The Charter and risk assessment

A second case on the interaction of two major Victorian statutes of 2006, the Charter and the Disability Act, is now online. Both involve supervised treatment orders under the latter statute:

193(3) A supervised treatment order must- (a) state that the Authorised Program Officer is responsible for the implementation of the supervised treatment order; (b) require the person to whom the supervised treatment order applies to reside in premises approved by the Authorised Program Officer; (c) refer to the treatment plan which must be attached to the supervised treatment order; (d) specify the period for which the supervised treatment order is to continue in force, being a period not exceeding 1 year.

The first case, MM (Guardianship) [2008] VCAT 1282, blogged about here, involved a narrow and unusual question: whether or not supervised treatment orders should be made in relation to someone who wants to be treated but, due to an intellectual disability, lacks the capacity to fully consent. Disappointingly, VCAT Deputy John Billings opted for a broad reasonable limits analysis – which, of course, the detention regime passed with flying colours – without applying the interpretation mandate to the specific provision in dispute. The new case, LM (Guardianship) [2008] VCAT 2084, looks at a much broader question about the limits of the detention regime and does a better, but still inadequate, job.

As always, the facts are heartbreaking. Following childhood behavioural problems, LM was diagnosed at the age of 13 with a ‘borderline to mild intellectual disability’ and a plethora of mental disorders, as well as non-epileptic seizures. As an adult, she attracted a criminal record, including for threatening a woman and a child in a McDonald’s toilet (in 2004) and, more recently, walking into traffic, carrying a controlled weapon and offensive public behaviour. She is presently on a good behaviour bond. Within various institutions, her behaviour included secreting knives and walking onto roads, both apparently with intent to suicide; aggression and threats towards staff; and repeated seizures. But there have been considerable improvements in her current location. Nevertheless, her current disability service provider considers it necessarty to lock the front door to that institution about 70% of the time (apparently so that she feels safe); to forcefully return her to the premises on a number of occasions when she climbed the back fence and headed for the road; to restrain her during seizures; and to engage the police to return her to the premises. They obtained an interim supervised treatment order to authorise these measures and now seek a non-interim order.

There’s little doubt that LM is unwell and poses some danger to herself. However, for better or for worse, treatment of those problems depends on other regimes, including other provisions of the Disability Act, the Mental Heath Act and the Guardianship and Administration Act. The supervised treatment order regime,  the sole regime permitting disability service providers to ‘detain’ anyone, is, by contrast, aimed at protecting others. No-one disputes that LM satisfies the threshold eligibility requirements for STOs: she has an intellectual disability, is in residential care and is being treated. But does she meet the core test of being a risk to others?:

191(6) VCAT can only make a supervised treatment order if VCAT is satisfied that- (a) the person has previously exhibited a pattern of violent or dangerous behaviour causing serious harm to another person or exposing another person to a significant risk of serious harm

What is ‘serious’ harm? The Disability Act doesn’t define the term, so VCAT Member Julie Grainger looked to definitions in the Cth and ACT Criminal Codes (defining serious harm as either life-threatening or longstanding) and the Migration Act (with a broader definition all sorts of potential hams.) She strangely didn’t consider the definition in in Victoria’s own Crimes Act – probably because it refers to ‘serious injury’, thus avoiding an Austlii search –  but it’s not a very helpful definition.

After noting that there’s a much stronger analogy between STOs and criminal punishment, Grainger opted for the Code definition, observing:

This definition is also compatible with, and promotes the human rights of persons with a disability by ensuring that human rights such as the right to recognition and equality before the law (section 8 of the Charter), the right to freedom of movement (section 12 of the Charter), the right to liberty and security of the person (section 21 of the Charter) and the right not to be tried or punished more than once (section 26 of the Charter) are limited only in the most serious of circumstances.

Fair enough. The reasoning here basically equates compatibility with ‘least possible intrusion’, which is fine, although it doesn’t really go beyond the traditional rule that requires strict construction of provisions that limit common law rights. The Charter supports a more nuanced interpretative approach:

21(2) A person must not be subjected to arbitrary arrest or detention.

(3) A person must not be deprived of his or her liberty except on grounds, and in accordance with procedures, established by law.

An important precondition for avoiding arbitrariness in detention and for ensuring compliance with lawful requirements is for the provision authorising detention to be as clear and precise as possible. So, it’s vital that any interpretation come up with a definition that is not merely minimalist but also not susceptible to widely inconsistent factual applications.

Grainger’s definition strikes me as fitting that bill, but her application of the test to LM strikes me as very problematic.

LM has threatened people with a knife on two occasions, which has exposed people to a significant risk of being stabbed, which was likely to have resulted in a physical injury that would have endangered their lives or that was significant and longstanding. The incident in McDonalds also exposed two members of the public to a significant risk of suffering a significant and longstanding psychological injury. In addition, LM’s propensity to walk onto or lie on roads has exposed disability workers and pedestrians to a significant risk of being hit by a car if they were to try to restrain LM or remove her from the road. This in turn exposes people to a significant risk of life threatening physical injuries or significant and longstanding physical or psychological injuries. LM’s propensity to walk onto or lie on roads also exposes drivers to a significant risk of serious harm if they were involved in a car accident as a result of trying to avoid hitting LM.

I dunno about all of that. There’s a big jump from threatening with a knife to stabbing or a ‘significant and longstanding psychological injury’. It surely depends on the type of threat and the type of knife too. And the traffic risks likewise depend on the nature of the road and what was involved in LM walking onto it or lying on it. I think it’s notable  that, despite the multitude of dangerous thing LM did, on no occassion did any of these risks of serious harm eventuate, even in the form of injuries a little lower in the scale.

The biggest danger in protective regimes like this one is an overly broad risk assessment. Potential serious risks arise in just about every part of life. They are, no doubt, a very major part of the life of every disabled person. So, people applying these tests have to be very careful not to take an approach that basically equates disturbing behaviour with behaviour that requires the most serious forms of intervention. Indeed, the Disability Act itself warns against being overly keen to limit the rights of disabled people just because of there are risks involved:

5(3) Disability services should-… (n) be provided in a way which reasonably balances safety with the right of persons with a disability to choose to participate in activities involving a degree of risk…

Another way of putting it is that Grainger failed to define ‘significant risk’ in s191(6)(a). If she’d looked to Victoria’s caselaw on the crime of endangerment, or to the Court of Appeal’s approach to the regime for supervising serious sex offenders – both regimes that use a risk assessment to determine whether people should have their liberty restrained –  she may well have hesitated to draw the links she did. In TSL, Callaway relied on the Charter to observe:

A person subject to an extended supervision order is a prisoner in all but name. The threshold would be far too low, in a free society, if a court had a discretion to make an extended supervision order simply because it was satisfied that there was “a substantial – a ‘real and not remote’ – chance” of his or her re-offending.

It’s worth thinking hard about whether supervised treatment orders should have the same threshold.

The Office of the Public Advocate took a different tack, pointing out that most of LM’s dangerous behaviour was actually a byproduct of attempts at self-harm. Grainger rejected that argument a little too easily:

In my view, the context in which the risk of serious harm arises is not relevant. Whilst I agree… that LM’s violent or dangerous behaviour appears to stem from a desire to harm herself and not other people, this does not alter the fact that her behaviour has been violent and dangerous and has exposed disability workers and members of the public to a significant risk of serious harm.

It’s true that harm is harm, no matter what the context. But it’s also clear that the STOs are not meant to be triggered by attempts at self-harm.  So, it seems to me that VCAT has to be extremely careful not to  equate self-harm attempts with dangers of serious harm to others, when it comes to STOs. Grainger’s reasoning, in focussing on fairly distant risks to others, including risks to disability workers trying to prevent self-harm, seems to come very close to doing this.

Once Grainger made her finding that LM posed a significant risk of serious harm to others, the OPA was really blowing against the wind in trying to argue that LM didn’t satisfy the remaining criteria:

191(6)(b) there is a significant risk of serious harm to another person which cannot be substantially reduced by using less restrictive means;

(c) the services to be provided to the person in accordance with the treatment plan will be of benefit to the person and substantially reduce the significant risk of serious harm to another person;

(d) the person is unable or unwilling to consent to voluntarily complying with a treatment plan to substantially reduce the significant risk of serious harm to another person;

(e) it is necessary to detain the person to ensure compliance with the treatment plan and prevent a significant risk of serious harm to another person.

The middle two grounds were uncontroversial in this case, but the OPA argued that guardianship and non-STO disability law could sustain the door-locking, supervision and police involvement deemed necessary to manage LM. As occurred in MM, the disability service provider argued against an ad hoc scheme created just for LM, in favour of the more familiar and certain STO scheme, noting, in particular, the police’s wish for the express detention authorisation that flows from granting an STO.

As in MM, VCAT’s finding that the grounds were established was followed (rather than preceded) by a Charter analysis of the entire STO scheme. Again, neither lawyer had much to submit on this question. And, again, the Charter analysis was basically an assessment of whether STOs engage and (if so) reasonably limit rights. And, once again, the finding was that they either didn’t engage rights or did so reasonably:

  • Charter s. 8: Grainger found that STOs limit the right to ‘equal and effective protection against [impairment] discrimination’ – shouldn’t that be the right to equal enjoyment of human rights? – but relied on the ‘least restrictive means’ language in the Disability Act to find that STOs satisfy Charter s. 7(2)
  • Charter s. 12: ditto (i.e. ‘the same considerations justify the limitation on LM’s “freedom to choose were to live”.)
  • Charter s. 21: Grainger found that STOs do not engage or limit this right because LM’s detention is in accordance with the law as required by Charter s. 21(3). She ignores the more nuanced concept of ‘accordance with law’ that comparative human rights law provides. And what about Charter ss. 21(1) and (2)?
  • Charter s. 26: Grainger held that there’s no double punishment because of the protective nature of STOs. That’s almost certainly correct here, but it affirm the need to be careful not to merely equate criminal behaviour (threats) with the danger of serious harm to others, and to be very careful about how the STO is implemented.

Grainger’s analysis is, at least, better than the one in MM. And, again, while I think some of the rights were poorly analysed, the conclusions are obviously defensible.

But, again, VCAT’s quick leap to a rights compatibility analysis – without any reference to operative provisions – obscures the real Charter issue. Apart from the risk analysis issue discussed above, there doesn’t seem to be any statutory interpretation questionto be resolved. So, the relevant Charter question is whether the STO scheme might breach the conduct mandate:

38(1) Subject to this section, it is unlawful for a public authority to act in a way that is incompatible with a human right or, in making a decision, to fail to give proper consideration to a relevant human right.

Grainger doesn’t address Charter s. 38 at all and therefore doesn’t consider the unresolved questions of whether VCAT has any supervisory role over disability service providers when it comes to their Charter obligations and whether VCAT itself is bound by the conduct mandate in its capacity in ordering STOs. Either way, the application of the conduct mandate doesn’t depend on whether or not STOs are reasonable limits under Charter s. 7(2) , but – as I’ve repeatedly argued in this blog – whether or not the Charter s. 38(2) test is satisfied:

38(2) Subsection (1) does not apply if, as a result of a statutory provision or a provision made by or under an Act of the Commonwealth or otherwise under law, the public authority could not reasonably have acted differently or made a different decision.

That is, whether or not options that are less restrictive of LM’s rights are reasonably open under the s191 of the Disability Act.

I know far too little about either the Disability Act or the treatment of LM to make that judgment. But one particular conclusion of Grainger’s strikes me as falling far short of the kind of analysis Charter s. 38(2) would require:

Mr Walkinshaw [for the OPA] urged the Tribunal to make a short order so that the order was reviewed as soon as possible. Mr Card [for DHS] and Mr Bayliss [for the service provider] urged the Tribunal to make a 12 month order on the bass that the treatment plan has been approved by the Senior Practitioner for 12 months and LM is distressed by the hearings. I consider that it is appropriate to order that the STO be reviewed by 9 July 2009, being the review date for the treatment plan, as the treatment plan dated 9 July 2008 is a detailed, well-articulated document that has been approved by the Senior Practitioner for 12 months and I cannot find any justification for a shorter order in the circumstances. I also note that the Authorised Program Officer undertook to seek a review, variation or revocation of the STO… if any of the matters specified… may have ceased to apply or if she believes there should be a reduction in the restrictive interventions contained in the treatment plan.

Given that everyone accepts that LM has been improving a lot recently, and the marginal nature of her falling within the s191 conditions, Grainger’s approach seems to get the burden of proof all wrong. If VCAT is bound to act compatibly with human rights except to the extent provided by Charter s. 38(2), then VCAT needs to issue the shortest possible STO that’s reasonable under the statute. The statute provides for a maximum of 12 months, but clearly contemplates a shorter order, so surely such an order should be carefully considered in the case of a subject who is showing signs of improvement and who barely scrapes into the precondtions for STOs in the first place. As I’ve argued previously, the willingness of the service provider to make reviews is no answer to a question about whether or not a tribunal should remain seized of the question. Indeed, European courts have held that constant review by a court or tribunal is mandatory when there is cause to think that a person’s relevant circumtsances might be quickly changing. (The Charter’s s21(7) complicates this issue by limiting such reviews to a formal court, rather than to a court-like entity such as VCAT. But Charter s. 24(1) probably has the same effect.)

To flog the same poor old horse one more time: a Charter s. 7(2) analysis is really only appropriate when you’re not sure what the law means. When you are sure, but are contemplating what particular decisions you’re going to make, a Charter s. 38(2) analysis – directing you to the least intrusive decision that the statutory scheme (properly interpreted) reasonably permits – is the way to go and the best way to avoid reducing the Charter’s human rights culture to a mere exercise in box-ticking.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s