Criminal Law Review Prosecutorial Policies Law General Essay

Published Date: 02 Nov 2017

Jonathan Rogers

Subject: Criminal law

Keywords: Assisted suicide; Constitutionality; Director of Public Prosecutions; Policies; Prosecutions

Legislation: European Convention on Human Rights 1950 art.8

Suicide Act 1961

Case: R. (on the application of Purdy) v DPP [2009] UKHL 45; [2010] 1 A.C. 345 (HL)

*Crim. L.R. 543 Abstract

There are several objections to the recent decision in Purdy. Further consideration should be given in future to preferring prosecutorial "systems" of enforcement to prosecutorial "policies", where the potential scope of an offence is so wide as to be a matter of public concern.

In R. (on the application of Purdy) v DPP1 the House of Lords was confronted with a tragic application on behalf of a desperately ill person. Mrs Purdy suffers from primary progressive multiple sclerosis. She needs an electric wheelchair and her paralysis is likely to spread. She anticipates that a time will come when she will feel that life is unbearable and would want to commit suicide, but she may by that stage be physically incapacitated from doing so. She would then want to be assisted to die in a country where this can be overseen by doctors, probably Switzerland (where assisted suicide is legal, provided that it is not done for "selfish" reasons).2 She would, however, need help even to get to Switzerland, and her fear is that her husband may be prosecuted for assisting her suicide if he were to help her to make that journey. Fear of that, and of that alone, would prevent her from making the journey.3 At the time of the litigation,4 s.2(1) of the Suicide Act 1961 provided that:

"A person who aids, abets, counsels or procures the suicide of another, or an attempt by another to commit suicide, shall be liable on indictment to imprisonment for a term not exceeding fourteen years."

*Crim. L.R. 544 Despite, or perhaps even because of, the brevity of this sentence, its content was thought to be clear.5 It proscribed even modest acts of assistance and it left no room for any common law defence of necessity which might otherwise have been available to Mrs Purdy's husband, should he be prosecuted.6 But Mrs Purdy did not challenge the scope of the substantive law, and she also accepted that the DPP cannot provide relief by promising in advance not to prosecute her husband.7 She argued instead that the inhibiting effect of the offence of assisted suicide engages art.8(1) of the European Convention on Human Rights, and that in order for that interference to be "in accordance with the law" it should be supplemented by a clear and accessible prosecutorial policy relating to its enforcement. Article 8(2) of the European Convention provides:

"There shall be no interference by a public authority with the exercise of this right except such as is in accordance with law and is necessary in a democratic society in the interests of … the protection of health and morals, or for the protection of the rights and freedoms of others."

This argument, though it had been unsuccessful in the lower courts, persuaded all five of their Lordships. Consequently they granted a mandatory order that the DPP should promulgate and publish a policy relating to the factors that he will consider relevant when deciding whether it would be in the public interest to prosecute anyone for assisted suicide in cases which are comparable to that of Mrs Purdy. The DPP published an interim policy8 on September 23, 2009, less than two months after the decision in Purdy, and he announced a 12-week consultation period on its contents. A final (revised) version was published on February 25, 2010.9

Much could be written about the Purdy decision, including the preliminary issue whether a competent and autonomous person's wish to be assisted to die engages the right to respect for her private life in art.8(1). The decision that art.8is engaged required their Lordships to prefer the (ambivalent) decision of the European Court of Human Rights in Pretty v United Kingdom10 on this point to their own (ambivalent) decision to the contrary earlier in the same case.11 But the engagement of art.8 had in fact been conceded by the DPP12 and we shall say nothing further about this part of the decision. Our attention will instead be addressed to the decision that compliance with art.8(2) might require the prosecutor to promulgate an offence-specific policy, the essence of which is inevitably to suggest that certain *Crim. L.R. 545 cases of assisted suicide are most unlikely to be prosecuted for public interest reasons.13 We shall argue that offence-specific policies of the type that was mandated in Purdy are inherently objectionable, and no less so in this difficult and tragic case. However it is strongly arguable that by virtue of the same art.8(2), the DPP should have been required to set up a special system for dealing with affected cases, so that the "chilling effect" of the offence does not have disproportionate effect.14

This article divides into four sections. First, we shall elaborate upon the decision in Purdy, and secondly we will outline the main points of the DPP's Final Policy. Thirdly, we shall criticise the decision in Purdy on the grounds that it does not address the constitutional objections to suggesting that certain conduct will systematically not be prosecuted; and we shall note that promulgating an offence-specific policy was not even implicitly required by the decision of the European Court in Pretty v United Kingdom. Finally we will develop the key argument that compliance with art.8(2) should instead require the DPP to maintain an efficient and fair prosecutorial system for dealing with assisted suicide cases. The distinction between prosecutorial policies and prosecutorial systems of enforcement is relatively unfamiliar in the literature15 and was not grasped in Purdy, but it may usefully be employed in other contexts too, especially where again the substantive law is wider than would seem to be absolutely necessary.16

1. Purdy: analysis of the decision

In this section, we seek to emphasise two particular points. First, that the DPP does not seem to have challenged Mrs Purdy's underlying construction of art.8(2) of the European Convention. Instead he seems to have fought the case on the basis that the information that Mrs Purdy sought was already to be found in the Code for Crown Prosecutors, which sets out how the "public interest" test operates in the general run of cases. This may limit the weight of the precedent in Purdy; indeed it may be limited substantially to its own facts. Secondly, the DPP was not given a free hand to produce any policy on prosecuting assisted suicide that he pleased. Rather, he was specifically mandated to formulate a policy which would seek to "protect the right to make an autonomous choice".

*Crim. L.R. 546 Purdy: an uncontested and limited precedent

As we have noted already their Lordships had held in Pretty that respect for the private life of the dying person does not require or even permit the DPP to decide not to prosecute anyone in advance of any offence being committed.17 So Mrs Purdy resorted to the next available argument; that the absence of any guidance on how the public interest element in cases such as hers would be assessed meant that any ex post prosecutorial discretion would not be "in accordance with the law". In turn, the DPP should have to declare his policy (or to devise one, if he had not already done so, and then declare it) regarding the enforcement of assisted suicide in cases where the victim was understood to be an autonomous individual who was under no apparent external pressure to commit suicide.18 This argument had not been anticipated in Pretty, but acquired its force from a line of Strasbourg cases suggesting that where a discretionary power exists (that affects the enjoyment of an art.8 right) then in order for that power to "accord with the law",

"… the law must indicate with sufficient clarity the scope of any such discretion conferred on the competent authorities and the manner of its exercise."19

Their Lordships accepted the argument, apparently untroubled by the fact that none of the authorities cited involved the exercise of prosecutorial discretion20 in relation to a substantive offence the scope of which is itself clear.

So the underlying assumption in Purdy is that it is discretionary enforcement of the criminal law, and not the criminal law itself, which constitutes the "state interference". Only then does it become pertinent that the criteria for that decision are not accessible and that its exercise may not readily be predictable. It is thus important to observe that the DPP did not seem to contest this construction of the applicability of art.8 to prosecutorial discretion. Rather he relied on the point that any such guidance as Mrs Purdy could expect to receive could already be found by interpretation of the Fifth Code for Crown Prosecutors (2004).21 Further, he had already applied the Code to a similar case where he had decided not to prosecute the family members of 23-year-old Daniel James. Mr James was a young man who had become tetraplegic following a serious spinal injury occasioned in a rugby match, whose family assisted his travel to commit suicide in a clinic in Switzerland. Mr James was accepted to be

*Crim. L.R. 547 "… a mature, intelligent and fiercely independent young man with full capacity to make decisions about his medical treatment whose determination to commit suicide was not in any way influenced by the conduct or wishes of his parents."22

The DPP had also published23 his reasons in this case. But, departing from the Court of Appeal, their Lordships found that the factors listed in the "general" Code bore too little relevance to prosecutions for assisted suicide. The DPP noted that the James case was not one of the more serious cases of assisted suicide on account of the settled willingness of the victim to die24; but the relevance of this does not derive from the "general" Code and in any event the DPP acknowledged that assisted suicide itself remained a serious offence. The other point of mitigation relied upon by the DPP was that:

"Neither Daniel's parents nor the family friend stood to gain any advantage, financial or otherwise, by his death. On the contrary, for his parents, Daniel's suicide has caused them profound distress. That is a factor against prosecution."25

However--and here lies the rub--there is nothing in the "general" Code to suggest that lack of greed, or the finding of any combination of grief, distress and regret from the defendant following an offence, may constitute "public interest" reasons not to prosecute. As stated by Lord Hope:

"The question whether a prosecution is in the public interest can only be answered by bringing into account factors that are not mentioned there. Furthermore, the further factors that were taken into account in the case of Daniel James were designed to fit the facts of that case."26

So Mrs Purdy succeeded because neither the general Code nor the publication of the reasons for not prosecuting in the James case made it sufficiently clear how the DPP exercised his discretion in assisted suicide cases. It might be countered that, among the 115 or more English people known to have been assisted to die in Switzerland, only eight cases were referred to the DPP, and none of them resulted in prosecution, including the rather telling case of Daniel James.27 But clarity about the decision-making process is more than a question of outcomes. Rather, the underlying principles needed to be promulgated in order for Mrs Purdy to receive reliable legal advice about the likelihood of prosecution. Lord Hope seemed to appreciate that it might be difficult for the families to attach much weight to past practice alone and said that "this issue will not go away"28 as things presently stand.

*Crim. L.R. 548 Although the DPP lost on account of the insufficiency of the general Code, it is at least clear that future applications of Purdy may be limited. Their Lordships did accept that in most cases, the content of any general Code for Crown Prosecutors is likely to offer sufficient guidance as to likely prosecutorial decision-making.29 Case law too may assist. So, it is surely wrong to suppose that the DPP will be required to produce many separate offence-specific policies in other areas where human rights are engaged. For example, it should be clear enough already that teachers who chastise their pupils for religious reasons should expect prosecution even if their rights under art.9(1) are thereby engaged. This is because their potential liability can be justified for the purpose of protecting the rights of children,30 and because the "general" Code for Crown Prosecutors specifies that the fact that "the person was in a position of authority or trust"31 points towards a decision to prosecute. The ready applicability of the general Code to most offences and indeed to most issues in the criminal law perhaps explains the tactical decision of the DPP to rely on it in Purdy. Unfortunately this proved to be exactly the wrong case in which to adopt that tactic.

The other purpose behind the DPP's policy

It should be emphasised that the DPP was not mandated simply to produce any policy which clarifies his approach to prosecuting assisted suicide cases. Specifically, he had to produce a policy which clarified how he would determine the likely autonomy of the victim and the weight that would be given to this factor. In this context it may be recalled that in the Pretty litigation in 2001 both the House of Lords and later the European Court agreed that the criminal prohibition itself can be said to be "necessary" under art.8(2) for the purpose of protecting the rights of vulnerable people. Lord Bingham had explained that this could justify punishment even in cases where the victim had needed no protection from any kind of manipulation by family members:

"On behalf of Mrs Pretty counsel disclaims any general attack on section 2(1) of the 1961 Act and seeks to restrict his claim to the particular facts of her case: that of a mentally competent adult who knows her own mind, is free from any pressure and has made a fully-informed and voluntary decision. Whatever the need, he submits, to afford legal protection to the vulnerable, there is no justification for a blanket refusal to countenance an act of humanity in the case of someone who, like Mrs Pretty, is not vulnerable at all. Beguiling as that submission is, Dr Johnson gave two answers of enduring validity to it. First, ‘Laws are not made for particular cases but for men in general.’ Second, ‘To permit a law to be modified at discretion is to leave the community without law. It is to withdraw the direction of that public wisdom by which the deficiencies of private understanding are to be supplied’ (Boswell, Life of Johnson, Oxford Standard Authors, 3rd ed, 1970, pp 735, *Crim. L.R. 549 496). It is for member states to assess the risk and likely incidence of abuse if the prohibition on assisted suicide were relaxed … . But the risk is one which cannot be lightly discounted. The Criminal Law Revision Committee recognised how fine was the line between counselling and procuring on the one hand and aiding and abetting on the other (report, p 61, para 135). The House of Lords Select Committee recognised the undesirability of anything which could appear to encourage suicide (report, p 49, para 239) …

… It is not hard to imagine that an elderly person, in the absence of any pressure, might opt for a premature end to life if that were available, not from a desire to die or a willingness to stop living, but from a desire to stop being a burden to others."32

The European Court in turn agreed that it was open to the United Kingdom to prohibit absolutely assisted suicide, for the reasons given by Lord Bingham.33 It also agreed that the DPP could not promise in advance to turn a blind eye to the offence34--even though after the event, a decision not to prosecute, in the light of all the relevant facts, might be quite unobjectionable. Indeed, the latter remedy--non-prosecution after the event--met with some approval. The European Court noted that:

"The Government [has] stated that flexibility is provided for in individual cases by the fact that consent is needed from the DPP to bring a prosecution and by the fact that a maximum sentence is provided, allowing lesser penalties to be imposed as appropriate … It does not appear to be arbitrary to the Court for the law to reflect the importance of the right to life, by prohibiting assisted suicide while providing for a system of enforcement and adjudication which allows due regard to be given in each particular case to the public interest in bringing a prosecution, as well as to the fair and proper requirements of retribution and deterrence."35

It is clear that the European Court in Pretty was referring here to possible discontinuance in cases such as that of anyone who might assist an autonomous individual to die at a time of his or her choosing. It follows from this that the policy which the DPP must produce in order to "accord with the law" under art.8(2) mustalso address the fact that the main justification for state interference under the same art.8(2) is to protect vulnerable sufferers, and that cases involving fully autonomous victims may properly be discontinued. Crucially their Lordships in Purdy thought that the decision to discontinue in the James case was itself humane36 and proper; their criticism had instead been that it was hard to see how the DPP had reached it by applying the Code for Crown Prosecutors. Lady Hale was perhaps the most explicit in joining the dots when she pointedly advised the DPP that "the object [of the guidance which he had to produce] must be to protect the right to make an autonomous choice. The factors which tell for and against such a genuine exercise of autonomy free from pressure will be the most important."37

*Crim. L.R. 550 So, the policy which the DPP must promulgate under art.8(2) to "accord with the law" must also "accord with" the state justification for enacting the offence. The issues are not distinct. This brings us to his Final Policy itself.

2. The DPP's Final Policy on assisted suicide

The DPP focuses on the "public interest" test in the Final Policy, though some legal clarification is also offered. He points out, presumably for the benefit of his own prosecutors, that in cases where the defendant (D) gives assistance to a suicide which in fact is never attempted by the victim (V),38 then the correct charge depends upon the date of the offence. D should be charged with an attempt to commit the offence of assisted suicide, under the Criminal Attempts Act 1981, where the offence is alleged to have occurred on or before January 31, 2010. But where the same activity is alleged to have occurred since February 1, 2010, then D may be straightforwardly charged with the recently amended version of the offence under s.2 of the Suicide Act 1961, which includes doing an act "capable of encouraging or assisting the suicide" with the appropriate intention.

The format of the "public interest" part of the policy is similar to that of the "general" Code. Factors are listed which would tend to suggest that prosecution is appropriate, followed by a list of factors that point the other way, but the content of one list is by no means just the converse of the other.39 Indeed there are sixteen "pro-prosecution" factors but just six "anti-prosecution" factors. Of those factors which point towards prosecution, no less than eight relate to the vulnerability or perceived lack of independence of V, for example his or her young age or doubted mental capacity, or because he had not clearly communicated his decision or had not sought the assistance of the accused "on his or her own initiative", or was pressured by D (or by another, to the knowledge of D). Most of the other pro-prosecution factors relate to unscrupulous or somehow "professional" defendants, who may be unknown to their victims or who have helped more than one victim, or accepted payment for their help, or who provided an environment which allows someone to commit a suicide. Most of the pro-prosecution factors had also appeared in the Interim Policy. Perhaps the two most important additions in the Final Policy are the factors that D had a history of violence or abuse against V40 (which is thought to be further relevant to V's autonomy, at least in relation to D) and that D was acting in a professional capacity (nurse, doctor, professional carer, etc.).41 This latter factor appears to recognise that an element of breach of trust aggravates the offence, even if D was motivated by compassion for a particularly desperate individual.

But the more interesting list is that of the "anti-prosecution" factors, reduced to just six in the Final Policy from a proposed thirteen in the Interim Policy. In the Interim Policy, most of these factors had related (but in a positive way) to the perceived autonomy of V. But in the Final Policy all the previous victim-oriented *Crim. L.R. 551 factors were condensed into just one factor (albeit that it appears at the top of the list), namely that "the victim had reached a voluntary, clear, settled and informed decision to commit suicide".42 The other five factors relate to the suspect, including that he was motivated "wholly" by compassion, that he should have sought to dissuade V from committing suicide and, echoing the DPP's reasons for not prosecuting the family of Daniel James, that "the actions of the suspect may be characterised as reluctant encouragement or assistance in the face of a determined wish on the part of the victim to commit suicide".43 This shift was heralded in the publicity surrounding the Final Policy, and the DPP said that "the policy is now more focused on the motivation of the suspect rather than the characteristics of the victim".44 It was separately explained in the Summary of Responses that whereas only a minority of respondents to the consultation exercise arising from the interim policy thought that any of the victim-oriented mitigating factors were appropriate, by contrast, a clear majority of respondents had supported all of the suspect-oriented factors that found their way into the Final Policy.45

It must be said that whilst only the one victim-oriented mitigating factor (that "the victim had reached a voluntary, clear, settled and informed decision to commit suicide") remains, it is exactly the one which their Lordships in Purdy had thought to be crucial.46 The CPS in its Summary of Responses can say little more than that it disagrees with the 62 per cent of respondents who had objected even to this being regarded as relevant to prosecutorial policy.47 But the other victim-oriented factors that had graced the interim policy were each agreed by the CPS to be unsatisfactory. The possible suffering by V from a terminal or degenerative condition as an "anti-prosecution" factor was dropped because on consultation it was widely thought to suggest that the lives of the disabled were generically less worthy of protection under the Suicide Act.48 Separate objection was made to the factor that V may have tried to take his own life before (apparently on the basis that some unsuccessful attempts are in fact cries for help rather than representations of a settled wish to die).49 The factor that V should have been a family member or close friend to D was also dropped from the Final Policy due to concerns that families may also have the most influence (including malign influence) over suicidal persons.50 This, then, is another example of asymmetry in the Final Policy; it is a "pro-prosecution" factor that V and D should be unknown to each other, but it is thought not to be an "anti-prosecution" factor in itself that they may have had a very close relationship.

However, the deletion of all of these factors does give pause for thought. For a start, one might wonder whether, even if each of the objections to those factors is sound in itself, cumulatively the factors seem likely to provide powerful mitigation. That is to say, if V is dying or in a helpless physical condition and has already *Crim. L.R. 552 attempted suicide (unaided) and is then assisted by a close friend or family member to do so, then it may seem to be going too far to add up all the separate objections in order to disregard the overall picture which seems to emerge. So a cumulative approach to these factors might have been one way to preserve the relevance of a factual combination which had influenced the DPP in the James case. But the more significant problem is that no guidance remains in the Final Policy as to how the DPP would determine the presence of the two most apparently51 weighty anti-prosecution factors, namely that V made a "voluntary, clear, settled and informed decision" to commit suicide and that D was motivated to assist "wholly through compassion". Perhaps a previous suicide attempt by V, supposedly irrelevant in itself, will nonetheless assist in determining that V had made a "settled and clear" decision. It seems especially likely that the closeness of the relationship between D and V, and the nature of V's irreversible physical condition, each supposedly irrelevant in themselves, will help to satisfy a prosecutor that D acted through compassion. One may even posit that a prosecutor would only readily accept that D acted wholly through compassion for someone who was seriously ill or dying (does not the word "compassion" entail relief from suffering?). At this point we notice that whilst these factors are excluded from the list of "anti-prosecution" factors in the Final Policy, nothing is actually said about their supposed irrelevance either. So their presence may still be thought to be relevant when applying the main tests of a "voluntary, clear, settled and informed decision by V" and "acting wholly through compassion by D". Alternatively, it seems to be open to prosecutors to consider them directly as miscellaneous public interest factors, since the list of such factors in the Final policy is said to be "not exhaustive".52

Thus the Final Policy is not as clear as it might have been. It may be said that the Policy is aimed at all sections of the public and that its governing principles should be readily comprehensible to all; and that to say (for example) that "V's terminal illness is not relevant in itself but may be evidence of D's compassionate motivation" would be too confusing for most laypersons. But if that is so, then we can see that it is very difficult for prosecutorial "public interest" guidelines to be clear and accessible, because clarity may tend to come at the price of concealing important nuances (thus undermining accessibility). It is a feature of discretionary judgments that they are exercised within a relatively sophisticated framework, where factors are relevant for some purposes but not others, or relevant not in themselves but only as evidence of something else, and so on. If these points are thought to be too difficult to explain then the purposes of a prosecutorial policy cannot satisfactorily be met. We may recall that Lady Hale advised that "the factors which tell for and against such a genuine exercise of autonomy free from pressure will be the most important". It can hardly be said that this advice has been reflected by a single sentence that it is an "anti-prosecution" factor that V made a "voluntary, clear, settled and informed decision" to commit suicide, without further definition or explanation.

*Crim. L.R. 553 It may be thought, contrary to the contention here, that many lay persons do understand the import of the Policy. Mrs Purdy herself is said to have welcomed the emphasis on D's compassionate motives in the Final Policy (though significantly, she still urges legislative change to the scope of the offence as the best way of providing clarity).53 But if Mrs Purdy now believes that the way is clear for her husband to assist her when the dreaded hour approaches, as seems to be the case, then this brings us to the fundamental problem addressed in this article. Is it proper for the prosecutor ever to be heard to suggest that certain breaches of the criminal law will generically not be prosecuted? The necessary implication from the Policy is that if (albeit rarely) all the anti-prosecution factors are present and all the "pro-prosecution factors" are absent in any case, then there will be no prosecution. This is not a matter over which the DPP can be criticised by promulgating his Final Policy, since he had tried to resist having to produce it. So we now turn to criticise the decision of their Lordships in Purdy to require him to produce such a policy.

3. The constitutional criticisms of Purdy

In this section we will outline the main criticism of the decision in Purdy, namely that promulgating a policy which suggests that non-prosecution may be in the public interest primarily on the basis of the circumstances of the offence itself offends traditional notions of the rule of law. Moreover, all the constitutional ills of such policies--i.e. their limitations in addressing broader issues in the substantive law and the potential complications in applying and adapting their contents--are present in this case. Finally, it is wrong to think that anything of the sort may have been required by interpretation of the passage in Pretty v United Kingdom where the European Court noted that it was important to the overall compatibility of the offence of assisted suicide with art.8 that we should have "a system of enforcement and adjudication which allows due regard to be given in each particular case to the public interest in bringing a prosecution".54

The rule of law--the basics

The DPP is generally expected to apply the law of the land, as it is understood to be.55 This does not mean that he is expected to prosecute every provable case. In exercising his discretion to discontinue in the public interest, he may take account of such factors as the welfare of the victim, the situation of the defendant or simply his own departmental resources, if the offence is minor and unlikely to be repeated.56 But he may not discontinue on the basis that the law in question is generally not worth enforcing. There is a thin line to be drawn here--because when the prosecutor discontinues on the basis that his resources are too stretched, that may seem to be another way of saying that the law is not worth enforcing. But it is not quite the same, provided that the reference to resources is genuine and not a cover for fearing *Crim. L.R. 554 unwarranted criticism over bringing a trivial or controversial prosecution. Even when resources are invoked, it is important that they should not be expected to be used indefinitely as a reason for not prosecuting.57 Further, even if the DPP does put the prosecution of the less serious instances of an offence "on the back burner", he would still not be entitled to refuse to consider a prosecution within the class, whatever the individual circumstances.58 This would seem to mean that it was already not wholly unsatisfactory that the police and the DPP should have been apparently routinely discontinuing compelling cases of assisted suicide such as that of the family of Daniel James.59 Indeed, the fact that the DPP had chosen to go public with his reasons for not prosecuting the James case suggests that he wished it to be understood that he would not apply the law in other suchcases--albeit without explicitly saying so. His reason for resisting the Purdy litigation can only have been that he desired to resist having to do something which would expose him to the criticism that he was openly (though indirectly) rewriting the law.60 Indeed, there was an Early Day Motion signed by 63 MPs which was laid before Parliament on December 1, 2009, calling upon the DPP to withdraw the Interim Policy on the ground, inter alia, that it overrode the will of Parliament.61 The fact that the law would be effectively rewritten for some classes of cases should Mrs Purdy succeed, notwithstanding all the caveats that might be given in any resulting policy about each case being considered on its own facts and about there being no guarantees of non-prosecution,62 had been recognised in the Court of Appeal. The Lord Chief Justice had said that:

"Without giving what in reality would amount either to immunity from prosecution or the promulgation of a policy which would effectively discount the risk of a prosecution in this particular case (which it is accepted cannotbe provided) Mrs Purdy cannot achieve her true objective."63

But the House of Lords in Purdy says nothing of the rule of law at all. So if the traditional notion of the rule of law is altered, we are not told to what extent this should be so. However, their decision reminds us of our obligation to consider exactly what the rule of law objections might be, at least in cases such as this, where many would agree that the substantive law is very widely drawn and where Parliament has seemed to rely on extensive use of prosecutorial discretion64 (for in this case, it seems that objections based on the "sovereignty of Parliament" are relatively weak). But it seems that even here, constitutional difficulties remain.

*Crim. L.R. 555 Constitutional difficulties

One drawback is the limitations of a prosecutorial policy in applying coherently to the substantive law as a whole. The DPP's Final Policy is restricted to cases of assisted suicide. But some cases hover ambiguously on the borderline between assisted suicide and murder, for example if D opens V's mouth for her (which V cannot do) and puts in the overdose of pills, which V then voluntarily swallows.65 But since the DPP is not required in Purdy to announce how he will prosecute murder cases involving willing and autonomous victims, nothing is said on this subject--even though in some cases it may be the subject on which families desire assistance most urgently of all. To add to their desperation, it has since transpired in a case involving one Mrs Gilderdale that if the DPP takes the view that such an allegation does amount in law to murder or at least attempted murder, then he will not apply his assisted suicide policy, even if all the anti-prosecution factors in his policy would seem to have been applicable.66 Other problematic scenarios in our common law of homicide offences can readily be imagined. Suppose that D does nothing to call for medical help for V after V has just attempted suicide, in circumstances where a duty of care might arise for him to seek medical help despite the unwillingness of V to receive treatment,67 and with fatal consequences. In a case like this, it would seem that gross negligence manslaughter is arguably committed but not the (amended) offence of assisted suicide68 and so, again, the Final Policy on assisted suicide should not apply. But if D had actively assisted V's suicide in the first place (instead of merely passively refraining from calling help after the event) then the Final Policy would apply, quite possibly with the expectation that no prosecution would be brought. This would be anomalous and indefensible.

These examples show that a wholesale legislative revision of the laws appertaining to euthanasia is rather what is needed. They also remind us of the limits of human rights law in reshaping the substantive criminal law.69 Here, the problems are caused by the fact that the offence of assisted suicide engages the right to respect to private life of suicidal persons, but it is unlikely that we would say the same of the law of murder or manslaughter. So the Final Policy only applies to the former offence and thus aggravates existing anomalies in the substantive law.

But perhaps the more problematic point from the constitutional perspective is the legal effect of the policy which their Lordships required the DPP to promulgate. It seems to be taken for granted that if the DPP were to misapply his Final Policy, *Crim. L.R. 556 or to draw irrational conclusions on the absence or presence of the relevant factors, then D could seek judicial review of his decision to prosecute him. Alternatively he may seek to stay the proceedings as an abuse of process on the same ground when his trial is due to commence.70 But to what extent will a court be prepared to review the DPP's application of his policy? If it is regarded as a straightforward judicial review case, then one would expect a Wednesbury71 level of review, which would be unfavourable to D. It would likely be argued that a more intrusive "super-Wednesbury" standard should apply, since the DPP's Policy exists to give effect to an art.8 right.72 But this runs into the argument that the art.8 right is not possessed by D himself.73 It is not clear whether the rights of V can be extended so that D (say, her husband) can also demand a higher level of scrutiny should he claim that his subsequent prosecution is unreasonable. It is submitted that the courts would apply the lower Wednesbury test to determining any review of the DPP's exercise of discretion. They would hold that the need to prosecute assisted suicide cases involving potential coercion or undue persuasion should take priority; and if the prosecutor had made an inaccurate (but bona fide) assessment of the facts that would not withstand close scrutiny, then D should have to be content to raise his version before the sentencing judge in mitigation.

Moreover, it is not clear that the DPP would not be able to change his policy.74 Indeed, should there be discovered evidence that his current policy is being abused, and that some less than willing dying people are being persuaded to travel to Switzerland, he is likely to come under pressure to do so; or he (or his successor) may independently judge this to be appropriate. But if he were to change his policy, then what should he do in respect of offences that have been committed already and which may be affected by the content of that change?

One argument is that such defendants may have a substantive75 legitimate expectation that their case will be considered on the more favourable criteria in the "old" policy. But not every policy creates legitimate expectations as to its continuance, and the courts would only tend to find such an expectation to have been legitimate if an explicit promise as to its duration were made and would (preferably) affect only a small class of persons.76 It surely cannot be supposed that the DPP's policy is expected to remain unchanged forever, and besides the number of persons affected is potentially quite large. Further, to allow such an *Crim. L.R. 557 expectation in cases of prosecutorial discretion creates special difficulties. An offence may be committed many years before it is detected, and it seems hard to imagine that an old policy would still apply to this case alone, many years later, even if the defendant had clearly relied on its contents at the time. By analogy, when sentencing guidelines change there is no suggestion of a transitory period whereby the old regime should apply to offences already committed,77 and it is not obvious why prosecutions should follow a different path.78 In all these cases we are dealing with defendants who knowingly committed an offence the nature of which the prosecuting or sentencing authorities later deem to be more serious than they had originally thought, and it is submitted that the courts would hold that the effect of departure in such cases is not unjust.

It is conceivable that a court which was asked to review the application of the DPP's Policy (or a change to it) might have regard to the sui generis nature of the Policy and hold that, in view of its history, including Purdy itself, it was intended to give effective rights to defendants which may be more generous than would normally apply in ordinary public law doctrine. It may even be reasoned that the Final Policy was to be treated as a de facto substitute for legislative change, and that since D would have been able to rely on a statutory defence that was in force at the time of his conduct, even if subsequently revoked, so too should he be able to rely on this particular Policy. But this is highly speculative.

The Strasbourg justification?

The final argument that may be made in defence of the decision in Purdy is that the legitimacy of the Final Policy may be gainsaid by interpretation of the decision of the European Court in Pretty v United Kingdom. It will be recalled that the Court had held that our wide offence of assisted suicide was justified in order to protect the rights of the vulnerable partly because "consent is needed from the DPP to bring a prosecution and … lesser penalties to be imposed as appropriate [upon conviction]".79 From that passage one might reason that the Court would approve of our DPP discontinuing such cases as that of the family of Daniel James; and by extension, would it not prefer that such decisions were made on publicly accessible criteria (assuming, possibly dubiously,80 that they can be spelt out in an accessible way)? Some such reasoning, which we might call "the Strasbourg justification", seems to have held some appeal for Lord Brown.81 It might be *Crim. L.R. 558 buttressed by recalling that in most or even all other cases, defendants can reasonably be expected to assess the likelihood of prosecution by reference to the "general" Code.82

However, there are several reasons to doubt that the European Court had had anything of the sort in mind. The Court had said in Pretty v United Kingdom that "strong arguments based on the rule of law could be raised against any claim by the executive to exempt individuals or classes of individuals from the operation of the law".83 Whilst this was said in the context of granting proleptic immunity from prosecution, it is optimistic to infer from these words that the court would be much happier with a requirement that the DPP should be systematically reducing the law of assisted suicide by any policy of non-prosecution in some cases. It might also be mentioned that many European countries--formally, at least--recognise a principle of legality which does not permit prosecutors to drop cases by reference to the public interest,84 which makes it less likely still that any policy permitting routine discontinuance had been contemplated in Strasbourg.

But most of all, the "Strasbourg justification" is founded on a misreading of Pretty. The European Court expected us to ameliorate the width of our law on assisted suicide by mixing a strategy of non-enforcement with prosecution followed by sentencing flexibility upon conviction. It refers to both possibilities in English law with no suggestion that one should be used to the exclusion of the other, and it later calls for a "system of enforcement and adjudication".85 Indeed, such an approach is exactly what prosecutors ordinarily expect to follow in cases where the substantive law may appear harsh (but clear). Thus the DPP personally authorised the prosecution of Mrs Gilderdale for attempted murder and he stuck by this decision even though it must have been very obvious that at all events Mrs Gilderdale would not face imprisonment. The decision to prosecute Mrs Gilderdale is not criticised, however: for it is the right approach as traditionally understood. The rule of law is not undermined by leniency in sentencing, and the institution of prosecution in such a case serves to communicate that the law remains unchanged. What is being criticised here is that their Lordships in Purdy did not advert to the nature of such a mixture of strategies, and did not consider that this is what was required in Pretty v United Kingdom.

But how should these alternative state responses (fairly) be mixed? Certainly it would not be satisfactory if the DPP were arbitrarily to decide to prosecute only some minor cases (trusting judges not to imprison the defendant86). Consideration of this is the heart of the Purdy problem. It leads to my argument that the correct *Crim. L.R. 559 solution to the problems arising in Purdy was the promulgation of an offence-specific system that may offer reassurance to those in the position of Mrs Purdy that their families would not be prosecuted arbitrarily (if at all).

4. Dealing with assisted suicide by a prosecutorial system

In this section I present my own proposals as to how the Purdy case should have been decided. The argument is that when reading the passage in Pretty where the European Court had said that our offence of assisted suicide was compatible with art.8 because we also had a "system of enforcement and adjudication", we should pay more attention to the word "system". A "system" of prosecutorial discretion would allow the DPP to change his approach periodically for good strategic reasons, but would ensure that no decision to prosecute would be arbitrary or oppressive. A key feature would be that he would have to engage actively with the suspect during the decision-making process.

The first point to make is that it should have been decided in Purdy that the "state interference" in Mrs Purdy's private life was the mere existence of the criminal offence of assisted suicide.87 Accordingly, if the scope of that offence is clear (which it is) then it is already "in accordance with the law" for the purposes of art.8(2). This means that, when a criminal offence is in issue, the words "in accordance with the law" in art.8(2) cover no more than the same ground as in art.7 of the Convention. But this apparent overlap is not at all problematic. Article 7 is meant to ensure that all criminal offences are reasonably clearly defined, regardless of the subject-matter and regardless whether another Convention right is engaged. Article 8(2), for its part, deals with all state interferences with private life (among which the enactment of a relevant criminal offence is just one possibility) in order to ensure that they are all clearly accessible to potentially affected citizens. So there is no basis for expecting that the words "in accordance with the law" in art.8(2) would "add" anything extra to the protection in art.7. It may be that where the substantive law itself is uncertain, prosecutors may undertake certain responsibilities to avert its effects, perhaps by preferring alternative charges where the law is relatively certain.88 But there is no previous authority, either domestically or in Strasbourg, to suggest that they incur any extra responsibilities where the substantive law is clear.

There was another and better way to justify granting a remedy to Mrs Purdy. As we have seen, the European Court in Pretty had held that our offence of assisted suicide could be justified for the protection of the rights of the vulnerable assuming that the state has a "system of enforcement and adjudication" to take account of cases involving fully autonomous victims. So having an offence-specific system should be regarded as part of the justification element in art.8(2) rather than as part of any quality of law test. It is submitted here that there are two essential facets to any such "system". First, there needs to be not only a consistent way determining the autonomy of the victim but also a coherent strategy for deciding when to *Crim. L.R. 560 prosecute cases involving fully autonomous victims, bearing in mind that routine and indefinite discontinuance would violate the rule of law. Observance of this point should remind us that suspects truly should commit the offence at their own risk. But secondly there does need to be a way of engaging suspects with the decision-making process so that they might be expected to accept any decision to prosecute them as being fairly reached. We now examine both these facets.

Determining the victim's autonomy and devising a strategy of prosecution

The first matter should be to devise a consistent approach to the determination of the victim's autonomy. One possibility that we have already considered is that, besides medical evidence, a combination of circumstantial factors such as the closeness between the parties, the terminal state of the victim and the victim's previous attempts to commit suicide unaided may suffice.89 This approach should not need to be published as a matter of practice. The prosecutor must be persuaded by the defendant of the autonomy of the victim, and otherwise to treat the matter as not established (meaning that prosecution will then almost inevitably follow). This is because the logic of the blanket ban on assisted suicide posits that families are not necessarily well placed to adjudge the autonomy of their loved ones90; and observance of the reasons for the law should thus dictate a precautionary approach on matters of proof.91 Moreover, a family member who does not accept that message and hopes to appeal to prosecutorial discretion may be thought to undertake the responsibility92 of having to prove his or her case on the autonomy of the victim. It would also be advisable for all such evidence to be assessed by specially appointed senior prosecutors (as opposed to any prosecutor in any of the 42 different branches of the CPS).93 This should mean in turn that all decisions in this unusual area may readily be compared with each other in order to ensure consistency.94 It should be noted that at this stage, having a "policy" of sorts would be helpful, but it would be primarily for the purpose of consistent fact-finding, mainly for internal use and is only part of a wider set of measures.

A strategy of prosecution would deal with the difficult subset of cases where the autonomy and willingness to die of the victim is established to the satisfaction of the prosecutor. We have already said that any such strategy needs to be more nuanced than routine discontinuance. One acceptable strategy would be presumptively to prosecute all such cases, in order to communicate the message behind the blanket prohibition in the Suicide Act 1961, as it was enunciated by Lord Bingham--namely that, even with the best will in the world, family members may wrongly assess the autonomy of their dying member or underestimate the *Crim. L.R. 561 constraints that he or she may perceive.95 But such a strategy would be subject to two provisos. The first proviso might be that the prosecutor should offer to accept the fact of the victim's autonomy for sentencing purposes so that the defendant can be reasonably assured of a non-custodial sentence, thus alleviating to some extent the defendant's stress of being prosecuted. Secondly, and more importantly, he should be prepared to discontinue those cases where the very process of undergoing prosecution will cause disproportionate harm to the accused, even if he were to offer to recognise the autonomy of the victim for sentencing purposes.96 Generally speaking, prosecutors must always have regard to the "harms of prosecution" (as distinct from the prospective "harms of punishment")97 and the "general" Code for Crown Prosecutors suggests that prosecutors might take into account the old age of the defendant or any "significant mental or physical ill health" as a reason not to prosecute.98 But it would be open to the DPP to list further"harms of prosecution" in cases of assisted suicide involving autonomous victims, such as stress arising from publicity and invasions of privacy, and financial hardship from paying legal fees. Additionally, the facts that the family members may be grief stricken and/or still partially reproaching themselves--rationally or irrationally--for what they have done may also be thought to make the harms of prosecution that much harder to bear. To some extent, this is how the DPP reasoned when he decided not to prosecute the family of Daniel James.99 So here, the strategy of prosecution in minor cases (involving autonomous victims) would be subject to a sensitive and informed appraisal of the likely effects upon the family.

But importantly this should not be the only possible strategy which mixes prosecution with non-prosecution.100 The DPP might alternatively decide in the immediate future to prosecute no case where the autonomy of the victim and his willingness to die are established. However, this decision should be reviewed periodically with an open mind. He may wish to reflect public concern over assisted suicide should it change direction, if for example there were to be plausible stories in the media about Dignitas clinics abroad adopting very casual assessments of the patients' physical and mental health. Similarly, if there were to be legitimate public debate over some cases where he had discontinued a prosecution, the DPP might become concerned that unscrupulous families would try to take advantage of what might be perceived as a weakness in the system. Prosecuting even minor cases for general deterrent purposes might start to make some sense.101 It is submitted that such a strategy would be flexible and would be better capable of commanding public confidence over a long period of time than the current policy which is not said to be subject to regular review. It certainly meets the requirements of the rule of law, providing that the initial decision not to prosecute minor cases *Crim. L.R. 562 truly is subject to review, for in such a case prosecutions would merely "be put on the back burner" for a period. Even so, any subsequent change in strategy to pursue all cases of assisted suicide should still be subject to a humane consideration of the effects of prosecution upon the individual defendant.

It will be noted that little has been said about the compassionate motivation of the defendant in this section. This is because it is the evidence, and the relevance, of the victim's autonomy which should be crucial, bearing in mind the justification for the width of the substantive law. But many defendants who did act compassionately should typically not be prosecuted--whatever the applicable strategy at any time--if it were thought inhumane to expose them to the strains of the prosecutorial process. Compassion then may be relevant but it should only be recognised indirectly as a humanitarian consideration, and as such it should only come into play when the autonomy of the victim has been satisfactorily determined.

A system as a mode of engagement with the defendant

The other crucial facet of the proposed notion of a prosecutorial system, as opposed to a policy, is that the suspect should be actively included in the decision-making process (either in person or via his legal team). Whilst there appears to be no direct authority in Strasbourg jurisprudence on what a prosecutorial system should consist of, a useful analogy may be drawn with the decision in Buckley v United Kingdom102 where a local authority wished to justify its relocation of a gypsy site. It was accepted that its action was "in accordance with law", but in determining whether the individual decision could be said to have been necessary for the purpose of enforcing planning controls, the court noted that:

"Indeed it is settled case-law that, whilst Article 8 contains no explicit procedural requirements, the decision-making process leading to measures of interference must be fair and such as to afford due respect to the interests safeguarded to the individual by Article 8."103

Moreover, the court found that there was no violation in this case precisely because there was a working system of appeals, which included a number of personal visits to the applicant's home in order to best assess her living conditions as well as the likely development needs of the immediate area. It is possible that state interferences with art.8, which protects "respect" for one's home and private life, may be especially subject to such personal attention by decision-makers. To "respect" someone's right may involve communication and engagement with them about the decision, as well as simply making a decision which does in fact take into account all the material factors.104

This is not how prosecutions generally work in England; indeed, the absence of any general duty to explain a decision to a prosecuted person is often criticised.105 But it would not be a wide departure from that norm to accept that such engagement may be required in the comparatively rare cases when an art.8 right is engaged. *Crim. L.R. 563 The suspect, after being interviewed by the police, might (absent any risk of absconding) expect to be bailed before a decision is made whether to charge him. At that stage he should be invited to supply any evidence that he might have to the prosecutor on the competence of the victim and her willingness to die, and also any evidence of his own emotional well-being if he wishes to argue separately that he should not be prosecuted for humanitarian reasons. This means that the prosecutor should be working from an early stage from evidence from the defendant as well as any evidence gathered solely by the police. The DPP's Final Policy rightly advises suspects that co-operation with the police in investigating the suicide is itself an anti-prosecution factor,106 but it is possible for the authorities to invite him to produce the best evidence at his disposal, potentially for his own sake.107

Further, if he does eventually decide to prosecute under some such "system", the prosecutor ought to outline whether he accepts any evidence presented to him by the defendant as well as his reasons for deciding to proceed. This may involve explaining how his decision fitted in with his general strategy in comparable assisted suicide cases. It is suggested that explaining a decision in strategic terms may make sense to a layman, even if he does not accept the decision as it affects him personally. By contrast, detailed policies, of the type mandated in Purdy, may be less amenable to gaining respect and acceptance from lay persons. In cases where there is a list of pro-prosecution factors as well as anti-prosecution factors, explaining a borderline decision to prosecute would seem to be especially difficult, and if there has not been appropriate emphasis in including the family in the decision-making process, the reasons may never gain acceptance.

Moreover, if the DPP were to announce a special system for dealing with assisted suicide cases which involved meeting the defendant and his legal team and explaining his decision, then any person in Mrs Purdy's position should feel assured that any decision to prosecute a family member will not be taken arbitrarily. This, in fact, is what seems to have worried Mrs Purdy most of all. In a newspaper interview which preceded the hearing before their Lordships, Mrs Purdy had said that she worried that the potential case of her husband, who only moved from Cuba to England when he met her, might be viewed differently from other potential defendants in comparable cases on account of his immigration status.108 Further, if this were to happen, then she feared that there would be no other close relative of the family to help him through the trauma of prosecution in this country.109 But fears over discriminatory prosecutions ought to recede if the DPP has a clearly defined strategy which he is prepared to discuss with affected defendants. It should remain possible to plead abuse of process should there be evidence that the DPP is motivated to prosecute by other factors, though the law on discriminatory *Crim. L.R. 564 prosecution is relatively undeveloped in England.110 But the operation of the system in practice would at least be scrutinised in court, and it is submitted that the defendant would air his grievances more effectively than if his counsel were restricted to a doomed attack on the prosecutor's application of his policy document under principles of Wednesbury reasonableness.

Conclusions

It is unprecedented that prosecutorial guidelines relating to enforcement should be required in order to render the law accessible and clear. It is not even clear to what extent a nuanced structure of reasoning can be made to be accessible. The real problem in Purdy was that the applicant and others in her position feel uncomfortable about being at the mercy of prosecutorial discretion. But the solution to that was not to require an offence-specific policy, at least not if that is the sole solution. Instead, the proper approach should be to require a prosecutorial system in assisted suicide cases, not in order to make the offence "accord with law" (which is quite unnecessary) but in order to render its effects proportionate to the need to protect the vulnerable. Under such a system, the DPP would have a choice of prosecution strategies, which may vary over time. He might accordingly prosecute some cases in which he was satisfied of the victim's autonomy; but in such cases, he should offer to accept the fact of the victim's autonomy for sentencing purposes, and be prepared to discontinue if the defendant seems truly unable to cope with the prosecution process. Provided that any such strategy is consistently applied until such time as it might be changed, there is no legitimate ground for complaint. But the prosecutor may have to explain his strategy to the suspect as well as its application in the instant case. One hopes that such a system would satisfy the victim who contemplates asking for assistance, but that is not the crucial point. A policy which suggests routine discontinuance is impermissible under closer examination of the rule of law and was not anticipated as a safeguard in Pretty v United Kingdom.

No doubt passing legislation would be the most principled way in which to resolve issues relating to euthanasia. The prosecution of minors for consensual sexual activity in the absence of any obvious sign of coercion is another difficult areas in which the legislature is fearful to tread. But until Parliament seizes the day, the DPP might consider prosecutorial systems (with consistent review of strategy, and active engagement with defendants) as offering the best way of preserving his future options whilst offering greater reassurance against fears of arbitrary prosecution.

I am grateful to the journal's reviewers for their comments. As ever, the final contents are my own responsibility.

Crim. L.R. 2010, 7, 543-564

1.

R. (on the application of Purdy) v DPP [2009] UKHL 45; [2009] 3 W.L.R. 403.

2.

Swiss Penal Code 1942 art.115.

3.

The same is said to be true of others in the same position: Purdy at [31].

4.

The Suicide Act 1961 s.2 has since been amended by s.59 of the Coroners and Justice Act 2009, which came into force on February 1, 2010. The effect is that "assistance" and "encouragement" of suicide are now the only modes of liability, but it now only needs to be shown that D has done an act which is capable of assisting or encouraging suicide, with the intention of assisting it. It need not be shown that D's act did in fact assist or encourage V to commit suicide, and the purpose of the amendment is to facilitate the prosecution of those who advise on methods of committing suicide via the internet. The offence in s.2 shall continue to be referred to as "assisted suicide" in this article, even though "encouraging or assisting suicide" may become the preferred label for the amended offence.

5.

The clarity of s.2(1) of the Suicide Act 1961 was accepted by Lord Hope in Purdy [2009] UKHL 45; [2009] 3 W.L.R. 403 at [18], [41], and see too the Lord Chief Justice in the same case in the Court of Appeal [2009] EWCA