The Section 31 Children Act 1989 Threshold Criteria

Before a court can make a care order, it must be satisfied that the ‘threshold criteria’ in Section 31 of the Children Act 1989 have been met. The order must also promote the welfare of the child. The main effect of a care order is to give parental responsibility for the child to the local authority. If a care order is made, the child can be removed by the local authority at any time. The threshold criteria, therefore, play a significant role in that they prevent care orders being made simply based on what is in the best interests of the child. However, as this paper will demonstrate, the threshold for state intervention at various stages of the child protection process has been extremely controversial.

Section 31 Children Act 1989: Threshold Criteria

One of the great problems in the law of child protection is that if the wrong decision is made, great harm may result. As Bainham said: “The law in this area has to strike a careful balance between enabling the protection of children at risk of harm, with protecting the rights to respect for family life for children and their parents” Not surprisingly, there is substantial case law on the interpretation of s.31 of the Children Act 1989 and the House of Lords have considered their interpretation in some important cases which will be assessed afterwards. An analysis would be made about whether the courts have interpreted the threshold criteria in a strict or lax way. If interpreted in a strict way, this would imply that it would be more problematic for the local authority to satisfy the grounds for a care order.

Under the first limb of threshold, the local authority must show that the “child concerned is suffering, or is likely to suffer, significant harm” when applying for a care order. Although quite straightforward, there have been concerns about when the alleged state of affairs must be shown to exist.

‘Is suffering’

In Re M, the key issue was the meaning of ‘is’ in the threshold criteria. In this case, the children’s father had murdered their mother. Three of the four children were placed with their aunt. The remaining child was placed with foster carers, but later joined her aunt. However, the local authority still wanted a care order just in case it became necessary to remove the child from the aunt’s house. By the time the case came to court, it was hard to say that the child was, at that time, suffering from significant harm or that she was likely to. Their Lordships hence explained that the correct test was “whether the child was suffering from significant harm at the time when the local authority first intervened”. Given that interpretation of the threshold, this was clearly satisfied. That decision is clearly correct, as their Lordships indicated or else it would be difficult for the local authority to obtain a care order in cases where children were put in excellent care. It was a lax interpretation of the threshold criteria as a strict interpretation would have made the law hard to operate.

‘Likely to suffer significant harm’

The alternative ground on which the local authority can satisfy the first limb of the threshold criteria under s.31(2)(a) of the Children Act 1989 is the likelihood of future significant harm. One of the major issues that local authorities encounter is that predictions that child abuse will occur are difficult to make. Removing a child on the ground of speculative harm is controversial as it is impossible to know whether or not the harm would occur.

In Re H, there were several issues for the court. The first was the meaning of ‘likely’. Their Lordships held that ‘likely’ meant that significant harm was a ‘real possibility’. It was not necessary to show that the harm was probable in the sense of ‘more likely than not’. This is a notably lax interpretation of the threshold criteria. They also held that it must be shown, on the balance of probabilities, that the threshold was satisfied. They rejected the view that the criminal burden of proof should be applied. However, rather confusingly, Lord Nicholls said that “where there was a more serious allegation, more evidence would be required to establish it on a balance of probability than a case of a less serious allegation”.

This dicta was reconsidered by the court in Re B where their Lordships made it clear that Lord Nicholls was not suggesting that, in cases of serious abuse, the criminal burden of proof should be used. The civil balance of proof should be applied in all cases under the Children Act 1989. Instead, what Lord Nicholls implied, was that “some allegations will be inherently unlikely and they will require more evidence to establish them than others”. This interpretation was followed in Re S-B. This aspect of the decision in Re B and Re H is perhaps best viewed as a lax interpretation, although it is probably not as lax as it could have been. Requiring a criminal burden of proof would have indeed made it very difficult for the local authority to obtain a care order. However, Lord Nicholls’ approach to the standard of proof, as well as its wider implications for protecting children at risk of harm has attracted strong academic criticism.

Re B also confirmed another aspect of the decision in Re H. Risk of significant harm can only be established based on ‘primary facts’ which would then have to be proved on the balance of probabilities. Mere suspicions are not sufficient. In Re H, a 15 year old girl alleged that she had been raped by her stepfather. The local authority sought a care order in respect of the girl’s three younger siblings who continued to live with the man. There was a strong suspicion that the older girl had been abused and that the younger girls were at risk of being harmed. However, as it had not been proved on the balance of probabilities that the girl had been abused, no primary facts had been proved and thus, no care order was granted. As Lord Hoffman in Re B stated, “either a fact happened or it did not and there was nothing in between”. If there are no facts to support a finding of risk of future harm, the court is powerless to proceed. This is, undoubtedly, a strict interpretation of the threshold criteria. The majority of their Lordships saw this issue in terms of parental rights; parents should not have their children removed on the basis of suspicions. However, it is suggested that this is not a safe approach to risk taking with children. The reason why it is unsafe is that it would be very difficult for the local authority to safeguard a child’s right to be protected from abuse even when there is a serious risk of danger.

As this analysis suggests, there are evidential problems and difficulties of predicting the future. The problems of proof partly explain the lengthy delays which can occur in child protection proceedings. With the introduction of the Children and Families Act 2014, there is now a 26-week time limit for completing care proceedings with the possibility of extending the time limit for up to 8 weeks, if this is necessary to resolve the proceedings justly. However, an important issue that arises here is whether this is achievable in complex cases. ‘Timescales can end up replacing professional judgment’.

‘Harm attributable to the care given or likely to be given or the child’s being beyond parental control’

Uncertainty about who caused harm to the child is also another issue which local authorities and courts generally encounter. The issue of the ‘unknown perpetrator’ was addressed in the case of Lancashire CC v B. In this case, it was clear that the child had suffered harm. However, it was not clear whether it was the parent or the child minder who had caused harm to the child. Their Lordships held that as long as it was clear that the abuse was caused by a parent or a child minder, it did not matter which had perpetrated the abuse. On the other hand, where it is not clear whether the harm was caused by a parent or someone who was not a primary carer of the child, then no care order could be made. Although the House of Lords provided a clear guidance on when the threshold criteria would be satisfied in the case of an ‘unknown perpetrator’, they provided limited guidance on how the court should deal with an unknown perpetrator when deciding whether a care order should be granted.

Their Lordships returned to that issue in Re O and N, where it was emphasised that “just because the threshold criteria was satisfied, it did not automatically mean that a care order had to be made”. In one of the appeal cases, it was evident that the child was harmed by one of the parents, who had since separated. The child lived with the mother. The issue for their Lordships was whether the suspicions that the harm may have been caused by the mother should be considered. Their Lordships held that suspicions could be considered at the welfare stage. Lord Nicholls however emphasised that social workers should be careful in such cases to treat the parents as potential perpetrators, not proved perpetrators. Therefore, in Re S-B, it was confirmed that if both parents were possible perpetrators, the court might decide to remove the child as they were at risk of harm. It is therefore submitted that in Lancashire, the House of Lords took a noticeably lax interpretation of the threshold criteria as the children could be removed from their parents even if they did not perpetrate the abuse. However, it was probably not as lax as it could have been as it was necessary to show that a primary carer of the child was harming the child.

‘Significant harm’

Even if the facts are known, there is much controversy over how much suffering the child should face before the local authority could intervene. Harm is very widely defined in s.31(9) of the Children Act 1989 as the “ill-treatment or the impairment of health or development.” ‘Health’ means ‘physical and mental health’. ‘Development’ includes “physical, intellectual, emotional or behavioural development”. As a result of the Adoption and Children Act 2002, the definition of ‘harm’ also includes the ‘impairment suffered by hearing or seeing the ill-treatment of another’. The legislation, however, does not define the line between ‘harm’ and ‘significant harm’.

The Court of Appeal in Re C (A Child) explained that to be significant, the harm had to be “great enough to justify the local authority interfering in the autonomous life of the family”. The test will therefore be subjective to the particular circumstances. This raises many questions. If a local authority finds that a child is living in a house where the family’s diet is unhealthy and where the children spend all their time in front of the television, what should be done? Joanna Nicolas, a child protection consultant, believes that “obesity should also be treated as a form of abuse as any type of under-feeding is, because of the physical impact on the child, the implications for their future health and the psychological impact.” However, many would argue that this kind of situation is not sufficiently serious to justify intervention. This puts social workers in a difficult situation as they do not know in which circumstances it will be appropriate for them to intervene.

Ward LJ also stresses the importance of Article 8 of the European Convention on Human Rights when assessing the significance of the harm, highlighting that Article 8 requires that there must be a ‘”relevant and sufficient” reason for crossing the threshold’. Additionally, if the state is to intervene in a child’s life, the level of state intervention must be proportionate to the risk that the child is suffering. There is a danger that a child who is genuinely suffering will be known to the local authority, but never, quite, be regarded as suffering sufficiently to justify intervention. In Re MA, the local authority found that a girl, who was not the biological daughter of the parents, had been badly treated by them. However, no care order was granted in respect of the parents’ other children as their Lordships found that there was no sufficient evidence of a risk of significant harm to their natural children. The decision in this case is controversial as the parents demonstrated a capacity for cruelty and thus gave rise to a real possibility that they would harm their own children.

In deciding whether the child is suffering from ‘significant harm’, the ‘child’s health or development must be compared with that which could reasonably be expected of a ‘similar child’. There are a number of issues in regards to the ‘similar child’ test. There is particular controversy over the extent to which the cultural background of the child should be taken into account. It is also unclear to which extent the characteristics or capabilities of the parents should be considered.

Reforms and recommendations

For the last 40 years, several reforms have been intended to improve the law on ‘child protection’ and compensate for failures in practice. Many of these reforms responded to the cumulative evidence inspections and high-profile reviews into children’s deaths including: the 1974 Maria Colwell inquiry which led to the Area Review Committees, the 1988 Cleveland inquiry which formed the early versions of the statutory guidance Working Together To Safeguard Children and the Victoria Climbié Report which contributed to the Every Child Matters green paper with recommended policies designed to ensure that it never happened again. Since the individual reforms of the past have all seemed intelligent and well-designed, it seems puzzling that they have not achieved their intended goals. It is submitted that there may have been too many unnecessary targets. Instead of addressing existing practical problems, such as poor system management and inadequate funding, the previous reforms have focussed too much on the process of case management and increasing regulation. This may have impeded the real issue of child protection. The Munro report has provided some interesting recommendations to improve the law on child protection with particular focus on early intervention, the transparency and accountability of the system and the expertise of the social work profession.

In conclusion, it is submitted that there is no consistent theme in the approach of their Lordships in regards to the threshold criteria. There is however increasing evidence to suggest that the thresholds need to be lower. Witnesses from the courts found little or no evidence of inappropriate removal of children and many instances where earlier removal would have been appropriate. This is backed by academic research. Professor Ward noted that “there is substantial evidence that many children remain for too long with or are returned to abusive and neglectful families with insufficient support”.

Word Count: 2500

Bibliography

Primary Sources

Cases

Lancashire CC v B [2000] 1 FCR 509

Re B (Children) (Care Proceedings: Standard of Proof) [2008] UKHL 35

Re C (A Child) [1993] 1 FLR 257

Re D (Care: Threshold Criteria) [1998] Fam Law 656

Re D (A Child) (Care Order: Evidence) [2010] EWCA Civ 1000

Re H and Others (minors) (sexual abuse: standard of proof) [1996] AC 563

Re L (Children) [2006] EWCA Civ 1282

Re M (A Minor) (Care Order: Threshold Conditions) [1994] 2 FLR 577

Re MA (Care Threshold) [2009] EWCA Civ 853

Re O and N (Children) (Non-accidental injury) [2003] 1 FCR 673

Re O (A Minor) (Care Order: Education: Procedure) [1992] 4 All ER 905

Re P (Care Proceedings) [2012] EWCA Civ 401

Re S-B (Children) [2009] UKSC 17

Re T (A Child) (Care Order) [2009] 2 FCR 367

Statutes and statutory instruments

Adoption and Children Act 2002

Children Act 1989

Children Act 2004

Children and Families Act 2014

Secondary Sources

Books

Herring J, Family Law (6^th edition, Pearson Education Ltd, 2013)

Harris-Short S and Miles J, Family Law: Text, Cases and Materials (2^nd edition, Oxford University Press, 2011)

Journal articles

Bainham A, ‘Striking the Balance in Child Protection’ [2009] CLJ 42

Hayes M, ‘Uncertain Evidence and Risk-Taking in Child Protection Cases’ [2004] CLFQ 63

Keating H, ‘Shifting Standards in the House Of Lords’ [1996] CFLQ 157

Lowe N and Cobley C, ‘The statutory “threshold” under Section 31 of the Children Act 1989-time to take a stock’ LQR 396

Masson J, ‘Reforming Care Proceedings- Time for a Review’ [2007] CLFQ 411

Websites

Department for Education, ‘Landmark Children and Families Act 2014 gains royal assent’ (Press release, 13 March 2014) <

Eileen Munro, ‘The Munro Review Of Child Protection’ (2010) <

Grady H, ‘Should Extremely Obese Children Be Taken Into Care?’ (BBC Radio 4, 16 August 2012)<

Jonathan Herring, Family Law (6^th edition, Pearson Education Ltd, 2013) 605.

Children Act 1989, s 1(1).

Jonathan Herring, Family Law (6^th edition, Pearson Education Ltd, 2013) 603.

Re T (A Child) (Care Order) [2009] 2 FCR 367.

S Harris-Short and J Miles, Family Law : Text, Cases And Materials (2^nd edition, Oxford University Press, 2011) 803.

Andrew Bainham, ‘Striking The Balance in Child Protection’ [2009] CLJ 42,45.

Heather Keating, ‘Shifting Standards in The House Of Lords’ [1996] CFLQ 157.

Children Act 1989, s 31(2)(a).

S Harris-Short and J Miles, Family Law : Text, Cases And Materials (2^nd edition, Oxford University Press, 2011) 832.

Re M (A Minor) (Care Order: Threshold Conditions) [1994] 2 FLR 577.

S Harris-Short and J Miles, Family Law : Text, Cases And Materials (2^nd edition, Oxford University Press, 2011)832.

Jonathan Herring, Family Law (6^th edition, Pearson Education Ltd, 2013) 611.

ibid.

Re H and Others (minors) (sexual abuse: standard of proof) [1996] AC 563.

S Harris-Short and J Miles, Family Law : Text, Cases And Materials (2^nd edition, Oxford University Press, 2011)834.

ibid.

S Harris-Short and J Miles, Family Law : Text, Cases And Materials (2^nd edition, Oxford University Press, 2011)834.

Re B (Children)(Care Proceedings: Standard of Proof) [2008] UKHL 35.

Jonathan Herring, Family Law (6^th edition, Pearson Education Ltd, 2013) 612.

Re S-B (Children) [2009] UKSC 17.

Mary Hayes, ‘Uncertain Evidence And Risk Taking In Child Protection Cases’ [2004] CLFQ 63,

ibid.

Re P (Care Proceedings) [2012] EWCA Civ 401

<>accessed 18 January 2015.

Eileen Munro, ‘The Munro Review Of Child Protection’ (2010) > <accessed 18 January 2015.

Mary Hayes, ‘Uncertain Evidence And Risk Taking In Child Protection Cases’ [2004] CLFQ 63,

Lancashire CC v B [2000] 1 FCR 509.

Jonathan Herring, Family Law (6^th edition, Pearson Education Ltd, 2013) 618.

Re O and N (Children) (Non-accidental injury) [2003] 1 FCR 673.

Re C (A Child) [1993] 1 FLR 257

Helen Grady, ‘Should Extremely Obese Children Be Taken Into Care?’ (BBC Radio 4, 16 August 2012) < > accessed 15 January 2015.

MA, SA and HA v MA, HA and City and Council of Swansea [2009] EWCA Civ 853

Re MA (Care Threshold) [2009] EWCA Civ 853.

Children Act 1989, s 31(10).

Re MA (Care Threshold) [2009] EWCA Civ 853

Ibid.