Even with a legal basis, complete with a nuanced suite of safeguards, there seems to be consensus that the success of any type of law reform in this area will be contingent on the resources that governments invest. Inclusion International suggests the realisation of a new fundamental right to make decisions will require a range of reforms, including the way supports are provided by governments and community, and that legal reform ‘must go hand in hand with strategies for building community supports’.
‘Simply eliminating all forms of substitute decision making without providing the necessary support for decision making would in effect deprive people with an intellectual disability of the power to make decisions.’
Flynn and Arstein-Kerslake state that a human rights approach to this reform suggests there can be no limit to the level of support provided for a person to be supported. The South Australian Public Advocate reported in its trial of SDM the importance of adequate resourcing. The ALRC Report itself notes submissions from a number of stakeholders, including the Queensland OPA, that argued with ‘careful guidance, education, training and advice’ a rights-based approach to decision making could be fraught with the same issues as currently existing practices, and advocated for codes of practice to guide decisions makers.
In terms of training material, the Autistic Self Advocacy Network (ASAN) has released ‘An Act Relating to the Recognition of a Supported Healthcare Decision-Making Agreement for Adults with Disabilities’, and accompanying educative material, including a toolkit providing information for health care providers on liability.
It seems probable that governments will want cost benefit analysis regarding the introduction of SDM, but Gooding advocates that such an assessment must also consider the comparative costs of the current system. Governments must also consider adequately resourcing advocacy groups that represent people with a disability and oversight agencies. This remains an ongoing issue in the rollout of the NDIS and cannot be underestimated as a key issue for the ACT Government to consider in any reform to guardianship law.
Gooding submits that informal efforts by people with disability, their families and others should not replace government resourcing. The Victorian trial of SDM offers a reasonable compromise, as it seeks to mobilize citizen volunteers utilising a paid facilitator, to develop relationships around isolated persons and build social capital in the community. The concept of independent monitors is a common feature of SDM, at least found in both the South Australian trial and the British Columbia representative agreement models. The monitor is someone to oversee the supporter, and seems worthy of further investigation as part of ACT trials and law reform.
One of the most discussed aspects of General Comment 1 and Article 12 is how it relates to those whose will and preferences may be very difficult to decipher. These could include people who have been institutionalised, those in a ‘persistent vegetative state’, or coma. Quinn notes that there are dangers that ‘at some point the supports and assistants will end up making decisions for and not with those individuals’. He argues that a return to substitute decision making is not the answer for this group. Instead, he submits that Article 12.3 requires the state to take additional steps, including advancing research that provides the possibility, even if remote, that a person’s brain can be read to detect preferences. In relation to those whose social connectedness has been severed due to institutionalisation, then the state should create social conditions to kindle some ‘kernel or preference’.
Nonetheless, Quinn accepts that for some individuals, decisions will need to be made ‘for’ them and that the State and law must be honest about this. Quinn suggests the role of the law remains vital for this group, as there are risks that the community around that person might ascribe to him or her a will or preference based on the community or culture in which they live. In doing so, the ‘right to be different’ would be smothered.
Similarly, Gooding questions how new regimes of SDM would apply to those in ‘a coma’ or who have no relationships of trust, and/or there is not enough intention expressed to guide decision making. Like Quinn, he agrees there are ‘clear risks to stretching the meaning’ of SDM to where decisions are ultimately made for the person.
In General Comment 1, the Committee contemplates a system that has been labelled ‘facilitated decision making’, in which the ‘best interpretation’ by friends or family, or if necessary, public officials, is used to determine a person’s wishes and preferences based on the person’s life story, values and beliefs. The Committee suggests that such a system must be based on the protections in Article 12(4), including respect for the person’s rights, will and preferences.
‘The ‘best interests’ principles is not a safeguard which complies with article 12 in relation to adults. The ‘will and preferences’ paradigm must replace the ‘best interests’ paradigm to ensure that persons with disabilities enjoy the right to legal capacity on an equal basis with others.’
Flynn and Arstein-Kerslake submit that even where a person has impairments that significantly affect communication, such a shift is possible. They suggest a support should do everything they can to still enable that person to exercise their legal capacity. This could include spending time learning their communication methods (including movements of eyelids, hand squeezing and smiling), researching past communications and speaking to those who know the person well to ascertain their values and belief systems. This would be very time-consuming as well as worthwhile, but it is worth noting that this would require the allocation of further resources at least in the case of public guardianship, as the current Public Advocate’s office is guardian of last resort for about 150 Canberrans.
However, Gooding identifies that State parties may struggle to implement such an approach for certain individuals, particularly for a person who is ‘non-verbal, requires intensive support and has no financial experience or apparent preference regarding personal finance’. As Gooding notes, prior to the release of General Comment 1, the Victorian Law Reform Commission recommended that in exceptional circumstances, the overarching goal should be to promote the personal and social wellbeing of the supported person, with other principles applied to support this, including how a ‘prudent person’ would administer a person’s finances. While worthy of consideration, such goals and principles Gooding submits are indistinguishable from the ‘best interests standard’. 
Gooding proposes somewhat of a hybrid of the VLRC model and that envisaged in General Comment 1, in which the best interpretation by supporters remains the principle driver, even as the other guiding principles advanced by VLRC are applied. In the situation of a person who is non-communicative, supporters could reasonably presume that a person wishes to be free from harm and vulnerability, and therefore prefer secure and stable investments. Even then, Gooding submits ‘robust mechanisms’ would be required to prevent exploitation, as would arbitration procedures, such as by ACAT, to resolve disputes.
The respect for the rights of such individuals is considered by the ALRC Report, which suggests where a person’s will and preferences cannot be determined, representatives must act to uphold the person’s right and in a way least restrictive of those rights, a protection consistent with both the ACT Human Rights Act and Article 12 (4).
Even with this combination of protections, Gooding notes that there are still scenarios where such a framework will require greater clarity or safeguards:
A notable barrier to the introduction of supported decision making systems is the lack of research into the various aspects and impact of their use. Carney suggests ‘some reform caution is warranted, as international evidence of exactly what forms of assisted decision making regimes work in practice is lacking. Key areas of knowledge that academics have noted include:
Gooding submits that further research is needed to inform how states should define ‘undue influence’, how to identify and address when it occurs, and the limits of the law in entering such interpersonal relationships. Kohn, Blumenthal and Campbell also suggest more evidence is needed to demonstrate that SDM in practice will empower persons with disabilities to make decisions for themselves, as opposed to having ‘the opposite effect’. As Gooding puts it:
‘Certainly, the gap in the empirical and social science literature on how to effectively realise a support decision-making regime does little to assuage the concerns of policymaker, academics, families and others about the concept.’
In undertaking a review of the experiences of SDM from 2001 to 2011, Davidson et al concluded more research is required, noting that ‘the available evidence on implementation and impact is limited, and mainly comes from the areas of intellectual disabilities and end-of-life care.
Similarly, Kohn et al raise concerns about the lack of clarity about how decisions are made in supported decision making relationships, the effect of such relationships on the individual’s involvement and the quality of the decisions that result. They also find a lack of material on demographic information about principals, supporters or monitors (those who oversee supporters). They also suggest more work is needed on substantive outcomes, including on the psychological impacts of SDM on principals and supporters, congruency between wishes and decisions, and quality of decisions, including the effect on outcomes.
‘…there is currently insufficient empirical evidence to know the extent to which (or the conditions under which) it can remedy the problems posed by surrogate decision-making process.’
However, it must also be noted that there is a lack of empirical research into the current dominant form of guardianship regime used in many jurisdictions. This is a point made by several commentators, including those who remain unconvinced about SDM. As Blanck and Martinis note, this is likely because guardianship has been assumed to be ‘benevolent and protective’ of people with disability. Gooding similarly questions if existing substitute decision making laws are ‘under-researched’.
A necessary ‘safeguard’ then for the ACT appears to be to continue trials locally, and research those occurring in other jurisdictions fully, prior to legally embedding such a system in the longer term. This should not necessarily be used as an excuse to retain current laws indefinitely. In light of the CRPD and corresponding General Comments, there are clear human rights issues arising from the current law.
One of the most complex issues regarding reform to the way the law (and society) treats decision making capacity is how success of such reform is measured. As Quinn has noted, most human beings, most of the time, think and act irrationally. While actions and reasoning may be ‘cloaked’ in a ‘garb of rationality’ the ‘wellsprings of both thought and action often run much deeper.’ He notes the importance of the ‘dignity of risk’ principle, but this must be about process as much as outcome.
‘Dignity of risk, I suggest, doesn’t just mean allowing freedom in order to enable people to learn and alter behaviour – it is valuable in itself because it allows for dignity in taking risk.’ 
Quinn notes that the test of decision-making applied to people with a disability has been ‘exceedingly high’ where poor choices and decision making is not tolerated to such an extent that people without disability would have difficulty surmounting it.
Finally, reflecting the barriers and issues identified above, while law reform is needed to make the current laws more human rights compatible, any transition to new regimes must be done carefully, systematically and in consultation with people with a disability and their carers. Even the most robust and appropriate regime of safeguards will be undermined by a rushed introduction of a new model.