When National revealed its "law and order" policy before the last election, I wrote this post on it. Now that Judith Collins reportedly is preparing to introduce legislation to deliver that policy, here are some more thoughts. 

Prior to the 2011 election, Judith Collins announced that National planned to legislate to permit the ongoing "civil detention" of offenders deemed at high risk of future sexual or violent offending even after their jail sentences were complete. Civil detention - now apparently called "Public Protection Orders", because who could disagree with "Public Protection"? - would thus be a retrospective restriction applied to some prisoners on top of the original sentence that they received for their crimes, based purely on the prediction that they inevitably will commit further offences when and if released.

At the time of Collins' announcement, the general assumption was (and still is) that the policy was aimed at preventing the release of Stuart Murray Wilson (aka "the Beast of Blenheim"), even though his name was not mentioned by Collins at the time. However, I'm starting to wonder whether that general assumption about the legislation's intent was misplaced.

First of all, if Wilson is the reason for this legislation, then we have a problem - or, rather, National does. You see, Wilson has to be released from prison on September 1, as his entire 21-year-long jail term will have been served at that point. However, September 1 is now less than five months away, and there is not even a Bill before the House to allow for Wilson's ongoing detention.

So the only way that a Bill could be introduced to the House, passed through the necessary stages and enacted into law in time to allow for the relevant applications to be made to the High Court to permit Wilson's ongoing detention is if some of those necessary stages are radically curtailed. In particular, the select committee stage - which normally allows for 6 months of public submissions and consideration - will have to be chopped back considerably.

I'm going to do the National Government the courtesy of assuming that it is not intending to truncate the process of parliamentary deliberation in order to enact a likely controversial piece of legislation that has significant human rights implications simply to target one individual. Because if it was intending to do that, it would be committing a gross breach of constitutional and parliamentary principle. And I'd prefer to believe that the New Zealand Government, no matter which party is in charge of it, does not set out to act in such a fashion - at least, I prefer to believe that until presented with evidence to the contrary.

What this means is that - assuming the National Government doesn't abuse parliamentary process in an indefensible way - Wilson is going to have to be released from prison on 1 September. But it doesn't necessarily mean he has to be allowed to go off on his merry way without anyone being able to watch over him. If he really does pose such a threat of future offending as to require new legislation to allow his continued detention behind bars, he already would appear to be a prime candidate for an Extended Supervision Order. (Note that one of the offences Wilson was convicted of was rape of a girl under the age of 14 - so he meets the offending threshold for such an order.)

[UPDATE: I may be wrong about Wilson being eligible for an ESO - having looked at the matter again, the Christchurch Press reported in 2007 that "In 1996, the court found that between 1971 and 1994 Wilson raped six women, raped a girl under 14, committed bestiality, assaulted two females, and indecently assaulted two children and four women." However, the more recent Sunday Star Times report by Adam Dudding refers to "how he allegedly raped a 14-year-old in front of her mother to teach the older woman a lesson ...".]

[UPDATE 2: I wasn't wrong about Wilson being eligible for an ESO - according to Radio NZ, the probation service is beginning the process of applying for one.]

Therefore, by the time the legislation enacting the civil detention regime is enacted into law, Wilson already will be in the community - possibly under intensive supervision (if he's really thought to be such a danger to those under 16) for a period of up to 10 years. And while his apparent attitude to such supervision is not great - witness his claim to not "give a stuff about" a post-release condition not to associate with children under 16 - there are teeth to such orders. If he breaches one, he can be jailed for up to another 2 years ... and then re-emerge with the order still in place over him.

All of which makes me wonder if perhaps the proposed Public Protection Orders regime actually isn't meant to catch Wilson, and genuinely is aimed at the mystery "5-to-12" inmates that Judith Collins has cited as likely recipients. I label these targets as "mystery" because we don't yet know the exact grounds upon which a Public Protection Order may be sought, thus we can't know who may qualify for one. Meaning that, for instance, when we see a claim like "it seems the Government had [John] Gillies, who has been thumbing his nose at society and the Parole Board for years, in mind", this may be no more than wishful thinking.

But while we await the details of the Government's proposed legislation, there's a couple of other things to think about. First of all, any system of Public Protection Orders will rely on being able to tell who really is a "dangerous" prisoner that will reoffend in a violent or sexual manner once released and who isn't one. And the methods of making such determinations are, it appears, not infallible.

For example, this paper on the concept of "dangerousness" in criminal justice policy (written by a prison manager from the UK) notes that:

In terms of risk assessment, it has been established that actuarial tools, drawn from statistical analysis of data on offenders, are generally more reliable than the clinical judgments of professionals alone, but even then, while the accuracy of actuarial tools will be high for very frequent minor offences, for serious violent and sexual offences their accuracy is not high and there are significant numbers of false positives. As a result, the ability of actuarial risk assessment tools to accurately predict risk in terms of dangerousness is less powerful than is commonly assumed.

The problems with predicting what any given individual may or may not do, based in large part upon what other individuals with shared characteristics have or have not done, is expanded on in this paper:

While actuarial methods may be more reliable than the alternatives, the application of group level data to individuals – especially where the group data relate to very rare occurrences such as incidents of serious violence – necessarily gives rise to errors, including both ‘false negatives’ and ‘false positives’. The calculation of margins of error is thus a critical part of working with these assessment tools.There are also more theoretical problems with actuarial risk assessment: notably, it can be argued that it is inappropriate for sentence severity to reflect an individual’s ‘risk factors’ that relate to social or demographic characteristics over which he has no control.

Consequently, it is unavoidable that any system of Public Protection Orders will capture some individuals who - despite the assuredness of those evaluating the individual - actually will not commit any further violent or sexual crimes if released. Furthermore, given the uncertainties inherent in the task (the "margin of error" involved), we might ask whether those involved in it may choose to err on the side of caution when conducting it. After all, no-one wants to be the psychologist, parole board member or judge who lets a "dangerous" inmate back out onto the street, only to see him (or her) create another victim. 

But, of course, the counter argument is that even if a Public Protection Order scheme sweeps in some current prisoners who actually won't reoffend in a violent or sexual manner, it is better to adopt the precautionary principle here. Or, in other words, better a non-reoffending prisoner be locked up than an offending prisoner be let out to cause more harm. 

That is a value call that doesn't have a definitively right answer. Indeed, it's a call we already make in our law with regards preventive detention - there are undoubtedly some people currently serving this sentence in New Zealand who actually wouldn't have committed the harms that the sentence was intended to prevent. But that said, the proposed Public Protection Orders differ from preventive detention in that they are imposed not because of a crime already committed, but rather purely because of predictions of a crime to come. 

That's a difference that has been held important by the European Court of Human Rights (see here), as well as the United Nations Human Rights Commission (see here and here). Both of these bodies have said it is OK for a country to sentence someone to an indefinate period of detention for something they have done (combined with a justified fear of what this shows they may do when released). However, altering a person's prison sentence once this has been imposed purely because of fears the person may do bad things in the future is a no-no from a human rights perspective.

While this fact does not stop the Government proposing (or Parliament enacting) Public Protection Orders, it is going to have to be confronted. So Judith Collins told Adam Dudding in the Sunday Star Times that: "drafting of the bill was [not] being stalled by human rights concerns, writing that 'the draft legislation will be assessed for consistency with the human rights affirmed in the New Zealand Bill of Rights Act 1990'." But that's precisely the point - you can be pretty sure the advice on that legislation is going to be that it isn't consistent with the NZBORA, and so the Attorney General will have to attach a s 7 notice to it informing the House of this fact. After all, note what the current Attorney General said back in 2009 in a s 7 notice attached to legislation amending the Extended Supervision Orders regime:

There is a difference between ordering detention as a response to previous criminal offending and ordering detention solely as a precaution against possible future offending. I think that detention which is imposed solely on the basis of possible future offending, rather than proved past offending, is inherently problematic.


That this Bill's goals could be achieved without contravening the right against arbitrary detention or double jeopardy does not provide justification for these inconsistencies. The state should not detain citizens solely on the basis of preventing future offending, nor should it punish offenders twice for the same offence.

And this view was expressed with respect to orders that simply allowed the Parole Board to tell released prisoners where to live, not orders that kept prisoners behind bars for an indeterminate length of time! So I think it unlikely that Chris Finlayson will have changed his mind on this point in the interim 3 years - meaning that if Judith Collins wants to get this legislation on the statute books, she's going to have to convince Parliament that it ought to (once again) legislate in a way that restricts fundamental human rights in a way that cannot be demonstrably justified.

John Banks has already indicated that, as the sole representative of a party of individual liberty and principle, he can't wait to do so. But I wonder how Peter Dunne feels about this particular issue?

Comments (6)

by Andrew Geddis on April 12, 2012
Andrew Geddis

I've been emailed the following comment by someone who tried posting it here, but was rejected for some unfathomable reason ... the guard-bots must have been overly cautious or something.

I've read Judith Collins' description of Public Protection Orders (PPOs) in the link you provided to her web-page (and that's all I've read, so I may not be up to date), and I'm not sure the difference between them and Preventive Detention (PD) is as stark as you suggest (that PPOs differ from PD in that they are imposed not because of a crime already committed, but rather purely because of predictions of a crime to come).

They both seem to be aimed at the same category of offenders, namely those with a history of serious violent/sexual offending who pose too great a risk of re-offending in similar ways to be sentenced and released conventionally.  Presumably PPOs, like PD, would be imposed because of a combination of a prisoner's criminal history (how else could he be labelled a serious violent/sexual offender?) and the risk that he will re-offend in a similar way in the future.  So to the extent that the PPO is based on the prisoner's criminal history, it would be imposed for offending already committed - like PD.  Similarly, PD can't be imposed unless, like with PPOs, the Court is satisfied that the prisoner poses a sufficient risk of re-offending in a similarly serious way in the future.  So they are potentially quite similar orders/sentences. 

The difference seems to me to be one of timing, specifically when that risk of future offending becomes sufficiently grave and apparent to do away with conventional sentences.  In the case of someone sentenced to PD, it will obviously be at sentencing (because that's when PD is imposed), whereas in the case of someone slapped with a PPO, it will be subsequent to sentencing (because otherwise, presumably, he would have been sentenced to PD).  Isn't that why you're able to say that PPOs are imposed purely on the basis of predicted offending, and PD for offending already committed (because, with PD, the risk happens to be apparent at the time he's being sentenced for a qualifying offence, so he can be sentenced to PD for that offence, even though the sentence is actually a reflection of his entire criminal history and likely future risk - like a subsequently imposed PPO would be were he not sentenced to PD)?

Take two prisoners (A and B) who have equally horrific criminal histories and who pose equally serious long-term threats to the public.  The only difference between them is that the threats posed by A are apparent to all and sundry at the time he's sentenced for his most recent serious offence, whereas, for some reason, those posed by B aren't.  So Prisoner A is sentenced to preventive detention, and Prisoner B to a conventional sentence.  Several years into Prisoner B's sentence, though, it becomes apparent that he poses an equally grave long-term threat to the public as Prisoner A.  They're effectively in the same boat now, only there's no guarantee that Prisoner B will not be released until he no longer poses such a threat.  So the Crown (or whoever) applies for a PPO in respect of Prisoner B.  Ultimately both the PD imposed on Prisoner A and the PPO imposed on Prisoner B, regardless of when they were imposed, reflect their criminal histories and the future threats they pose to the public.

Sorry, I could feel that blowing out as I was writing it.  It'll be interesting to see what criteria and justifications the Government comes up with for PPOs.  Incidentally, was a s7 flag ever pinned to preventive detention, perhaps when it was amended in 2002?
My only comment would be - yes, assuming there is difference in the test used for preventive detention/PPO's, then the issue would come down to retrospectivity and double jeopardy. That's because I'm not sure how you'd have: "two prisoners (A and B) who have equally horrific criminal histories and who pose equally serious long-term threats to the public.  The only difference between them is that the threats posed by A are apparent to all and sundry at the time he's sentenced for his most recent serious offence, whereas, for some reason, those posed by B aren't." More likely (as I speculate in my previous post, linked in the intro to this one), you'll have two prisoners (A and B) considered of equal risk ... but A committed crimes at a time when PD wasn't an available sentencing option, while B committed crimes when it was (and so got it). The effect of this change would be to (in essence) now make A subject to PD.

As for whether "was a s7 flag ever pinned to preventive detention, perhaps when it was amended in 2002" - not that I know of. But that's not that surprising - prospective PD regimes applied at time of sentence for crimes committed have been given green lights by human rights courts/bodies around the world.
by mickysavage on April 16, 2012

It is interesting that the ESO sentence has a provision specifically confirming that it has retrospective effect (s 107(c)(2) Parole Act 2002) and the Courts have recognised (in Belcher v CEDOC and also in CEDOC v McDonnell) that the ESO sentence is punitive in nature and that the retrospective element of the ESO sentence appears to breach the BORA protection against double jeopardy.


This has not stopped the Court from making ESO orders but there may be a civil suit possible against the Crown by an affected individual for having their rights breached.

by Andrew Geddis on April 16, 2012
Andrew Geddis


The retrospective nature of the ESO regime was a basis for the AG issuing a s 7 report under the NZBORA back in 2003 (when it was first introduced to Parliament). It also caused 4 of the Greens' MPs to vote against the Bill.

However, now it is law, I don't know that you could launch a civil suit against the Crown "by an affected individual for having their rights breached". You'd essentially have to argue that even though Parliament can legislate inconsistently with the NZBORA (see s 4), the Courts nevertheless may award damages for any breach that results from such legislation. I can see a couple of problems with this:

(1) It requires the Courts to pass direct judgment on the merits of the law Parliament has made (and remember, the Courts haven't even yet said for certain they can make declarations of inconsistency under the NZBORA);

(2) How is "the Crown" responsible for the law that Parliament makes? They aren't the same thing, and so if "the Crown" is liable in damages, you need to show how and why it is accountable for what Parliament does.

by mickysavage on April 16, 2012



My comments are based on the Belcher decision (no 2) where there was some discussion on whether or not a declaration of inconsisetency with the NZBOR can be made in a criminal proceeding.  The Court said [para 16] that a declaration should be sought in a civil proceeding commenced in the High Court essentially to overcome jurisdictional problems in the District Court and also on appeal.  That way the section 5 "reasonal limits" issue can be litigated outside of the criminal trial process.


My comment about a civil suit hinted that damages may be available.  I agree this is a really long stretch and it may be that a delcaration of inconsistency is all that can be achieved.


In relation to parties the Crown is the only one that I can think of to sue.  An action against Parliament itself would no doubt fail.

by Andrew Geddis on April 16, 2012
Andrew Geddis


Yes - sorry - misunderstood what you meant by "civil action". Although I'm pretty cynical about "declarations of inconsistency" ... I mean, what do they tell us that the AG's s7 notice hasn't already said? So even if you can get one (which hasn't yet been established for sure), I'd query their practical utility.

by DeepRed on April 28, 2012

Minority Report is no longer fiction.What's next, pass laws?

Post new comment

You must be logged in to post a comment.