of English child law that the court should make the least interventionist order capable of protecting the child. Applying the same, or a similar, approach to that required under s 31 helps ensure that the court properly recognises the effect of the proposed order and adequately considers whether that intervention is necessary and proportionate.
Unless the court has a clear understanding of what significant harm the child has suffered or is likely to suffer, it becomes difficult to assess the future risk of harm at the welfare stage or to evaluate whether the proposed plan is actually capable of protecting the child from that harm. Child protection proceedings necessarily require the court to identify what, specifically, the child requires protection from. By analogy, the same reasoning applies when an SGO in favour of a foster carer is being considered at the conclusion of care proceedings.
A proper threshold analysis also helps ensure that the court makes clear findings on disputed matters which are significant to its decision. That was particularly important in Re E, where the risks said to arise from the mother’s mental health and the relationship between the mother and maternal grandmother were central to the proposed permanent placement of E outside her family. Identifying the harm relied upon, and making clear findings about the facts said to establish that harm, provides the foundation for the later welfare and proportionality analysis.
Comment
Re E continues the line of cases in which the courts attempt to strike the right balance with regard to state intervention in keeping a child from their family when it is in the child’s best interests. In future, courts may be less able to rely on the fact that a child is settled and doing well in a foster placement as sufficient justification for making that arrangement permanent. Instead, greater attention has to be paid to the evidence said to justify keeping the child outside their family, particularly where the risks relied upon are disputed or may have changed over time.