Although this paper deals with the law under the Family Law Act 1975 and under the Children and Young Persons (Care and Protection) Act 1998, my knowledge of the latter is limited, and on that topic this paper will be neither expert nor comprehensive. Although in the 1970s I did some work in child welfare law, since 1993 I have been mainly involved, both as a judge and as an academic, in the operation of the Family Law Act 1975 (Cth). I should also mention that this paper does not deal with that remarkably resilient jurisdiction, the parens patriae jurisdiction still exercised by the Supreme Court of New South Wales.3
My main focus will be the High Court’s decision about ‘unacceptable risk’, and how it plays out in the different jurisdictions of the family law courts (the Family Court of Australia and the Federal Magistrates Court) and the children’s court. But I’d first like to say something about the two jurisdictions.
2. The two jurisdictions similarity and difference
There are obvious similarities in the two jurisdictions. It can broadly be said that in each jurisdiction the child’s interests are to be paramount. Under the New South Wales act, the first principle in s 9 is
In all actions and decisions made under this Act (whether by legal or administrative process) concerning a particular child or young person, the safety, welfare and well-being of the child or young person must be the paramount consideration. In particular, the safety, welfare and well-being of a child or young person who has been removed from his or her parents are paramount over the rights of the parents.
The equivalent section in the Family Law Act 1975 is s 60CA:
In deciding whether to make a particular parenting order in relation to a child, a court must regard the best interests of the child as the paramount consideration.
Next, there are some modifications of the rules of evidence, and other provisions that take the processes of both courts some distance away from the traditional adversary model.4
There are two big differences, however.
The first relates to standing to apply for orders relating to children. Only DOCS can apply for a care order: s 61. By contrast, almost anyone can apply for any order relating to children under the Family Law Act. Section 65C provides that any of the following can apply for a parenting order: a parent, the child himself or herself, a grandparent, or ‘any other person concerned with the care, welfare or development of the child’.
The second difference is even more fundamental. If one asks ‘What does it take to have the court determine the child’s best interests?’ each jurisdiction gives a different answer.
The family law jurisdiction’s answer is: “Nothing, except that someone applies for an order’. Once the application is made, the court engages with the case, applying the paramountcy principle. The court can make a wide range of parenting orders. There is no need for the applicant to establish any threshold requirement.
By contrast, even though it is DOCS that makes the application, the children’s court cannot make a care order unless it makes a crucial finding, namely that the child is in need of care and protection, there being a series of specified grounds for such a finding (s 71). This is a distinct part of the proceedings, called, I understand, the ‘establishment phase’. The Children’s Court may make a care order in relation to a child or young person only if it is satisfied that the child or young person is in need of care and protection for any of the specified reasons, which include:
(c) the child or young person has been, or is likely to be, physically or sexually abused or ill-treated,
(d) subject to subsection (2), the child’s or young person’s basic physical, psychological or educational needs are not being met, or are likely not to be met, by his or her parents or primary care-givers,
(e) the child or young person is suffering or is likely to suffer serious developmental impairment or serious psychological harm as a consequence of the domestic environment in which he or she is living.
Considering the differences about the threshold
Does this difference make sense? I think it does, up to a point. The law gives each parent ‘parental responsibility’, the equivalent of what used to be called ‘guardianship’.5 If they can’t agree, the court can decide. The parties to the proceedings will be the parents, each of whom has been entrusted by the law with responsibility for the child’s care. The court has jurisdiction because we need a circuit-breaker.6
This analysis works nicely where the proceedings are between parents. But it breaks down where the parties are other people. The application might be by a grandparent against one or even both of the parents. It might be by a foster parent. I once had a case where the applicant was a child’s older sister. She argued that her younger sister should be removed from the care of the parents: they had sexually abused her (the older sister) and they were now similarly abusing the younger sister. In the children’s court, such an application would have to be made by DOCS, and the court could act only if the threshold test had been satisfied. But in the family courts, once the application was made, the only question is what is best for the child.
It’s interesting to reflect that the law might have imposed a threshold requirement in such cases, but it didn’t. The big chance to impose one, perhaps, arose in a House of Lords case, J v C (1970).7 The contest was between the child’s parents (who were poor, and Spanish) and the child’s foster parents (who were English, and who had looked after the child for some years at the request of the parents). The House of Lords simply dealt with the substance of the matter, holding, famously, that the paramount consideration principle applied even between parents and non-parents. It held that although the Spanish parents had done nothing wrong, it was now best for the child to remain with the English foster parents. The law in England and Australia has, I think, never departed from that approach, whether one looks at the old parens patriae jurisdiction of the Chancery Division or today’s law under the Family Law Act 1975, and the relevant English legislation.8
The difference in some circumstances is thought-provoking. If a grandparent or another person who has had some care of the child thinks that the child would be better off living with them, that person can apply under the Family Law Act and the court must decide the case according to what it thinks is best for the child. In theory, if the child were found to be even marginally better off in the care of the applicant, the court would have to place the child with the applicant, because of the paramount consideration principle. But if DOCS thinks the child would be better off in someone else’s care, for example with a foster parent, it cannot go to the children’s court on the basis that the child will be better off in foster care – it must establish one of the grounds for a care order.9 Yet if the child had been placed in temporary foster care, the foster parents could apply to a family court, and would only have to show that the child would be better off, even marginally, if placed with them.
In practice, perhaps, this theoretical oddity does not cause real trouble. Foster parents mostly don’t go to the family courts because, I assume, even if they won they would not necessarily be entitled to support from DOCS, and if they were to lose they might be liable for costs. And when the family courts come to decide cases between parents and non-parents, they are likely to take into account that other things being equal children are probably best with their parents, so in practice the non-parent would bear something like an onus of proof. And maybe that onus would not be very different, in practical terms, from the grounds needed for a care order.
3. The interaction of the two systems
exercise of jurisdiction
Both systems are about children: how do they inter-relate? In terms of jurisdiction, the legislation deals with the question in a way that is, I think, reasonably clear.10
Had there been no provisions dealing with this issue, by virtue of s 109 of the Constitution orders made by the federal courts would prevail over any inconsistent orders under the state system. However the Family Law Act (a federal Act) provides, in substance, that the state system will prevail: s 69ZK.11
The first strand of the section is s 69ZK(1), which says that when a child is under the care of a person under a (state) child welfare law, the family law court must not make an order in relation to the child.
There are two qualifications to this restriction. First, the family law court12 can make an order about the child if it is expressed to come into effect when the child ceases to be under such care. Thus the family law court could make an order, for example, that if in the future the child ceases to be under state care, the child is then to live with a certain person. In practice, the family law court would often be reluctant to make an order of that kind, since it would come into operation at an unknown time in the future, and in unknown circumstances. Normally any proceedings of the family law court would be terminated or adjourned when the child is in care under the child welfare laws. The second qualification is that the family law court can make orders in relation to such a child if the (state) child welfare officer has given written consent to the institution or continuation of the proceedings.
The second strand of s 69ZK is to preserve the operation of the state system. Section 69ZK(2) provides:
(2) Nothing in this Act, and no decree under this Act, affects:
(a) the jurisdiction of a court, or the power of an authority, under a child welfare law to make an order, or to take any other action, by which a child is placed under the care (however described) of a person under a child welfare law; or
(b) any such order made or action taken; or
(c) the operation of a child welfare law in relation to a child.
These two strands ensure that the operation of the federal system (the Family Law Act) cannot interfere with the operation of the state child welfare system. To put it another way, in the event of any overlapping or inconsistency, the state system prevails over the federal system. Although there were some technical problems with earlier versions of the Act, it is my impression that there is now no ambiguity or uncertainty about this matter.
Examples of the operation of the provision:
After family law court parenting orders placing a child with X, DOCS brings proceedings in a Children’s Court and obtains orders that the child should be removed from X and placed in the care of Y. The Children’s Court order would prevail: s 69ZK(2).
A family law court has ordered that no further medical examinations should be made of a child without the court’s permission. DOCS makes arrangements, valid under the child welfare law, for a further medical examination. Despite being inconsistent with the family law court order, the medical examination can be carried out, because the child welfare law prevails: s 69ZK(2).
While proceedings in a family law court are pending, DOCS obtains Children’s Court orders placing the child in care. Unless the Minister consents in writing to the family law court proceedings, the family law court cannot now make any orders relating to the child (even by consent of the parties) except orders that would come into force only when the child leaves care: s 69K(1).
DOCS intervenes and has a child removed from the mother and placed with the mother’s sister. Later DOCS identifies the father as a viable carer, and with his consent withdraws its Children’s Court proceedings. The father starts Family Court proceedings and is granted interim residence. Some months later, however, by parental agreement, the mother is granted residence and the child is returned to her. It is open to DOCS to take action under the Care and Protection Act to remove the child.13
A family law court makes a recovery order14 in relation to a child. Then, before the order is acted upon, the child comes under the state child welfare system. The recovery order remains in force until the family law court revokes it; but it cannot have any effect that is inconsistent with the operation of the child welfare law: s 69ZK(2). If, for example, the child is placed in the care of a foster parent under the child welfare law, the recovery order cannot be put into effect so as to remove the child from the care of the foster parent.
DOCS intervenes in a proceeding in a family law court, and orders are made that the child should live with X. DOCS later has reason to believe that the child is at risk of abuse from X. Even though it is a party to the family law court proceedings, it seems that DOCS could use its authority under the state law to remove the child from X, notwithstanding the family law court order: s 69ZK(2).
It should be emphasised that these are hypothetical examples, designed to illustrate the technical legal operation of the provisions. In such cases in practice, DOCS would no doubt have regard to other factors, including the protocols with the Family Court of Australia and Federal Magistrates Court.
The interaction of the two systems: some practical problems
Although I think the legislation deals satisfactorily with the exercise of jurisdiction, there seem to be persistent problems in the way the system operates. This is not my main topic today, and discussing it would take us to Magellan, protocols, resources, confidentiality and information-sharing, and much else. But I want to refer briefly to the view of the matter from the family courts’ vantage point.
From what I have heard, those working in the family courts tend to be frustrated by DOCS’ unwillingness to intervene in some cases that seem to involve children at serious risk. There are two aspects to this. I hear repeated comments to the effect that DOCS often declines to intervene when asked by the family courts to do so. Typically, such a request is made when material before the family law court indicates that there are serious doubts about the capacity of either party in the proceedings to give appropriate care to the child. In one such recent case, Benjamin J held that the Family Court had power to order the relevant State Department to intervene, and to make the Department subject to its orders,15 although the decision was overruled on appeal, the Full Court holding that the Family Court has no such power.16 The case is a vivid example of a problem that has long troubled the family courts, seen from the family courts’ perspective. The seriousness of these cases is such that those in the family courts wonder how it can be that the child protection agencies do not intervene in some cases.
One view, that I have often heard, is that those in the state and territory child protection agencies must assume that the child is not in danger if the case is before that family courts. I doubt that this could be right, because such an assumption is clearly untenable. The task of the family courts is to resolve disputes. Although they have the advantage of independent children’s lawyers in many of the more difficult cases, they do not have an investigative capacity, and in substance can only rule on applications made to them. Alternatives such as foster care or institutional case are not available.
Another view is that awful though these family law court cases are, where DOCS does not intervene, this is because the cases still fall short of the level of awfulness that DOCS requires – no doubt because of its limited resources – before it can justify intervening. The Children Legislation Amendment (Wood Inquiry Recommendations) Act 2009 amended a number of sections, including s 30, by substituting ‘at risk of significant harm’ for the existing ‘at risk of harm’. If that amendment raises the threshold of awfulness before DOCS takes action, as it appears to do, the difference between the two jurisdictions might thereby be accentuated.
In considering these issues, it is important to consider in each case just what would be achieved by DOCS intervening.17 Without intervening, DOCS can assist the family law courts greatly by providing existing reports and other documents that might be subpoenaed, and, perhaps, co-operating with the Independent Children’s Lawyer. Those in the family law courts who wish that DOCS would intervene more often need to identify just what intervention would achieve that could not be achieved in other ways. In some cases the answer might be that if DOCS intervenes, the court might be in a position to consider possibilities of placing the child with persons other than the parties to the litigation. In others, perhaps, further investigation and report by DOCS might assist the court. There are interesting and tricky questions here, including whether it might sometimes be preferable for DOCS to commence proceedings in the children’s court rather than intervene in the proceedings in the family law court. But to this is a famously difficult topic, and would require consideration of protocols, resources and practices – matters on which I am currently ill-equipped to speak, and, matters, happily, beyond the scope of this paper.
A radical solution?
I take it everyone here will be aware of the fact that for many years people have considered more radical approaches to the problem, such as a cross-vesting scheme for family law and care and protection matters,18 or a Commonwealth Child Protection Service,19 or having just one court, with comprehensive jurisdiction over what is now called ‘parenting law’ as well as the child protection law now administered by the children’s courts. I do not propose to explore these issues, however. I don’t think I was asked to do so, and a more practical focus on the present law is no doubt more relevant to children’s court magistrates.
4. The High Court decision in M v M: introduction
M v M (1988)20 is the key High Court decision on child sexual abuse,21 and it remains the most authoritative statement of the law.
The mother had suspicions that the father had sexually abused the parties’ 4 year old daughter, stemming from things the child said to her. She took the child for examination, but it was inconclusive, and the child then made no disclosures. The child did later make disclosures to a police officer, albeit after leading questioning by the mother and the officer. The child also implicated the father in sexual abuse in later interviews with a child psychologist, who believed she had been abused, although not necessarily by the father. The mother, it seems, genuinely believed that the allegations were true. The father denied them.
The trial judge said he was unable to determine whether or not the father had abused the child, but thought there was a possibility that the child had been abused by the father and terminated access by the father. His Honour ruled out supervised access as not having any benefit for the child, but that aspect was not the subject of the appeal.
The majority of the Full Court dismissed the father’s appeal, although Nicholson CJ, dissenting, held that access should not be refused because of a mere possibility that access would expose the child to a risk of sexual abuse – there must be ‘a real or substantial risk of such abuse occurring as a matter of practical reality’.
The High Court dismissed the father’s appeal in a single unanimous judgment. Although the trial judge had used the term ‘lingering doubts’, the High Court agreed with the majority of the Full Court that on a correct reading of his judgment, he had found that ‘as a matter of practical reality’, access would have entailed a risk that the child might be sexually abused and her welfare endangered. In those circumstances, the High Court, like the majority of the Full Court, upheld the trial judgment.
The High Court thought that the difference of opinion in the Full Court might have reflected different views about what the judge had decided. In dissent, Nicholson CJ had formulated a test that included the words ‘as a matter of practical reality’, and so, on the High Court’s interpretation of the trial judgment, it would have passed the Nicholson test. This is important, because it emphasises the important point, to which I will return, that the High Court did not say that any possible risk is enough.
The father’s submission and why it was rejected
To understand the decision it is important to identify the argument that was addressed to the High Court. The appellant father argued in the High Court that in such cases the court must first determine, on the balance of probabilities, whether the parent has abused the child. If the complainant fails on that issue, that is the end of the matter because the rejection of the complaint necessarily entails a negative answer to the second issue, which is whether there is a risk of sexual abuse occurring if the access order is made.
It is quite clear, in my view, that the father’s submission was untenable. Evidence that a person has previously abused the child would indeed strongly suggest that the child would be at risk if placed in the care of the person. But it is not necessary for a finding of risk. There could well be evidence of a risk of child sexual abuse other than evidence that the accused person had previously abused the child. The person could have abused other children, the house could be full of child pornography, the person could have been grooming the child, even boasting about his intentions to abuse the child, and so on. This sort of evidence could easily lead to a finding that the child would be at risk of abuse in the care of the person. And it would make no difference, of course, if it had also been alleged that the person had actually abused the child in question, but the court was unable to make that finding: the other evidence could still lead to a finding that the child was at risk of abuse.
Less obviously, although a finding that the person had previously abused the child would strongly suggest a future risk, we could imagine other evidence that would lead the court not to make the finding of risk. Suppose, for example, that there was good evidence that the previous abuse had occurred when the person was mentally ill, and the illness was now cured or controlled by medication. More obviously, if the proposed order was for supervised contact, the court might well be persuaded that it would involve no risk to the child of sexual abuse. (Of course, whether the supervised contact would benefit the child is a different question.)
Thus, the father’s submission in M v M could not possibly succeed. A finding that the person had previously abused the child is not a necessary condition for a finding that orders placing the child with that person would expose the child to risk of abuse. Nor is it a sufficient condition for such a finding of risk, because there could be circumstances that lead the court to conclude that even though the person had previously abused the child, the proposed orders would not expose the child to risk.
It’s not surprising, then, that the High Court rejected the submission, essentially for these reasons. Here is what it said:
The fact that the proceedings involve an allegation that the child has been sexually abused by the parent who seeks custody or access does not alter the paramount and ultimate issue which the court has to determine, though the court’s findings on the disputed allegation of sexual abuse will naturally have an important, perhaps a decisive, impact on the resolution of that issue…. the resolution of an allegation of sexual abuse against a parent is subservient and ancillary to the court’s determination of what is in the best interests of the child.
As this passage shows, there are two separate questions about which findings might be made: (1) whether the person seeking to have contact with the child previously abused the child and (2) whether the proposed contact orders expose the child to such a risk of abuse that they should not be made. We will take them in turn. In relation to the first, the Court discussed when such a finding could be made, and when it might be appropriate to make such a finding. In relation to the second, the Court laid down the ‘unacceptable risk’ principle.
Findings that a person has sexually abused the child: when can they be made?
As to the determination of sexual abuse allegations, there were no surprises: those allegations were to be determined ‘according to the civil standard of proof, with due regard to the factors mentioned in Briginshaw’.22 This basically remains the position, although as we will see it is now necessary to bring the test up to date by reference to later cases and to the Evidence Act 1995.
Findings that a person has sexually abused the child: when should they be made?
The Court made two comments on this question. First, it acknowledged that the evidence will often fall short of what is required to make such a finding:
No doubt there will be some cases in which the court is able to come to a positive finding that the allegation is well-founded….there will be very many cases, such as the present case, in which the court cannot confidently make a finding that sexual abuse has taken place.
The Court also commented on whether, in circumstances where the evidence would justify a finding that a person has abused the child, the court should, in fact, make that finding:
And there are strong practical family reasons why the court should refrain from making a positive finding that sexual abuse has actually taken place unless it is impelled by the particular circumstances of the case to do so.
Making parenting orders: the unacceptable risk test
The High Court then went on to say something about the assessment of risk of sexual abuse, and laid down the famous test of ‘unacceptable risk’, which relates to the making of what are now called parenting orders. The key passage starts:
In resolving the wider issue the court must determine whether on the evidence there is a risk of sexual abuse occurring if custody or access be granted and assessing the magnitude of that risk. After all, in deciding what is in the best interests of a child, the Family Court is frequently called upon to assess and evaluate the likelihood or possibility of events or occurrences which, if they come about, will have a detrimental impact on the child’s welfare. The existence and magnitude of the risk of sexual abuse, as with other risks of harm to the welfare of a child, is a fundamental matter to be taken into account in deciding issues of custody and access…
The Court then referred to various earlier attempts to formulate a helpful test, such as ‘risk of serious harm’, or ‘an appreciable risk’, but said those formulas were ‘striving for a greater degree of definition than the subject is capable of yielding’. The Court concluded:
In devising these tests the courts have endeavoured, in their efforts to protect the child’s paramount interests, to achieve a balance between the risk of detriment to the child from sexual abuse and the possibility of benefit to the child from parental access. To achieve a proper balance, the test is best expressed by saying that a court will not grant custody or access to a parent if that custody or access would expose the child to an unacceptable risk of sexual abuse.
The principles stated
It is not difficult to identify the principles emerging from M v M. In a valuable article published shortly after the decision,23 the former Chief Justice of the Family Court, Alistair Nicholson, expressed them accurately and succinctly:
(a) It is [not] necessary to make a positive finding of child abuse and the Court should avoid doing so except in the most obvious cases.24
(b) If such a finding is made, the standard of proof to be apply is that provided in Briginshaw v Briginshaw (1938) 60 CLR 336.
(c) In resolving the issue as to what form of order is in the best interests of the child, the court must determine whether on the evidence there is a risk of abuse occuring if custody or access be granted and assessing the magnitude of that risk.
(d) If the risk is assessed to be unacceptable, then custody or access should not be granted.
5. The problematical ‘Re W’ decisions
Before leaving this general topic I should say something briefly about what I consider a troubling line of cases, the ‘Re W’ decisions.25 Both John Fogarty and I have been critical of some of the reasoning in these cases. To put the matter very briefly, those decisions give a great deal of emphasis to the benefit to a child from continuing contact with both parents, and stress that courts should be very reluctant to make findings that a child has been sexually abused. I have no problem with either of these propositions, but my argument, and I think John Fogarty’s, is that some of the language in the cases is so strong on one side that it distorts the balancing process that the legislation and the High Court decision require. I can briefly indicate what I see as the problem with these decisions.
The language from the Re W decisions includes the following:
In children’s matters under Pt VII of the Family Law Act, where the issue is a child’s contact or residence with a significant person in his or her life, the grave consequences of a finding of sexual abuse cannot be overstated.26
The risk that the court will find heinous behaviour when none has occurred needs to be born in mind at all times. The harm and injustice that flows to both parent and child from an erroneous positive is almost too horrible to contemplate.27
The problem with such passages is, as John Fogarty put it:28
The gravity of such consequences has always been acknowledged. But to start with an entrenched view that the consequences ‘cannot be overstated’ or are ‘too horrible to contemplate’ warps the process explained by the High Court.
The NZ Court of Appeal has also stressed the need for a balanced approach:29
the difficulty of reaching a firm conclusion as to whether sexual abuse has or has not occurred and the necessity to proceed with great care and caution before either finding that such allegations have been established or finding that they are without foundation…
In the Family Court, Justice Sally Brown has made the same point:30
With respect to the Full Court, one might as well say that the harm and injustice that flows to a child from an erroneous negative finding is almost too horrible to contemplate, that harm including repeated sexual abuse of the child. Nevertheless, I am bound by the exposition of principle in the judgment.
Brown J also convincingly showed that the three decisions rely on a flawed summary of a particular journal article in the United States that has, in my view, limited relevance to the courts’ approach to child abuse cases.31 In my view it is unsafe to rely on some of the statements in these cases.32
There is a considerable body of case law since M v M. In my view the steps the court needs to take, and the principles involved, can be simply stated. In sexual abuse cases the court should first consider whether it is open to make a finding that a person has abused the child, applying the civil standard of proof having regard to s 140. If it is, the court should then consider whether it should do so. It may properly refrain from making the finding when it is otherwise able to justify the proposed parenting orders, for example by making a finding that exposure of the child to the person would expose the child to unacceptable risk, or for other reasons would not be in the child’s interests. In relation both to the application of the s 140 test and the choice of making or not making an available finding, the court should carefully and systematically consider the consequences of the decision for the child, the parents, and others involved. The passages in the Re W decisions that emphasise the dangers of making false positive findings of abuse should not be allowed to distract judges from carefully weighing up all the factors involved.
6. Conclusions - Back to basics
In the end, the most helpful and authoritative summary of the present law seems to be the decision in Johnson v Page,33 in which the Full Court specifically endorsed34 the following seven-point summary by John Fogarty:
1 The decisive issue is and always remains the best interests of that child. All other issues are subservient.
2 The nature of the risk is best expressed by the term ‘unacceptable risk’. It is an evaluation of the nature and degree of the risk and whether, with or without safeguards, it is acceptable.
3 Where past abuse of a child is alleged it is usually neither necessary nor desirable to reach a definitive conclusion on that issue. Where, however, that is done the Briginshaw civil standard of proof applies.
4 The circumstance, if it be so, that the allegation of past abuse is not proved in accordance with Briginshaw, does not impede reliance upon those circumstances in determining whether there is an unacceptable risk.
5 The concentration in these cases should normally be upon the question whether there is an unacceptable risk to the child.
6 The onus of proof in reaching that conclusion is the ordinary civil standard.
7 But the components which go to make up that conclusion need not each be established on the balance of probabilities. The court may reach a conclusion of unacceptable risk from the accumulation of factors, none or some only of which, are proved to that standard.
The Full Court added, in relation to the last paragraph:
We assume point seven of that summary is directed to the requisite standard of proof. We think a Judge may be cautious in coming to a finding of unacceptable risk if none, rather than some only, of the accumulation of factors considered, satisfy the standard of proof (but see Malec v J C Hutton Proprietary Limited (1990) 169 CLR 638).
The point here, I think, is that unacceptable risk is a prediction of the future, based on findings of fact. The confidence one will have in the prediction will be, in part, a reflection of the confidence one has in the factual findings that base the prediction. The decision cited, Malec, contains a short but interesting discussion of the approach to making predictions (in that case, relating to the future impact of injury in a damages action) as distinct making findings about past facts.
In relation to the paragraph 4, as the Full Court pointed out in Johnson, the position relating to the burden of proof for findings that a person has abused a child has been slightly refined, the early reference to Briginshaw now being a reference to s 140 of the Evidence Act 1995.
The principles as stated in Johnson provide a more balanced and more authoritative guide than the language in the Re W cases to the effect that the consequences of a false positive finding ‘cannot be overstated’, and I suggest that such statements cannot now be taken as representing the law. We don’t yet have a substantial Full Court discussion of the way the court might weigh up various considerations in deciding whether to make a finding of abuse in circumstances where that is open on the evidence. However I have suggested that in considering this, the court would be entitled to look at all aspects, especially those relevant to the child’s best interests – and in this respect a continuing involvement of the other parent, while of course very important, would not be the only factor to be considered. Although the amendments of 2006 gave renewed emphasis to the benefits to children of a relationship with both parents, there is no reason to think that the need to protect children is any less than it ever was, as Brown J has pointed out.35
Ultimately the court must try to work out what is in the best interests of the particular child, and in cases where there are allegations of child abuse, the challenging tasks include not only what parenting orders to make but whether to make positive or negative findings about the abuse allegations. The decision in Johnson may be seen as answering John Fogarty’s call for a ‘return to basics’. It does not make these decisions any easier – they are inevitably full of difficulty and anguish – but it does help guide practitioners and trial judges to what the High Court surely intended, namely a careful and balanced assessment of all the options, and all the relevant factors, when preparing judgments and making orders about the best interests of children.
7. Unacceptable risk and the Children and Young Persons (Care and Protection) Act 1998
Finally, how does all this fit in with the Children and Young Persons (Care and Protection) Act 1998 (NSW)?
First, as pointed out earlier, under the Family Law Act there is no equivalent to the threshold test under the NSW Act. Thus the ‘unacceptable risk’ test is applied when the court is considering what orders are most likely to be in the child’s interests. On the face of it, then, the test is relevant to the dispositional phase of care cases, not the establishment phase.
Is this correct? I think it is. The relevant language of the establishment phase is that the court may make a care order if the child is in need of care and protection for any of the following reasons: (c) the child or young person has been, or is likely to be, physically or sexually abused or ill-treated, and (d) the child’s or young person’s basic physical, psychological or educational needs are not being met, or are likely not to be met, by his or her parents or primary care-givers. I have not had a chance to check any case law on this, but this language does not seem to me to invite application of the ‘unacceptable risk’ test.
Is the test applicable when it comes to the dispositional stage? The Act spells out the objects of the Act (s 8), the principles to be applied (s 9) and the principle of participation (s 10). Those principles surely apply to the exercise of jurisdiction by the children’s court, as well as to other aspects of the legislation.36 Then we have s 71, which says that the children’s court may make a care order when a child is in need of care and protection for one or more of the reasons specified – the grounds. The Act then sets out the various orders that can be made when a child has been found to be in need of care and protection: ss 73-77 and 79, 81, 82, 88. There are also provisions relating to care plans (ss 78, 80), permanency planning (s 78A), and other matters. It seems to me that, broadly speaking at least, none of these sections qualify the application of the principles, most relevantly for us, the principles in s 9.
The question, then, seems to be whether the ‘unacceptable risk’ test applies when the court is considering what orders to make, and applying the principles in section 9.
I’m sure the objects and principles are familiar, but it’s necessary to set them out here, although I will abbreviate those that are of little relevance to the present topic. To some extent, there are similar provisions in the Family Law Act 1975, and I will indicate them in footnotes. Broadly speaking, I think that the provisions that are not echoed in the Family Law Act 1975 have to do with the nature of child protection law, involving external intervention and possible placement outside the family.
Section 8, Objects
… to provide:
(a) that children and young persons receive such care and protection as is necessary for their safety, welfare and well-being, taking into account the rights, powers and duties of their parents or other persons responsible for them,37 and
(b) [institutions, services and facilities to provide safe environment and foster children’s developmental needs] and
(c) [assistance to people responsible for children].
Section 9 Principles
The principles to be applied in the administration of this Act are as follows:
(a) In all actions and decisions made under this Act (whether by legal or administrative process) concerning a particular child or young person, the safety, welfare and well-being of the child or young person must be the paramount consideration.38 In particular, the safety, welfare and well-being of a child or young person who has been removed from his or her parents are paramount over the rights of the parents.
(b) [Children’s views to be heard and given weight.]39
(c) [Culture and other characteristics of the child to be considered.]40
(d) [Intrusive intervention in the life of the child should be the least that is ‘consistent with the paramount concern to protect the child or young person from harm and promote the child’s or young person’s development’.]
(e) [Child without family entitled to special protection and assistance from the State, and name, identity etc to be preserved.]
(f) [Out of home care – safe, and early decisions to be made.]
(g) [Out-of-home care: child entitled to a safe, nurturing, stable and secure environment; unless contrary to best interests, will include the retention of family etc relationships.]
Is the statement by the High court in M v M applicable under this legislation? I think it does, although for the reasons already given it does little more than pose the question the children’s court must address. In cases where there is a risk that a particular order will expose a child to risk of abuse (or other risks, for that matter), the test asks the court to consider whether that risk is ‘unacceptable’. Doing so involves weighing up all the matters relevant to the assessment of what will be best for the child, having regard to the s 8 objects and s 9 principles, and the various constraints and guides to be found in the other provisions. In some cases, it may require the court to assess whether the benefits of contact with a parent outweigh the risk of abuse. In other situations, I assume, the risk of abuse must be considered along with other possible benefits and detriments that might be entailed by the various possible orders that the court has to consider.
Although the test applies, then, it doesn’t take us very far. As I hope to have demonstrated, this is because the test was formulated by the High Court in the course of rejecting a foolish submission, namely that a positive finding that a parent had abused the child is a necessary and sufficient condition for ordering that the child should have no contact with the parent. The significance of the test in the context of the Family Law Act is mainly that it re-stated the obvious, namely that the court must try to work out what is best for the child, weighing up all the factors involved. Stating this obvious point was also significant because it rejected the idea that some more specific guideline should be applied, for example a guideline that the risk must have some particular quality before it can be taken into account.
This leads to a somewhat surprising conclusion, namely that although the ‘unacceptable risk’ test applies to the dispositional phase in the children’s court, its real function, in the children’s court as well as the family courts, is to tell us not to be distracted, and to get on with the often agonisingly difficult task of weighing up all the relevant matters when determining what orders are most likely to be best for the child.
I know some people have been puzzled by M v M. But the answer to the mystery of M v M is that there is no mystery. The High Court could perhaps have formulated some particular approach, or indicated that some factors are of particular importance. Wisely, in my view, it declined to do so. Its message for judicial officers determining child abuse cases is: get on with it.
8. Postscript: Sophie’s case
An instructive example of how the law can be misunderstood is the decision of the NSW Court of Appeal in Sophie’s Case in 2008.41
Facts: 5 year old Sophie was diagnosed with a sexually transmitted disease (‘NG’), that the father had contracted from a prostitute, and DOCS made a care application based on the likelihood that the father was the source of her infection. The children’s court made a care order; the District Court allowed the appeal; the Court of Appeal allowed DOCS’ appeal and ordered a rehearing.42
At the children’s court:
The Magistrate relied on the observations of the High Court in M v M (1988) 166 CLR 697 at 76–77 that the resolution of an allegation of sexual abuse against a parent is subservient and ancillary to the court’s determination of what is in the best interests of the child and that, accordingly, the court must assess the risk of sexual abuse occurring even when it is unable “confidently to make a finding that sexual abuse has taken place”. The Magistrate understood M v M to mean that the Court’s inability to resolve an allegation of sexual abuse should not distract it from assessing both the child’s welfare and the risks involved in competing proposals for the child’s custody or care.43
In my view this is correct. The establishment phase is based on s 71, that ‘the child or young person has been, or is likely to be… sexually abused…’ In the disposition phase, the magistrate appeared to apply M v M impeccably.
On this basis the Magistrate, although not making any finding that the father had sexually abused Sophie, concluded that contact with the father posed an unacceptable risk “in terms of her sexual safety and of her emotional wellbeing”. The Magistrate ordered that Sophie be placed in the parental responsibility of the Minister until the age of 18…
Again, in my view the first sentence is an orthodox application of the law.
Things seemed to go awry in the District Court appeal. As Sackville AJJA rather tactfully put it,
For reasons that are not clear, the proceedings in the District Court appear to have been conducted on a different basis than the care application in the Children’s Court. The sole factual issue addressed by the primary Judge in his judgment was whether Sophie had contracted NG as a consequence of sexual misconduct by the father. The answer to this factual question seems to have been regarded by the parties as determinative of the District Court appeal. The primary Judge, presumably because he was not invited to do so, did not consider by reference to M v M whether Sophie should be regarded as “likely to be … sexually abused” for the purposes of s 71(1)(c) of the Care Act, even if the Court was unable to determine whether the father had in fact sexually abused Sophie.
The District Court judge had concluded:44
The court, on the evidence cannot say with certainty that sexual interference took place, nor can it say that it did not. That the disease may have been contracted in circumstances that are neither unreal or fanciful, despite being highly improbable, does not mean that in this particular case, having regard to the evidentiary test, the case for intervention has been made out.
Applying those standards, in my view, the appeals should be upheld and the orders of the Children’s Court quashed.
There are two things wrong with this. The first one led the Court of Appeal to allow the appeal. It is that the wrong standard of proof was applied, namely the test of ‘certainty’. That is not the Briginshaw test, even on its most fearsome interpretation.45
The second thing that is wrong was not developed by the Court of Appeal (because of the way the District Court proceedings had been conducted), but in my view is clear. The District Court judge did not ask the question required by the application of the paramount consideration test, and by M v M, namely whether refusing to make protective orders would expose the child to an unacceptable risk of sexual abuse. The children’s court magistrate had correctly addressed that question, but the District Court - and, apparently, those who argued the case - wrongly overlooked it.
1 This article derives from a paper prepared for the Children’s Court Conference, Parramatta, 1 September 2010.
2 BA, LLB, BCL; AM; Honorary Professor of Law, University of Sydney; Visiting Fellow, ANU College of Law; former Judge of the Family Court of Australia.
3 As noted in PVS v Chief Executive Officer, Department for Child Protection  WASC 172 (Supreme Court of WA, Murray J), ‘the Supreme Court of New South Wales has held that the parens patriae jurisdiction will only be exercised in extraordinary circumstances, requiring the intervention of the court, although the matter is properly before the Children’s Court’, citing Re Georgia and Luke (No 2)  NSWSC 1387.
4 See Family Law Act 1975 Part VII, Div 12A. Details about ‘less adversarial’ trials can be found on the Family Court’s website: go to ‘about the court’, then ‘court initiatives’.
5 Family Law Act 1975 s 61C.
6 Similarly, where two trustees differ, either of them can seek a direction from the court.
7 J v C  AC 668.
8 By contrast, I understand that in the United States there are often provisions, based on a notion of family privacy that privileges the nuclear family, that prevent others, including grandparents, from applying for orders relating to the children, at least without overcoming a procedural threshold such as the need to seek leave.
9 An eloquent explanation of this is contained in the opening paragraphs of the speech of Lord Hale in B (Children), Re  1 AC 11,  UKHL 35,  3 WLR 1,  Fam Law 837,  2 FLR 141,  2 FCR 339,  Fam Law 619,  4 All ER 1;  UKHL 35 (11 June 2008) URL: http://www.bailii.org/uk/cases/UKHL/2008/35.html .
10 I wrote a detailed account of this as honorary consultant for the Wood Special Commission of Inquiry into Child Protection, and I’d be happy to make it available to anyone interested.
11 There are reported decisions dealing with the effects of earlier versions of this provision, but the current wording resolves the issues that arose in those cases and it is unnecessary to discuss them.
12 I use this term to include the Family Court of Australia, the Federal Magistrates Court, and the Family Court of Western Australia, all of which exercise jurisdiction under the Family Law Act 1975. State and territory magistrates courts do as well, but only to a limited extent.
13 This example is based on a case study in which, however, the (Victorian) department did not subsequently intervene: Family Law Council, Family Law and Child Protection: Final Report (September 2002) Case Study 1, p 39 (available on the Council’s website). The Family Law Council Report suggests that there is a problem in such cases of the department withdrawing once there has been Family Court involvement. If so, that is not a problem of lack of legal power to do so.
14 That is, an order requiring a person to return the child, and authorising the police or others to assist if necessary.
15 Ray and Males  Fam CA 219 (Benjamin J, 31 March 2009).
16 Secretary of the Department of Health and Human Services & Rollinson & Cheeseman & Ors  FamCAFC 258.
17 I am grateful to Rod Best, Director of Legal Services in the NSW Department of Community Services, for emphasising this point.
18 ALRC Seen and heard: priority for children in the legal process (Report 84, 1997).
19 Family Law Council, Family Law and Child Protection - Final Report (September 2002).
20 M v M (1988) 166 CLR 69; 12 Fam LR 612; FLC 91-979; 82 ALR 577.
21 The similar appeal heard at the same time, B v B (1988) 12 Fam LR 612, does not require comment.
22 Briginshaw v Briginshaw (1938) 60 CLR 336, at 362.
23 The Hon Chief Justice Alastair Nicholson, “Child sexual abuse - problems in family law” (1989) 4 Australian Family Lawyer, 1-5, at p 3.
24 Nicholson actually wrote “It is no longer necessary…”. I have modified the phrasing only to simply the matter by removing the implication that under the previous law it had been necessary to make such a finding. I don’t think that is so, but the point is not worth pursuing here.
25 K v B (1994) 17 Fam LR 722, WK v SR (1997) 22 Fam LR 592; (1997) FLC 92-787, and Re W (Sex Abuse: Standard of Proof (2004) 32 Fam LR 249; FLC 93-192.
26 WK v SR (1997) 22 Fam LR 592; (1997) FLC 92-787 at paragraph 47.
27 Re W (Sex Abuse: Standard of Proof (2004) 32 Fam LR 249; FLC 93-192 at paragraph 21.
28 John Fogarty AM, ‘Unacceptable risk - A return to basics’ (2006) 20 Australian Journal of Family Law 249,
29 S v S  NZFLR 657.
30 McCoy v Wessex (2007) 38 Fam LR 513, paragraph 33.
31 Hemiro & Sinla  FamCA 181.
32 A detailed discussion of the significant cases is attempted in R Chisholm ‘Child Abuse Allegations in Family Law Cases: a Review of the Law’ (2011) 25 (1) Australian Journal of Family Law, 1-32.
33 Johnson and Page  Fam CA 1235; (2007) FLC ¶93-344 (May, Boland and Stevenson JJ).
34 Johnson, paragraph 68.
35 McCoy v Wessex (2007) 38 Fam LR 513. Interestingly, the protection of children from violence and abuse would be given ‘greater weight’ under the Family Law Amendment (Family Violence) Bill 2010, released for consultation in November 2010: see item 17, adding new (2A) to s 60CC.
36 They are ‘principles to be applied in the administration of this Act’: s 9.
37 The Full Court attempted to summarise the gist of the the elaborate provisions of the Act since its amendment in 2006 in this way: “…there is a legislative intent evinced in favour of substantial involvement of both parents in their children’s lives, both as to parental responsibility and as to time spent with the children, subject to the need to protect children from harm, from abuse and family violence and provided it is in their best interests and reasonably practicable (Goode v Goode (2006) 36 Fam LR 422; (2006) FLC 93-286, paragraph 72).
38 Similar to Family Law Act 1975 s 60CA.
39 Similar to Family Law Act 1975 s 60CC(3)(a).
40 Similar to Family Law Act 1975 s 60CC(3)(g).
41 Director General of Department of Community Services; Re Sophie  NSWCA 250 (Sackville AJJA, Giles JA, and Handley. AJJA agreeing.)
42 Technical issues about the nature of the appeal are not relevant to the issues in this paper.
43 Sophie’s case, above, paragraph 18.
44 Quoted in paragraph 41.
45 Paragraph 67 (‘It was not appropriate to find that the Director-General had failed to satisfy the burden of proof on the balance of probabilities simply because his Honour could not exclude a hypothesis that, although consistent with innocence, was “highly improbable”. To approach the fact-finding task in that way was to apply a standard of proof higher than the balance of probabilities, even taking into account the gravity of the allegation made against the father.’)