Submission: Public safety and serious offenders: a review of preventive detention and post-sentence orders
About the New Zealand Council for Civil Liberties
- The New Zealand Council for Civil Liberties (‘the Council’) is a voluntary, not-for-profit organisation which advocates to promote human rights and maintain civil liberties.
Introduction
- The Council would like to commend the Law Commission on its excellent issues paper on preventive detention and post-sentence orders (issues paper). While the Council acknowledges that different goals prompt different outputs, we are continually disappointed that neither the Ministry of Justice nor Attorney General discuss these issues with this openness and detail when they are supporting legislation.
- The Council opposes preventative detention and post-sentence orders in principle. Our reasons are precisely those discussed in chapter 3: these measures are arbitrary detention prohibited by the International Covenant on Civil and Political Rights (ICCPR); and are second punishments prohibited by the New Zealand Bill of Rights Act 1990 (NZBORA).
- The Council reluctantly acknowledges that the rest of our systems are in such an unsatisfactory state that we can not recommend abolishing preventative detention at this time. There are small numbers of people who are dangers to themselves and others who are currently managed though preventative detention. We recommend the eventual decommissioning of preventive detention in favour of other measures outside of the Justice sector.
- The second strategic area for change in Corrections Hokai Rangi strategy (2019) is “Humanising in healing”. The recommendations in this area are variations on “Safety is seen through a lens of manaaki, positive relationships,and uplifting wellbeing (i.e. well people are safe people).” The Council first notes that substantive progress has not been made despite the passage of several years. Primarily, the Council notes that Corrections itself acknowledges the need to move away from punitive measures or framings.
- The Council agrees with Hokai Rangi’s goal to replace punitive measures with healing measures. The Council firmly supports that rehabilitation and transformative work should dominate preventative detention, especially given that those measures are intended to dominate all of Corrections’ work.
- This submission will cover a few areas in detail before ending with brief responses to the dozens of questions and proposals in the issues paper.
Learning from others
- The Council believes that first principles analysis has a place in policy development. However, we are surprised that the options paper does not include a survey of overseas practises.
- The Council has not conducted a thorough review of the literature. What we have reviewed, for example the Council of Europe’s Committee of Ministers’ recommendation CM/Rec(2014)3, broadly agrees with the issues paper.
- The Council recommends the Law Commission delays completion of this report until it conducts a survey of preventative detention throughout the OECD.
Breaches of Human Rights are never justified
- Question 6 of the issues paper asks: Do you think the law is justified in providing for preventive measures that may breach human rights? If so, what types of measures are justified and why?
- Even asking this question reveals a profound misunderstanding of Human Rights and New Zealand Law.
- Section 5 of NZBORA, a law of constitutional significance, reads:
the rights and freedoms contained in this Bill of Rights may be subject only to such reasonable limits prescribed by law as can be demonstrably justified in a free and democratic society. (emphasis ours) - The issues paper repeatedly frames the key balancing between the rights of the person being deprived of liberty and the rights of people in the community to live free of violence as a “trade-off” involving “public safety” or “community safety.” By doing so the issues paper sets up a false conflict to which it then repeatedly falls victim. “Community safety” is nothing more than shorthand for the rights of people in the community. Therefore, as the paper usually concludes, the question is how great a limitation of one person’s liberty is reasonable and proportionate to the predicted increase of the liberties of all of the other people in the community. This is not a question of when to abandon our human rights, but one of picking the correct balance when balancing the conflicting rights of different individuals.
- The Council will note at this point, that all imprisonment, including the initial sentence, are only justified if the incarceration of the people in prison results in greater liberty for those not in prison. Despite their vast funding our prisons fail to demonstrate that they have this positive effect. This gives weight to the prison abolition movement, of which the Council is not a member but whose evidence seems strong to us.
Evidence of Reoffending Risk
- Since our 2015 merger with Tech Liberty New Zealand the Council has had technology governance as a focus area.The Council has taken the leading role in civil society on Statistics New Zealand’s Algorithm Charter, and related matters like the Open Government Partnership.
- Section 9.12 of the report repeats a common fallacy, that by giving a human the power to make the ultimate decision, mitigates the worst algorithmic errors. As comforting as this thought is, it has been scientifically disproven. In practice, people give more credence to stronger signals from algorithms, and these outlier signals are less likely to be accurate. Putting a human in the loop does not mitigate algorithmic errors.
- The Council agrees with the issues raised in sections 9.15 to 9.20.
- The Council contacted Corrections about section 9.23. From that interaction, the Council is confident that Corrections has not, in the past, taken the responsible and appropriate steps that section 9.23 expects. The Council doubts that Corrections will take those steps now, and therefore suggests that the report’s conclusion on Question 28 is not well founded.
- The Council rejects section 9.26’s suggestion that the Algorithm Charter provides the necessary governance framework to create a just outcome. The Council noted in 2019 that “the Charter is a short page of principles without supporting documents”. Despite our continued efforts in the intervening years, this remains true. The Algorithm Charter is a vague and voluntary standard. Statistics New Zealand itself is openly critical of the Algorithm Charter1 p58 of The Government Data System Future Design Narative says “the algorithm charter is insufficient for protecting Māori rights and interests”.
Section 6.3 of the Māori Data Governance Model (co-authored by Stats) discusses the charter at length before concluding “While the Charter is a welcome initiative, responsible algorithm design and implementation requires governance, frameworks and organisation that go further.” - Regardless of how little one expects from the Algorithm Charter, Corrections has failed to abide by it. The charter’s transparency requirements are:
Transparency – maintain transparency by clearly explaining how decisions are informed by algorithms.
This may include:
- plain English documentation of the algorithm
- making information about the data and processes available (unless a lawful restriction prevents this)
- publishing information about how data are collected, secured and stored.
This information is not available for the Corrections algorithms cited in the report. It is a fact, not a risk as suggested by sections 9.19-9.20, that the tools have not been adequately scrutinised.
- The Council is sceptical that Corrections have met the other standards set out in the Algorithm Charter. They did not offer any evidence of having done so when we prompted them for it.
- Luckily, in an ideal world, all of these algorithm governance failings are moot. All of the algorithms discussed in chapter 9 are supervised learning algorithms. Without exception, supervised learning algorithms are machines for repeating past behaviours. When an organisation has identified a need to make a clean break with a deeply flawed past, as Corrections does routinely, then it can’t use a supervised learning algorithm on the behaviours which need to change. The only possible contribution that these algorithms could have is to embed the mistakes of a racist past, and thereby to contribute to the failure of numerous all-of-government initiatives. The Council therefore opposes all uses of predictive algorithms in sentencing or preventative detention.
- If we are to insist on using algorithms in sentencing or preventative detention, then we should not even consider using algorithms trained on our history of failures. The Council’s fundamental opposition in this regard can be avoided by adopting an algorithm, or training our own algorithm on data, from a foreign jurisdiction, for example Denmark or Sweden, whose penal system has the characteristics of justice which we intend to introduce to our own.
Responses to Questions
Questions on which the Council has no position have been omitted. The question or proposal has been omitted when its length made that seem reasonable.
Chapter 3
Q7: If the law is to continue to provide for preventive detention, do you agree the law should be reformed to demarcate more clearly the first and second periods of preventive detention to align with human rights law?
The Council completely supports reform to clearly demarcate preventative detention.
Q8: Do you think that people who are detained after completing what may be regarded as their punitive prison sentence should be managed in different conditions to prison?
The Council supports managing people in preventive detention in different conditions than prison. The Council notes that all people in prisons have the right to be treated better than they currently are.
Q9: Do you think the preventive regimes should have a stronger focus on therapeutic and rehabilitative treatment and provide stronger rights to treatment for people detained?
The Council supports vastly expanded and improved therapeutic and rehabilitative treatment for all people in prisons. If preventative detention is allowed to continue, then the people detained must be provided with even more treatment.
Chapter 4
Q10: Do you agree with the issues we have identified regarding the fragmentation of the law? Are there other issues we should consider?
The Council remains outraged that Public Protection Orders (PPO) are heard by civil courts. These orders are extreme measures of dubious morality and the defendant has a right to the strongest protections available, which necessitate that hearings only be conducted in criminal courts.
The Council further agrees that a single, coherent hearing helps to minimise the burden on the liberties of the defendant.
Chapter 5
Q11: Do you agree that preventive detention is not an appropriate measure for responding to risks of serious reoffending by young adults who have been convicted of serious sexual or violent offending?
The Council reiterates that preventative detention is not an appropriate measure for anyone.
The Council agrees that preventative detention makes significantly less sense when applied to young adults, and concurs that it should not be used on young adults.
Chapter 6
Q12: Do you think the qualifying offences are serious enough to justify making a person eligible for a preventive regime?
If preventative detention is to be retained, the Council would support the issue’s paper’s recommendation that indecent assault be removed from the qualifying offences.
Q13: Should the same offences be qualifying offences for all preventive regimes? If so, which offences should qualify?
The qualifying offences should be the same for all. None of the Films, Videos, and Publications Act offences are serious enough to warrant being on the list.
Q14: Do you consider any of the offences we discuss that are omitted should be qualifying offences for preventive detention, ESOs and PPOs?
The Council opposes expansion of preventative detention. The Council agrees that the qualifying offences must be the same for offences committed domestically as it is for offences committed overseas. That should be accomplished by removing any offence which does not currently qualify in both jurisdictions.
Q15: Do you agree that strangulation should be a qualifying offence for preventive detention, ESOs and PPOs?
The Council agrees that the omission of the new offence of strangulation is inconsistent with the handling of other offences of similar severity. The Council entirely believes the hypothesis that the omission is administrative error rather than intent.
Q16: Do you agree that incest should be removed as a qualifying offence for preventive detention, ESOs and PPOs?
The Council is opposed to the intervention of the state into the sex lives of consenting adults. Given that the state has defined incest primarily to interfere with the lives of consenting adults, the Council fully supports the removal of incest from the qualifying offences.
Chapter 7
Q19: Should a person be eligible for an ESO on the basis of overseas offending that would not come within the description of a qualifying offence if committed in Aotearoa New Zealand, if:
- the person has been convicted of an offence overseas that would constitute an imprisonable offence in Aotearoa New Zealand;
- the person was sentenced to more than one year of imprisonment for that offence;
- the person is returning or has returned to Aotearoa New Zealand more than six months after release from custody; and
- immediately before the person’s return to Aotearoa New Zealand, the person was subject to monitoring, supervision, or other conditions for the offence, or to conditions imposed under an order in the nature of an ESO or a PPO.
No. The Council believes that people should never be imprisoned or otherwise punished in Aotearoa New Zealand for acts which are legal here.
Q20: Are there any issues arising with the timing of ESO applications for overseas offenders, or with accessing information that require legislative reform?
Section 7.27 notes that an ESO for a returning Citizen must be made within six months, and suggests that this timing is insufficient. The Council agrees that the timing should be revised in legislation. However, the appropriate adjustment would be to reduce the time allotted to 30 days, not to allow any extension. That anyone is imprisoned for lack of sufficient evidence discredits all of us.
Chapter 8
Q22: Do the legislative tests for preventive detention, ESOs and PPOs focus on the right level of likelihood of possible future reoffending?
The Council again reminds the reader of our opposition in principle to all preventative detention. If we are to retain preventative detention, the highest threshold of probability should apply to all types of orders, with the same threshold for violent and sex offences.
Q23: Do you think there are any issues with the qualifying offences that a person must pose a risk of committing for the court to impose preventive detention, an ESO or a PPO?
The Council agrees with section 8.21. The offences listed in 8.21 should not be qualifying offences for preventative detention.
Q24: Do you think that it is an issue that the human rights considerations the courts apply when imposing a preventive measure are not referred to in the primary legislative tests?
The Council prefers that the primary legislation is clear and complete. We therefore support changing the legislation to align it with practise.
Q25: Do you agree with the issues we have raised concerning the traits and behavioural characteristics in the legislative tests for ESOs and PPOs?
The Council strongly agrees with sections 8.33 to 8.45. The current system basically makes mental illness a pre-condition for preventative detention. Instead the criteria should be fundamentally altered to avoid detain mentally disabled people and instead direct them to appropriate supports.
Q26: Do you agree with the issues we have identified with the legislative tests a court will apply to decide whether to impose preventive detention, an ESO or a PPO?
The Council considers the open ended timeframe for recidivism to be unreasonable. Reductio ad absurdum, If people were immortal then probabilistically everyone would be subject to preventive detention, as the even the most minute non-zero probability multiplied by infinity is certain. More practically, the Council feels no reason to question section 8.48’s conclusion that predictions beyond five years are of insufficient merit to deserve consideration.
Chapter 9
Q28:Do you agree with the issues we have identified regarding evidential matters and our preliminary conclusion that legislative reform is not generally needed to address these issues?
The Council strongly disagrees. Legislative reform is badly needed in Aotearoa New Zealand in order to catch up to the standards of Algorithm Governance prevalent in OECD. To cite the easiest and most important example, the idea that our algorithm governance is equivalent to the EU’s GDPR is laughable.
Q29: Do you think the possibility that risk assessment tools may be inappropriately used on Māori is an issue requiring reform? If so, why, and what reforms should be implemented?
As discussed previously, the Council has no doubt that risk assessment tools are being used inappropriately against Māori, Pasifika, and other minorities by the justice system in Aotearoa New Zealand. The harms caused by those tools are significant and widespread.
Q30: Do you think that the legislation should promote opportunities to address the court or provide information to the court for the person’s whānau, hapū or iwi or any person who has a shared sense of whānau identity around a particular kaupapa with the person?
The Council agrees. However, while the Council acknowledges and supports strong and broad interpretations of Te Tiriti, access to community support should not be limited by ethnicity. Everyone should be eligible to access the proposed supports.
Chapter 10
Q31: Do you think that the law relating to the conditions and management of people subject to release on parole from preventive detention and ESOs should appropriately allow for Māori-designed and Māori-led initiatives?
[Note the addition of the word should to clarify the question]
Yes, the law should accommodate Māori-designed and Māori-led initiatives. It should also accommodate such initiatives from other minorities whose rights do not stem from Te Tiriti.
Q32: Should the legislation build in tests or guidance to ensure that decisions about conditions are made in accordance with the NZ Bill of Rights?
The Council agrees with the Supreme Court that NZBORA is a document of constitutional significance. From that perspective, question 32 is offensive. NZBORA already binds the crown in all of the ways proposed. Supporting that interpretation of NZBORA is one of the Council’s objectives.
Q33: Do you think the term “residential restrictions” should be defined in the legislation?
The Council believes that residential restrictions are both significant limitations on civil liberties and counter productive. As such they should be abolished not codified.
Q35: Do you think the guiding principles of the Parole Act should be amended to state that people subject to ESOs must not be subject to conditions that are more onerous, or last longer, than is consistent with the safety of the community?
Yes, but you’ve backed into the correct position to avoid a proper definition of Human Rights. The limitation of one person’s rights is only ever justified by offsetting benefits to the rights of other people.
Q36: Do you think there are any issues arising from the division between the order- making and condition-setting jurisdictions for ESOs that require legislative reform?
The Council believes that preventative detention should be heard exclusively in criminal court. We further agree with the issue paper’s supposition that the Parole Board would benefit from improvement.
Q41: Do you think that the requirement not to associate with persons under 16 should be removed from the standard ESO conditions?
The Council believes that the state should not have the power to prevent people from associating with their whanau. The Council agrees with section 10.121 that this restriction is a “serious impairing of a person’s rights”. The Council agrees that such restrictions should not be standard.
Chapter 11
Q43: Should the courts have greater responsibilities for reviewing preventive detention instead of leaving the task of determining release on parole to the Parole Board?
The Council believes that preventative detention should be heard exclusively in criminal court.
Q44: Do you think the test for release from detention for people sentenced to preventive detention should expressly recognise their right to liberty except when justified by compelling reasons relating to community safety?
Yes
Q45: Do you think the test for release from detention for people sentenced to preventive detention should require “increasing justification” over time?
While agreeing wholeheartedly with section 11.45, that the initial justification must be high, the Council also believes that 11.43 is correct. The crown should need ever stronger evidence for every extension of detention.
Q46:Do you think that the test for cancelling an ESO should mirror the test for imposing an ESO?
Yes
Q47: Do you agree that an ESO should be suspended if an interim detention order is made?
Yes
Q48: Do you agree that an ESO should come to an end if a PPO is ordered?
Yes
Q50: Do you think that breaching an ESO condition should be an offence or that another mechanism should be used for ensuring compliance with ESO conditions?
The Council firmly believes that breach of conditions, be they ESO conditions or any other conditions, should not be an offence. We wholeheartedly support the reasons provided in sections 11.63(b) and (c). It behoves our community to support progress rather than demand perfection.
Chapter 12
Proposal 1:
The Council supports proposal 1. Proposal 1 is vague enough that it is difficult to see how the Crown can uphold its Te Tiriti obligations or Corrections implement Hōkai Rangi without proposal 1.
The Council believes that other minorities, especially Pasifika, have a right to similar affordances
Proposal 2:
The Council fully supports proposal 2. As noted in section 12.17 the UNHRC has ruled that this is a requirement under the ICCPR. And as noted in section 12.18, the Court of Appeal has ruled that this is a requirement under NZBORA.
Proposal 3:
The Council supports the repeal of preventative detention as proposed under P3C.
Proposal 4:
The Council prefers the single statutory regime proposed as P4B. As the issues paper has noted repeatedly, the courts have already taken it on themselves to consider all of the Acts when any of the orders is considered, so P4B both communicates and confirms current practise.
Proposal 6:
If the law continues to provide for preventive detention as a sentence, the law could provide that a person must be aged 25 years or older at the time of conviction for the qualifying offence in order to be eligible for preventive detention.
Regardless of how preventive detention occurs, the Council agrees that it should not be applied to people under the age of 25.
Proposal 7:
The same offences could be qualifying offences for preventive detention, ESOs and PPOs.
We agree.
Proposal 8:
The Council opposes the expansion of preventive detention. As noted earlier, we oppose the inclusion of offences under the FVPC Act.
Proposal 10:
To impose a preventive measure, the legislation could omit any requirement that the court be satisfied that a person displays any specific traits or behavioural characteristics other than the risk they pose to community safety by reoffending.
Agreed, except that we would suggest that the word could in the second phrase should be either should or must.
Proposal 11:
To impose a preventive measure, the legislation could require the court to assess the risk that a person will commit a qualifying offence within a certain timeframe. The requirement in the Public Safety Act that the offending be “imminent” could be repealed.
We agree.
Proposal 12B:
To impose a preventive measure, the legislation could state that the court must not impose a preventive measure unless it is satisfied that the limits the measure would impose on rights affirmed under the New Zealand Bill of Rights Act 1990 are justified.
As previously noted, the Council firmly believes that proposal 12B is already a constitutional requirement, and if it were not we utterly support it.
Proposal 13:
Section 3(b) of NZBORA reads “by any person or body in the performance of any public function, power, or duty conferred or imposed on that person or body by or pursuant to law.” It does not read, “by any person or body except for probation officers.” As such, probation officers are already bound to administer conditions, and perform all of their other duties, in ways which are consistent with NZBORA. The Council firmly believes in this constitutional requirement, and utterly supports it. We therefore support proposal 13.
Proposal 14:
The court could be responsible for setting special conditions of ESOs at the time it makes an ESO.
Agreed. These special conditions are major limitations on a person’s liberties. The Council firmly believes that any such limitation should only be made by criminal courts.
Proposal 16:
The standard ESO condition not to associate with persons under 16 years could be removed but be available as a special condition.
The Council supports this improvement to the liberties of the people subject to ESOs.
__________________
- 1p58 of The Government Data System Future Design Narative says “the algorithm charter is insufficient for protecting Māori rights and interests”.
Section 6.3 of the Māori Data Governance Model (co-authored by Stats) discusses the charter at length before concluding “While the Charter is a welcome initiative, responsible algorithm design and implementation requires governance, frameworks and organisation that go further.”