Book Review: Justice for Hedgehogs by Ronald Dworkin

Even lawyers who are not interested in ethics and morals should be fascinated by the brilliant book Justice for Hedgehogs (2011) by Ronald Dworkin.

His treatment of parliamentary sovereignty points the way for twenty-first century policy in the light of law as an interpretive concept. This contrasts to Lord Neuberger MR's recent conservative treatment of sovereignty which may leave us wondering why history stopped after the civil war, the Glorious Revolution of 1688, the Bill of Rights 1689, and the Act of Settlement 1701. Has nothing happened since? Wasn't there a Holocaust? Do people not need protection against abuses by majorities? What if Parliament abolished judicial review or the ordinary role of the courts? Neuberger refuses to test the need for change against hard cases, saying that such arguments are unreal, [1] they involve "postulating a wholly different Parliament from that which we have ever known" and – I sarcastically interpose - everyone in Northern Ireland would cheer to that, but he adds "and, if that arose, there would presumably be a very different judiciary from that which we have ever known." Amazingly, he continues "Further, a Parliament which was prepared to prevent citizens having access to the courts would presumably be unconcerned about the rule of law, in which case questions of constitutional sovereignty would be of no real significance in practice." Well, no, not if the judges sat on their hands. The issue would have to be fought out in the streets.

For those who dislike the idea of revolution as the only answer to a moral crisis of that magnitude, Dworkin's argument is compelling. His thinking has developed since (brace yourself, I'm going to use the cliché seminal) his seminal "Taking Rights Seriously" (1977). He now sees law as a branch of politics, rather than as a system separate from but interacting with morality. The change is because Dworkin has recognised the circularity of treating law and morality as separate but interrelated systems. How would they interrelate? If that is a legal question, the answer depends on an assumption about the role of morality in reading legal material; the alternative is to assume it is a moral question. This logical difficulty, says Dworkin, is what led to the "concept of law" jurisprudence, which treated the relationship between law and morality as neither a legal nor a moral question, but as a "conceptual" one. Here the problem is that people don't agree on which propositions of law are true. That, however, is unavoidable, and Dworkin looks to an integrated network of political value to construct a conception of law. Ethics produces personal morality which produces, as a subdivision, political morality which produces, as a subdivision, law. There may be, therefore, some valid laws that are too unjust to enforce.

But I have focused here on his final chapter, while on the way Dworkin has discussed a huge range of fascinating ethical and moral questions. His aim is to demonstrate how we can coherently think about questions of values. The integration of our understanding of interpretive concepts produces a unity of values. Interpretive concepts are those about which we share a general understanding, but we may disagree about their application in particular cases. Justice, right or wrong, helping, harming, keeping promises, obligations arising from relationships, equality, liberty, democracy, and law are all interpretive. The truth of an interpretation is determined by its consistency with the two principles of dignity. The first principle of dignity is that the success of one's own life has objective importance, and, as a corollary, all lives are equally important. The second principle of dignity is that we each have personal responsibility for our own life; we must each make our own decisions and so lead authentic lives. Self-respect comes from taking our own life seriously and in that sense living well: we must strive to make our own life a successful performance by creating value in it. The two principles of dignity interact to guide ethics (how to live well) and morality (how to treat others), which in turn are mutually reinforcing: living well requires treating others well, and respect for others enhances self-respect.

That is an outline of the skeleton of Dworkin's conception of the unity of value. There have been times when, as a lawyer, I have wished that the people who write complicated statutes would reveal the whiteboard diagram which they devised to guide their drafting. So too with this book. It has to be worked at rather than skim-read. Notes have to be scribbled in margins and end pages filled with references. Yes, the first chapter serves as a travel guide to the journey ahead, but there are times when Part One (Chapters 2 to 5) slows down while Dworkin deals with anticipated challenges from academic philosophers, requiring detailed conceptual distinctions on issues that most of us, since we have bought the book, will be prepared to take on trust just for the sake of getting to his point. But it would be wrong to be discouraged by the initial labours, as once we have appreciated the book as a whole we can go back to the details. Some readers might think a glossary would have been useful, but the index is a well designed work-around.

For an indication of what you will find in Justice for Hedgehogs, here is Dworkin's essay from a recent issue of the New York Review of Books (10 February, 2011).

This wonderful book will delight anyone who wants a systematic approach to deciding what is right. It would be a dull person indeed who did not feel the resonance of Dworkin's conclusion:

"Without dignity our lives are only blinks of duration. But if we manage to lead a good life well, we create something more. We write a subscript to our mortality. We make our lives tiny diamonds in the cosmic sands."

[Update] For a critique of Dworkin’s early work on rights and principles, see Roger Scruton, Fools, Frauds and Firebrands – Thinkers of the New Left (2015), Chapter 3. Of Dworkin’s style of argument, he says:

 “ … While purporting to give a general theory of law, Dworkin’s real interest is one of advocacy, on behalf of a political position towards which, in the conservative view of things, the law is at most neutral, and in some respects deeply opposed.” 

 “… he assumed that it was never he but always his opponent who bore the onus of proof. For Dworkin, as for the writers for the New York Review of Books generally, the left-liberal position was so obviously right that it was for the conservative to refute it.”

 “… For a conservative it is a matter of common sense that constant liberalization, constant remaking of law in the image of the elite lifestyles of New York, may eventually threaten the community with harm.”

 In contrast to Dworkin’s principles, Scruton places the common law:

 “…the knowledge that we need in the unforeseeable circumstances of human life … is bequeathed to us by customs, institutions and habits of thought that have shaped themselves over generations, through the trials and errors of people many of whom have perished in the course of acquiring it. Such is the knowledge contained in the common law, which is a social bequest that could never be adequately replaced by a doctrine, a plan or a constitution, however entrenched that constitution may be in a vision of individual rights. … The common law is concerned to do justice in the individual case, not to pursue some far-reaching reform of the manners, morals and customs of the community as a whole.”

 Whereas Dworkin sees law as emerging from a political morality that is part of personal morality which in turn is a product of ethics, Scruton’s view is more recognisable:

 “…the common law of the English-speaking people … [has] been in existence for a thousand years, with precedents from the twelfth century still authoritative in our twenty-first-century courts. It has developed according to an internal logic of its own, maintaining continuity in the midst of change and welding English society together through all national and international emergencies. It has shown itself to be the motor of history and the initiator of economic change ….” 


The opposing camps – conservatives and new leftists (to follow the terminology of Scruton’s title) – have set up conceptual frameworks for working out their positions on issues. Are the positions fixed for all time? I doubt that conservatives now oppose women having the vote. If agreement can come with time, are values relative to social conditions? Not all conservatives would, I surmise, now side with Devlin against Hart on the issue of the lawfulness of homosexual practices, although many might. Some conservatives may accept laws for abortion procedures, just as some new leftists may be against them. The methods for working out positions on contentious issues do not dictate answers. Dworkin may only support legalisation of homosexual practices because of the weights he gives to the values he recognises, and Scruton may take the opposite view because of his individualised weightings. A person may be conservative in some ways and new leftist in others. Answers to social issues are individualised, even where people approach them in the same way. A vote count, or a dictator, may determine the course society will take.

What to make of all this arguing

Can a government be forced to avoid making immoral laws? Who is to say what is moral? 

 

One way of settling this is by warfare. In the immediate aftermath of the Second World War the trials at Nuremburg imposed penalties on leaders who had acted according to the laws of the defeated government.

 

As a preferred alternative to warfare, is there a way, internal to the legal system under which it exists, of constraining the acts of a government?

 

To some extent, if there is a written constitution, government powers can be constrained, but only if the courts have the power to declare infringing legislation and governmental actions invalid.

 

Dworkin’s task was to find a way of extending the courts’ powers to hold immoral laws and governmental actions invalid.

 

The obstacle he faced, voiced by so-called conservatives, is the interpretation of history that recognises parliament as sovereign. This means that enacted laws are valid. There is no reasoning with such a position, because it takes sovereignty to be axiomatic.

 

Equally stubborn is Dworkin’s stance: laws are only valid if they are both properly enacted and morally right.

 

The difficulty with axioms is in proving that they are correct. You don’t have to be Gödel to realise that some things just can’t be proved using the methods of argument accepted within a system. Hence, warfare.

 

Dworkin can only attempt to persuade us that he is right. Many of his examples used to illustrate the unattractiveness of his opponents’ positions are appeals to emotion. But, the conservatives may say, just because an outcome is unattractive doesn’t mean it is illegal.

 

At its highest, Dworkin’s argument is that law and morality are not separate things. They are interpretive concepts.

 

Why, then, are people ordinarily punished for disobeying laws that they think are wrong? When does the majority view, expressed through the government, of what is a just law, become subject to the views of a small group, such as judges?

 

International laws try to provide an answer, through conventions and declarations. If governments accept these by ratification, they acknowledge that the final interpretation of their laws may be decided by international courts. But governments may refuse to ratify these conventions and declarations, or may “exit” from their jurisdiction.

 

The effort to get governments to do what they don’t want to do really comes down to political pressure, economic sanctions, and if all else fails, the horrifying disaster of warfare.

_____________________________

[1] Looking back on this from the future, it can be accepted that refusing to address difficult hypotheticals is not a wise tactic. The Supreme Court in Miller v Prime Minister [2019] UKSC 41 at [43] said that it was no answer to the points before it that the court should decline to consider extreme hypothetical examples.

Interpretation of unfettered powers

The granting by the legislature of unfettered powers on a court does not destroy all lawful restraint. In Lacey v Attorney-General of Queensland [2011] HCA 10 (7 April 2011) the High Court considered statutory interpretation, legislative purpose, jurisdiction, power, and the kinds of appeals that can exist.

The majority, French CJ, Gummow,Hayne, Crennan, Kiefel and Bell JJ, held that a provision saying an appellate court "may in its unfettered discretion vary the sentence and impose such sentence as to the Court seems proper" means that the appellate court must first be able to identify an error in the decision of the court below, and then the "unfettered discretion" gives the appellate court the power to do what it considers the lower court should have done [62].

The principle of legality is "the presumption that, in the absence of unmistakable and unambiguous language, the legislature has not intended to interfere with basic rights, freedoms or immunities" [43], and "judicial findings as to legislative intention are an expression of the constitutional relationship between the arms of government with respect to the making, interpretation and application of laws" [ibid].

The distinction between a power and the jurisdiction to exercise it is such that "The power given to the Court may inform the characterisation of its jurisdiction but does not necessarily define its content" [48].

Here the jurisdiction was indicated by the heading to the relevant provision, here simply the word "appeal". This encompasses the jurisdiction, and the question is, what kind of jurisdiction does it create? [56].

The majority set out the relevant classes of appeal for present purposes [57] (footnotes omitted):

"1. Appeal in the strict sense – in which the court has jurisdiction to determine whether the decision under appeal was or was not erroneous on the evidence and the law as it stood when the original decision was given. Unless the matter is remitted for rehearing, a court hearing an appeal in the strict sense can only give the decision which should have been given at first instance.

2. Appeal de novo – where the court hears the matter afresh, may hear it on fresh material and may overturn the decision appealed from regardless of error.

3. Appeal by way of rehearing – where the court conducts a rehearing on the materials before the primary judge in which it is authorised to determine whether the order that is the subject of the appeal is the result of some legal, factual or discretionary error. In some cases in an appeal by way of rehearing there will be a power to receive additional evidence. In some cases there will be a statutory indication that the powers may be exercised whether or not there was error at first instance."

Here the appeal was of the third kind [60] where the jurisdiction arose once the appellate court determined that an error had occurred in the court below.

Heydon J dissented, holding that the appellate court had first to decide whether it disagreed with the sentence imposed by the lower court to the extent that interference was merited, and if it did, then to vary the sentence in accordance with its perception of what was appropriate [94].

So the difference here was in practical terms about which sort of appeal was created by the legislation, and this difference was reflected in different use of legislative materials (here, speeches in Parliament). The majority noted [61] (footnote omitted):

"The Minister's words ... cannot be substituted for the text of the law, particularly where the Minister's intention, not expressed in the law, affects the liberty of the subject. In any event the Minister's Speech left open the question of the content to be given to the word "appeal" and thereby to the jurisdiction conferred upon the Court."

The last thing I want to do is to reveal my unsubtle mind, but I must be frank and admit I find the distinction between the first and third classes of appeal as described by the majority rather elusive, apart from the possible add-ons in the third. This, I trust, is the judges' fault, not mine. More clarity in future please.

Dangerous hypotheticals

One of the common judicial rear-protecting stratagems is to reason in the form "even if I'm wrong, I'm right."

That is, a decision on the basis of particular assumptions would be the same if those assumptions were wrong.

For example, if evidence is ruled admissible on the basis that there was no impropriety in the way it was obtained, a judge may say that even if the alleged impropriety had existed the evidence would still have been excluded.

I criticised that sort of reasoning a while ago: "Shaheed balancing: three propositions" [2004] New Zealand Law Journal 475. So did the Chief Justice last month, when with Blanchard, Tipping and McGrath JJ she delivered the reasons for granting leave to appeal in Hamed v R [2011] NZSC 27 (24 March 2011) at [15]:

"[15] Although the Court of Appeal indicated its view, in application of s 30 of the Evidence Act, that all the evidence should be admitted even if unlawfully obtained or obtained in breach of s 21 of the New Zealand Bill of Rights Act, we have some doubts as to whether it was in a position to express a view on the proportionality of exclusion. An assessment of this kind requires the gravity of the infringement to be carefully balanced against the benefits of admission of the evidence. Since the Court of Appeal took the view that there was no infringement, it was an unnecessary and perhaps artificial exercise for it to undertake the s 30 assessment in respect of the evidence it considered to have been lawfully obtained under s 198 warrants."

The misuse of hypotheticals is quite common. The Supreme Court did it itself in a passage that I have only now noticed: in Condon v R[2006] NZSC 62, an important decision on the status of the right to a fair trial, noted here in August 2006. In its penultimate paragraph the Court says [89]:

"After considering all of these matters, we have not been persuaded that the outcome of the trial would necessarily have been the same if Mr Condon had been legally represented. In our view there was therefore unfairness in the trial and accordingly a substantial miscarriage of justice has occurred."

The trouble here is that the result if there had been no error (ie if legal representation had been obtained), that is, if the trial had been fair, is irrelevant to the question of whether the trial had been fair. If the trial was unfair the conviction had to be quashed, regardless of the strength of the prosecution evidence. If an error at trial caused a real risk that the jury would not be impartial, that would render the trial unfair. A partial jury could nevertheless reach a correct verdict, but it would still be a partial jury delivering a verdict at an unfair trial.

On this point the Court had reasoned better in Sungsuwan v R [2005] NZSC 57, which was noted here in August 2005, and I observed that the majority of the Privy Council in Howse v R (New Zealand) [2005] UKPC 31 had fallaciously asked what course the trial would have taken if the relevant errors had not occurred. The way to avoid this artificial exercise is to ask, given that the errors occurred, what was their effect on the fairness of the trial.

In Sungsuwan the effect of the error on the verdict was a sufficient, but not a necessary, condition for a finding that the trial was unfair. It is one of the ways in which unfairness may be manifest. Another is where partiality existed although the verdict was inevitable. The sufficient condition should not be treated as if it was a necessary condition, as occurred in Matenga v R [2009] NZSC 18 at [31] footnote 39 where the Court defined a substantial miscarriage of justice as one which affects the result of the trial.

Abuse of process, stay of proceedings, and integrity

The inherent power of a court to stay criminal proceedings on grounds of abuse of process is a discretionary matter, not to be fettered by rigid classifications of kinds of official misconduct: Warren v Attorney General of the Bailiwick of Jersey (Court of Appeal of Jersey) [2011] UKPC 10 (28 March 2011).

There are, said Lord Dyson delivering the leading judgment, two categories of case in which a stay may be justified: (i) where a fair trial could not be held; (ii) where continuing the proceedings would offend the court's sense of justice and propriety [22, quoting R v Maxwell [2010] UKSC 48 at 13, a decision not yet published].

These should not be confused, and it is unhelpful to refer to fairness when discussing the second category (Lord Kerr at [84]).

The signal here [this is me now, not a Lord] is that what is often referred to as "public policy fairness" should now be called something like "public policy grounds" or "the public policy category".

The circumstances of each case are critical, and the classic cases - R v Looseley [2001] 1 WLR 2060; R v Latif [1996] 1 WLR 104; R v Horseferry Road Magistrates' Court, Ex p Bennett [1994] 1 AC 42 - must be read in this light. Lord Dyson added [26]:

"The Board recognises that, at any rate in abduction and entrapment cases, the court will generally conclude that the balance favours a stay. But rigid classifications are undesirable. It is clear from Latif and Mullen [[2000] QB 520] that the balance must always be struck between the public interest in ensuring that those who are accused of serious crimes should be tried and the competing public interest in ensuring that executive misconduct does not undermine public confidence in the criminal justice system and bring it into disrepute. It is true that in Bennett the need for a balancing exercise was not mentioned, but that is no doubt because the House of Lords considered that the balance obviously came down in favour of a stay on the facts of that case (the kidnapping of a New Zealand citizen to face trial in England)."

Any virtue that might have been thought to attach to Mr Bennett, through giving his name to a leading case on police misconduct, is dispelled in the present case by Lord Hope, who adds at [64-68] a detailed examination of the facts of that case.

Does the emphasis on the facts of each case erode the value of these cases as precedents? Does the balancing exercise leave the law unacceptably vague? On appeal the question will be whether the decision to refuse a stay was perverse or was one which no reasonable judge could have reached [51].

Judgments may tend to pull in two directions: to condemn the misconduct of officials while at the same time holding that a stay was not required. That was the position here. Looking at the misconduct from a (considerable) distance – the placing of audio surveillance equipment in a car without lawful authority and giving the impression that permission had been obtained – this was not a case of serious misconduct. But one would think it was, given the stern protestations of Lord Dyson [45-46], Lord Hope [61-62, 68], Lord Rodger [70-71], Lord Brown [78], and Lord Kerr [81-82]. There were circumstances that mitigated the wrongfulness, summarised by Lord Dyson at [47-50].

The Board found useful the summary of the balancing exercise appropriate to the second category of abuse of process cases (no balancing being appropriate where a trial would not be fair) given by Professor A L-T Choo in Abuse of Process and Judicial Stays of Criminal Proceedings, 2nd ed (2008), at p 132:

"The courts would appear to have left the matter at a general level, requiring a determination to be made in particular cases of whether the continuation of the proceedings would compromise the moral integrity of the criminal justice system to an unacceptable degree. Implicitly at least, this determination involves performing a 'balancing' test that takes into account such factors as the seriousness of any violation of the defendant's (or even a third party's) rights; whether the police have acted in bad faith or maliciously, or with an improper motive; whether the misconduct was committed in circumstances of urgency, emergency or necessity; the availability or otherwise of a direct sanction against the person(s) responsible for the misconduct; and the seriousness of the offence with which the defendant is charged."

Lord Kerr summarised [83] the principles that have emerged from recent jurisprudence. He emphasised that stays are not imposed to discipline the police, but instead are designed to protect the integrity of the criminal justice system.

The decision in this case may have been finely poised, but it is not surprising: given the extenuating circumstances and the relatively slight (compared to being kidnapped or entrapped) breach of the defendant's rights, and the fact that the offending was importation of cannabis (180 kg, worth over £1m), it is not unreasonable to conclude that the integrity of the justice system did not require a stay of proceedings.


No one likes to disagree with the Privy Council, but I have to doubt whether it is appropriate to address the need for a stay of proceedings as a balancing exercise. Either misconduct is sufficiently egregious to require a stay, or it is not. The balancing factors mentioned in this case are essentially those applicable to the discretion to exclude improperly obtained evidence. The balance is between factors relevant to the seriousness of the impropriety and factors relevant to the seriousness of the alleged offending. This similarity in reasoning raises the question of the relationship between exclusion and the stay of proceedings.

The stay is clearly the more extreme response, and factors justifying it can be put at the end of an imaginary continuum of impropriety, most of which could be a basis for exclusion of evidence unless the offending is too serious. The seriousness of the offence is properly part of the balancing exercise used to determine the admissibility of improperly obtained evidence. But when the level of impropriety is sufficiently high to take the potential remedy beyond exclusion of evidence to a stay of proceedings, is there a level of offending that would make a stay inappropriate? If there were, it would also have to make exclusion of the improperly obtained evidence inappropriate, otherwise where that evidence was central to the prosecution case the proceedings would terminate. Once evidence is ruled admissible after the balancing exercise, because the alleged offending outweighed the impropriety, an application for a stay would be futile. If, on the other hand, the evidence was excluded because the impropriety outweighed the seriousness of the offending, an application for a stay might in extreme cases be successful and if so, that would be because of the seriousness of the impropriety. 

There would be some cases where the impropriety, although sufficient to require exclusion of the improperly obtained evidence, was insufficiently serious to warrant the response of a stay. That would be because of the nature of the impropriety, not because of the seriousness of the alleged offending. But where the impropriety was sufficiently serious to justify a stay, it would be sufficient for that purpose no matter how serious the offending. The assessment of the seriousness of the impropriety would be an exercise of judgment, taking into account matters relevant to the impropriety. The judgment would be about whether in those circumstances a stay was necessary to uphold public confidence in the integrity of the judicial process. It would be wrong to say that a prosecution should proceed (because of the seriousness of the alleged offending) even though proceeding would compromise the integrity of the judicial process.

For a glance into the future, see Wilson v R [2015] NZSC 189 (14 December 2015), where the Court described the decision as to whether a stay was warranted in circumstances of investigatory impropriety as a balancing exercise, where the seriousness of the impropriety was balanced against the need to uphold public confidence in the integrity of the judicial process (at [60], Elias CJ dissenting on this point at [149]). I side with the Chief Justice on this.

The stay of proceedings can be used in wider contexts than improperly obtained evidence, and in those the seriousness of the alleged offending may well be a relevant factor. Delay, multiplicity of charges, re-litigation of decided issues, improper motive in charging or prosecutorial reneging on an agreement as to charge can all give rise to issues of the appropriateness of a stay, and the seriousness of the alleged offending could properly be a factor in the decision. 

Disobedient fact-finders

For an interesting article on the adverse effects of a rule (as opposed to a discretion) excluding improperly obtained evidence, and the ways in which fact-finders can be biased by knowledge of excluded evidence or by suspicion that evidence has been excluded, see Tonja Jacobi, "The Law and Economics of the Exclusionary Rule", available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1783863

The author cites studies that suggest juries do not obey judicial directions, and even go to the extent of counter-intuitively reasoning that "if I am innocent I may as well commit offences because if tried I would be likely to be convicted even if innocent." We need to know how valid are the studies that are cited in this paper.

Using relevance to prevent unfairness

Today's decision of the High Court of Australia in Stubley v Western Australia
[2011] HCA 7 is noteworthy on three points.

The amazing legislation in Western Australia governing the admissibility of propensity evidence.

Section 31A(2) of the Evidence Act 1906 (WA) provides:

"(2) Propensity evidence or relationship evidence is admissible in proceedings for an offence if the court considers —

(a) that the evidence would, either by itself or having regard to other evidence adduced or to be adduced, have significant probative value; and

(b) that the probative value of the evidence compared to the degree of risk of an unfair trial, is such that fair-minded people would think that the public interest in adducing all relevant evidence of guilt must have priority over the risk of an unfair trial."

It is para (b) that makes sensible people sit up and wonder. It posits a situation where fair minded people might think that something can be more important than the risk of an unfair trial. Normal people would think that when a risk of trial unfairness becomes unacceptably high, the risk must be avoided (by exclusion of the challenged evidence), regardless of what public interest there might be in adducing evidence of the defendant's guilt. That is, even people who appear to be guilty of the most hideous crimes are entitled to be tried fairly. But not, apparently, in Western Australia.

Differing analyses of relevance

The facts of the case on appeal require attention to illustrate this point, but more generally the case illustrates how senior judges can differ on what is in issue in a case. The majority (Gummow, Crennan, Kiefel and Bell JJ) held that the evidence in contention was not relevant to any issue, and a retrial was ordered, whereas Heydon J dissented, holding that there were live issues on which the evidence had probative value.

Stratagems and spoils (smile, Bard)

Heydon J draws attention to the significance of concessions by the defence on the scope of admissible prosecution evidence. The majority did not consider that the case required examination of this.

A clever defence tactic is to spoil the prejudicial effect of detailed prosecution evidence by conceding that aspects of it are not in issue. Where there is no issue, the theory goes, the point is not a matter on which proof is needed, or indeed permitted. Evidence must be relevant to a matter in issue.

Heydon J recognised that this was not an appropriate case to explore these stratagems, as no such techniques were used here.

Sentencing the reformed offender

The theme of maximum flexibility for a judge in fixing a sentence that fits both the crime and the offender, recently discussed here in relation to guideline judgments, is also present in Pepper v United States, USSC No 09-6822, 2 March 2011.

Pepper illustrates the relevance of post-sentencing rehabilitation when a sentence is reconsidered on appeal. The advisory guidelines (United States v Booker, 543 U.S. 220 (2005), discussed here on various occasions from 13 December 2007) could be departed from to recognise rehabilitation, and legislation to the contrary effect was declare invalid.

Possession of the fullest information possible on the offender's life and characteristics has long been held to be essential to selection of an appropriate sentence: Sotomayor J delivering the opinion of the Court in Pepper, citing Williams v New York, 337 U.S. 241, 246-247. Thomas J dissented in Pepper, on the grounds that the guidelines should be mandatory. He has held to this line previously, apparently not thinking that he should revise his views to conform to the law which he has no chance of changing.

Confronting imaginary emergencies

Michigan v Bryant, USSC No 109-150, 28 February 2011, continues the line of cases which purports to bring the reliability exception to the rule excluding hearsay into interpretation of the Confrontation Clause of the Sixth Amendment. See my 27 June 2008 comments on Giles v California, 554 U.S. 353 (2008). Bryant is a rather questionable decision on its facts, as Scalia J vigorously points out in his dissent.

To achieve admission of the evidence - the mortally wounded victim told the police who had shot him - the majority (Sotomayor J, joined by Roberts CJ and Kennedy, Breyer and Alito JJ, with Thomas J separately concurring) assessed the facts "objectively" from the points of view of the victim (the "declarant") and of the police who questioned him. Perhaps they were pushed to this curious approach because there was no evidence that the emergency which was necessary to escape the exclusion of the evidence actually existed: see Scalia J, slip op., pp 9–10; the majority said that the declarant and the police should have thought it did. That is, the majority held that there was an objective emergency and the declarant's objective purpose (not his actual purpose, but the purpose he should have had) was to help the police meet this emergency, so what he said was not testimonial and the confrontation clause did not apply.

The Confrontation Clause states:

"In all criminal prosecutions, the accused shall enjoy the right . . . to be confronted with the witnesses against him."

Witnesses are those who bear testimony, and testimony means "[a] solemn declaration or affirmation made for the purpose of establishing or proving some fact": Crawford v. Washington, 541 U. S. 36, at 51.

Those of us who do not have to worry about the Confrontation Clause can still take an interest in the majority's strange reasoning. I exempt Thomas J from this criticism, as he simply focused on whether the declarant's statement was "testimonial" within the meaning of that term in the Sixth Amendment and held that it was not because the interrogation here did not have the characteristics of a formalised dialogue with indicia of solemnity. He thus avoided trying to reconstruct the primary purpose of the participants. But the reasoning that constitutes the opinion of the Court is very odd.

The Court addresses the purpose component of testimony objectively (slip op., p 13):

"... the relevant inquiry into the parties' statements and actions is not the subjective or actual purpose of the particular parties, but the purpose that reasonable participants would have had, as ascertained from the parties' statements and actions and the circumstances in which the encounter occurred."

This undermines the basis for the reliability exception, which is that the circumstances as perceived by the speaker who is no longer available as a witness were such as to make fabrication unlikely. But the Court would say that even if the declarant did not realise that he needed to speak urgently, he should have realised that urgency existed and accuracy was required.

Further, as Scalia J points out, the Court is giving judges the power to do what they like:

"If the dastardly police trick a declarant into giving an incriminating statement against a sympathetic defendant, a court can focus on the police's intent and declare the statement testimonial. If the defendant "deserves" to go to jail, then a court can focus on whatever perspective is necessary to declare damning hearsay nontestimonial. And when all else fails, a court can mix-and-match perspectives to reach its desired outcome. Unfortunately, under this malleable approach "the guarantee of confrontation is no guarantee at all." Giles v. California, 554 U. S. 353, 375 (2008) (plurality)."

The Court recognised that the existence of an emergency was relevant, but in relation to the interrogators' primary purpose, calling this a "context-dependent inquiry". At slip op., p 22 it held:

"The inquiry still focuses on the understanding and purpose of a reasonable victim in the actual victim's circumstances, which prominently include the victim's physical state." [emphasis added]

The alarming thing here is that the actual victim here did not seem to appreciate any urgency. After being shot he fled some six blocks and at a gas station 25 minutes later a 911 call was made and the police quickly arrived. Then five officers in turn questioned him about the shooting.

"[the victim's] pressing medical needs do not suggest that he was responding to an emergency, but to the contrary reinforce the testimonial character of his statements. He understood the police were focused on investigating a past crime, not his medical needs. None of the officers asked [him] how he was doing, attempted more than superficially to assess the severity of his wounds, or attempted to administer first aid." [Scalia J's opinion, p 7]

Justice Ginsburg agreed with Scalia J that the declarant's intent is what counts. She noted that the issue of whether this case was one of a dying declaration was not before the Supreme Court because it had been abandoned by the prosecution in the Michigan Supreme Court. So although the Court had recognised that the dying declaration exception to the hearsay rule was an exception to the confrontation requirement in the common law inherited from England (Crawford v Washington, above), it was not necessary here to decide whether it survived the Court's recent confrontation decisions.

Bryant is a case where, as the prosecution recognised, the easy answer - that the declarant was making a dying declaration - was not available on the true facts, so the plurality had to uphold the admission of the evidence on these inconvenient facts by creating law that escapes the Confrontation Clause by pretending the statements were in response to an emergency and so were not testimonial. With the "facts" being what judges think they should have been, judges can do as they wish.

Aiding or standing by?

Aiding an offence requires a positive act of assistance in its commission. This is not new law, but it is usefully illustrated in Robinson v R (Bermuda) [2011] UKPC 3 (9 February 2011).

One of the ways in which the appellant had been alleged to have been guilty of the murders of twins was that he

"intentionally conveyed to … Burgess [the principal] by his presence and behaviour that he was assenting to and concurring in the commission of the offence"

Sir Anthony Hughes, for the Board, recognised the danger in this form of allegation [14]:

"The Board is disposed to agree that to frame an allegation of aiding in [this way] does carry danger and is best avoided unless carefully qualified and explained. It courts the risk that insufficient attention is paid to the undoubted requirement that aiding imports a positive act of assistance. Of course that positive act of assistance may sometimes be constituted by D2 being present, and communicating to D1 not merely that he concurs in what D1 is doing, but that he is ready and willing to help in any way required. ... If D2's presence can properly be held to amount to communicating to D1 (whether expressly or by implication) that he is there to help in any way he can if the opportunity or need arises, that is perfectly capable of amounting to aiding ... . It is, however, important to make clear to juries that mere approval of (ie "assent" to, or "concurrence" in) the offence by a bystander who gives no assistance, does not without more amount to aiding. It is potentially misleading to formulate aiding [in the way mentioned above] without that qualification and without explaining that the communication of willingness to give active assistance is a minimum requirement."

The prosecution may prefer to allege aiding rather than the more complicated form of extended secondary liability. An opportunity to do this can occur where although a common plan may have been departed from, the aider continues to assist [18]:

"... an aider (D2) is guilty in respect of acts which he assists the principal offender (D1) to commit, knowing what D1 is about, so that if D1 steps right outside what was contemplated by D2, the latter will not be guilty. That, however, assumes that D2's assistance ceases upon the fundamental departure by D1. It is clearly otherwise if D2 continues to render assistance after a change of direction by D1."

So, just a reminder of some fundamentals here.

Thinking for the jury

One of the requirements for a fair trial is the correct application of the law. When a jury has doubts about what the relevant law is, it may ask the judge for guidance. In R v Miljevic, 2011 SCC 8 (16 February 2011) the Supreme Court of Canada split 4-3 on whether the judge had correctly responded to questions from the jury.

In this case there was no need for the court to decide new law. There was no doubt about what the applicable law was. The jury had to choose between murder and manslaughter, depending on what it found the accused's state of mind to be. The law on this is settled, but the jury wanted clarification.

The jury had asked: "In 'layman terms' what is the difference between murder 2 and manslaughter? Examples? … A specific definition of manslaughter?"

The judge answered not by giving examples or by giving a definition of manslaughter, but instead by repeating the definition of murder. He did not address manslaughter because he did not want to confuse the jury, and because he did not want the jury to disobey his instruction to convict on one charge by acquitting on both.

The minority (Fish J, with McLaughlin CJ and Deschamps J concurring) held [8]:

"...no 12 jurors should be required by a trial judge to convict the accused placed in their charge of one or the other of two offences without understanding how the elements of both might relate to the evidence before them (see R. v. MacKay, 2005 SCC 75 (CanLII), 2005 SCC 75, [2005] 3 S.C.R. 607, at para. 1, citing Azoulay v. The Queen, 1952 CanLII 4 (S.C.C.), [1952] 2 S.C.R. 495, at p. 503). Yet that is what happened here."

There may be occasions where only one of the two available alternatives needs to be decided, but was this a case where the mens rea for murder could be determined without comparing it to the mens rea for manslaughter? The jury had to assess foreseeability of serious bodily harm (and it seems the judge misdirected on this: [22]) in the context of the accused's impairment through intoxication. Given that intoxication does not excuse the failure to foresee harm that a reasonable person would have foreseen (for manslaughter), could a direction on manslaughter possibly have made any difference here?

The majority (Cromwell J, with Abella, Charron and Rothstein JJ concurring) agreed with the majority in the Alberta Court of Appeal, holding that on the agreed facts the accused was guilty of manslaughter (he had thrown what he claimed was a heavy baseball bat into a group of people), so the only issue at trial was whether the accused was guilty of murder. Also, the judge had invited the jury to ask questions if they had difficulty.

The Court of Appeal had not addressed the question of whether the only issue in the case could be decided without comparison with the mens rea for manslaughter. The Supreme Court majority concluded [3]:

"...There is no reasonable possibility that the jury could have misunderstood what had to be proved in order for them to return a guilty verdict on the charge of second degree murder."

This treatment of the single-issue without comparison with the criteria for the alternative is wrong. A direction on manslaughter, even though the accused was guilty of this at a minimum, could have helped the jury understand the mens rea for murder. Intoxication may have had a bearing on what the accused knew of the likely consequences of his act, and the fact that a reasonable person, who is by definition sober, would have recognised the risk is not quite the same thing. Comparison with manslaughter would have clarified the subjective nature of the requirements for mens rea in murder.

Securing trial fairness

The primacy of a defendant's absolute right to a fair trial is preserved in aspects of Canadian evidence law concerning the withholding of disclosure in the interests of national security: R v Ahmad, 2011 SCC 6 (10 February 2011). This is because the court can, in the event that absence of disclosure compromises the right to a fair trial, order a stay of proceedings.

The legislative scheme under consideration in Ahmad is ss 38 to 38.16 and 39 of the Canada Evidence Act, R.S.C. 1985, c. C-5. Section 38.14 provides:

Protection of right to a fair trial

38.14 (1) The person presiding at a criminal proceeding may make any order that he or she considers appropriate in the circumstances to protect the right of the accused to a fair trial, as long as that order complies with the terms of any order made under any of subsections 38.06(1) to (3) [permiting disclosure] in relation to that proceeding, any judgment made on appeal from, or review of, the order, or any certificate issued under section 38.13 [prohibiting disclosure].

Potential orders

(2) The orders that may be made under subsection (1) include, but are not limited to, the following orders:

(a) an order dismissing specified counts of the indictment or information, or permitting the indictment or information to proceed only in respect of a lesser or included offence;

(b) an order effecting a stay of the proceedings; and

(c) an order finding against any party on any issue relating to information the disclosure of which is prohibited.

The Court, in a unanimous judgment, held [2] in relation to a conflict between the interests of national security which require concealment of information and the interests of the defendant,

"Where the conflict is irreconcilable, an unfair trial cannot be tolerated.
Under the rule of law, the right of an accused person to make full answer and defence may not be compromised."

This requirement of trial fairness was the criterion for constitutional validity of the legislation [5]. The important consideration was the flexibility of the legislated scheme [7]:

" ... the net effect is that state secrecy will be protected where the Attorney General of Canada considers it vital to do so, but the result is that the accused will, if denied the means to make a full answer and defence, and if lesser measures will not suffice in the opinion of the presiding judge to ensure a fair trial, walk free. While we stress this critical protection of the accused's fair trial rights, we also note that, notwithstanding serious criticisms of the operation of these provisions, they permit considerable flexibility as to how to reconcile the accused's rights and the state's need to prevent disclosure."

Important here is the context in which a stay of proceedings may have to be considered. Usually the stay is described as a remedy of last resort, but here a stay of proceedings may be required even though not all the information has been disclosed to the judge who therefore could not say that it was necessarily the only appropriate remedy [34-35].

There is no obligation on the defence to assist the court (for example by undertaking not to disclose to the defendant information given by the prosecution to counsel, see [30]) to avoid the need to order a stay [78]:

"... the defence is under no obligation to cooperate with the prosecution and if the end result of non-disclosure by the Crown is that a fair trial cannot be had, then Parliament has determined that in the circumstances a stay of proceedings is the lesser evil compared with the disclosure of sensitive or potentially injurious information."

The trial judge must be able to "conclude affirmatively" [35] that the right to a fair trial has not been compromised.

For my analysis of trial fairness in various leading appellate courts, click here.

Second thoughts

A witness's privilege against self-incrimination

Cases of alleged domestic violence are among those where a complainant may wish to deny the truth of her earlier complaint or of her evidence in pre-trial proceedings. She would then be admitting wasting police time by making a false complaint, or perjury. Will she be able to claim a privilege against self-incrimination so that she is not forced to give evidence incriminating the defendant at his trial?

This depends, as the New Zealand Supreme Court held in DK Singh v R [2010] NZSC 161 (17 December 2010), on how "likely" (s 60(1)(b) of the Evidence Act 2006[NZ]) it is that provision of the information sought would be used to incriminate the witness, that is, on whether there is a "real and appreciable" – as opposed to a "merely imaginary and fanciful" risk of incrimination (Singh at [31], citing Cockburn CJ in R v Boyes (1861) 1 B & S 311 at 330, 121 ER 730 (KB) at 738). In the circumstances of Singh the Court assessed this likelihood as sufficiently low to justify denial of the privilege.

The Court added that the privilege belongs to the witness, and it is not open to the appellant to make the claim on her behalf if she had waived it, applying R v Kingslake (1870) 11 Cox CC 499 (QB) and noting the consistency with s 60(4)(b) of the Evidence Act 2006.

Hostility and prior consistent statements

In the trial in this case the Crown had obtained a ruling that the witness was hostile, so that by cross-examination it was revealed that she had previously stated that the alleged offences had occurred. The defence then sought to have some of her prior consistent (that is, consistent with her denials of the offending) statements admitted under s 35 of the Evidence Act. But on the facts here the Court assessed those statements as not having sufficient probative value to make them "necessary to respond" to the Crown's challenge to the witness's accuracy or veracity, and held that their selective nature would make admitting them unfair to the prosecution and would require a time-wasting diversion (s 8 Evidence Act).

Trial unfairness

The appellant's fundamental argument was that the trial had been unfair. This is dealt with at the end of the judgment [59-61]. This argument was put on the grounds (as I would paraphrase on the basis of my analysis of trial fairness) that the jury would not have assessed the evidence impartially because the witness's credibility had been improperly undermined. It seems from the judgment that the unfairness argument was based on the jury not having been given a direction that they might consider that her clumsy attempts to deny that the offending had occurred were due to her fear of being prosecuted for perjury. That is, there was a real risk that the jury's assessment of the value of her evidence was not an impartial assessment because all relevant considerations may not have been taken into account.

The Court concluded that the jury had been given sufficient information to be able properly to assess the witness's credibility. A fear of prosecution would not have affected the way she gave evidence.

Res judicata or double jeopardy?

The different foundations of the special pleas (autrefois convict/acquit) and issue estoppel are called to mind by Coke-Wallis, R (on the application of) v Institute of Chartered Accountants in England and Wales [2011] UKSC 2 (19 January 2011). Lord Collins noted at [59] that the principles of the special pleas do not apply in civil cases, and that Lord Bridge had been wrong in Harry Lee Wee v Law Society of Singapore [1985] 1 WLR 362, 368 (PC) to suggest that they do.

Re-litigation of issues may be permitted, subject to the court's duty to prevent an abuse of its process. Re-litigation in civil cases is controlled by the doctrine of res judicata, which has generated the rules concerning cause of action estoppel and issue estoppel. In criminal cases the prevention of double jeopardy is the conceptual source of the special pleas.

In civil cases, cause of action estoppel prevents different evidence being used to prove a suit that had previously failed, as occurred in Coke-Wallis. If the earlier proceedings had been criminal, different considerations would apply to the admissibility of determinations of issues in the subsequent civil case: see Z v Dental Complaints Assessment Committee [2008] NZSC 55 (25 July 2008), noted in the entry for 25 July 2008. Acquittal in criminal proceedings does not prevent the same issues being decided in a subsequent civil case. This reflects the generality of the criminal verdict. The higher standard of criminal proof makes appropriate the use of evidence of a previous conviction in a subsequent civil case, which is likely to be permitted if relevant although there may be exceptions (as is recognised for example in s 47 of the Evidence Act 2006[NZ]). And, still focusing on where the second case is civil, evidence of a judgment or of a finding of fact in an earlier civil case is likely to be admissible (and not disputable) within the principles of res judicata but not otherwise.

A particularly clear outline of the position where the second case is criminal is given in R v Carroll [2002] HCA 55; 213 CLR 635; 194 ALR 1; 77 ALJR 157 (5 December 2002). In criminal law a more flexible approach may be permitted to prevent abuse of process. On double jeopardy, Gleeson CJ and Hayne J observed at 23-24:

"[23] It is ... important to recall that the four considerations which we have mentioned (the imbalance of power between prosecution and accused, seriousness for an accused of conviction, prosecution as an instrument of tyranny and the importance of finality) are not the only considerations which find reflection in the criminal law system. At the very root of the criminal law system lies the recognition by society that some conduct is to be classified as criminal and that those who are held responsible for such conduct are to be prosecuted and, in appropriate cases, punished for it. It follows that those who are guilty of a crime for which they are to be held responsible should, in the absence of reason to the contrary, be prosecuted to conviction and suffer just punishment.

"[24] Reference to the general propositions we have mentioned is important not because the answer to the issues now being considered can be found by deductive reasoning which takes any or all of them as a premise but because they are values to which the criminal law can be seen to give effect. They are values that may pull in different directions. There are, therefore, cases in which a balance must be struck between them. To take only one obvious example, it is accepted that in order to acquit the innocent, some who are guilty will go unpunished. But conversely, to punish the guilty, some who are innocent will suffer the very real detriments of being charged and tried for an offence they did not commit. It follows that to argue from any one of the considerations we have identified to some rule of universal application is to invite error."

Where the first case was criminal and there was an acquittal, it may be impossible to say that a given fact was determined to any standard. This makes the application of estoppels inappropriate (Carroll at [35]). Instead, it is recognised that there is a discretion which underlies the special pleas: Connelly v Director of Public Prosecutions [1964] AC 1254 per Lord Pearce at 1364:

"A man ought not to be tried for a second offence which is manifestly inconsistent on the facts with either a previous conviction or a previous acquittal. And it is clear that the formal pleas which a defendant can claim as of right will not cover all such cases. Instead of attempting to enlarge the pleas beyond their proper scope, it is better that the courts should apply to such cases an avowed judicial discretion based on the broader principles which underlie the pleas." (emphasis in original)

This discretion is the inherent power of the court to prevent an abuse of its process (Carroll at [39]). So, although a plea of autrefois acquit may not technically be available on a particular issue, considerations of double jeopardy may indicate that the laying of the second charge was an abuse of process. Outside of the technical application of autrefois acquit, it may be possible to reveal double jeopardy by comparing the issues in the first trial with those raised in the second. Carroll is an illustration: at the first trial the charge was murder, the only issue was identity, and the accused gave evidence. He was acquitted but subsequently charged with perjury on the basis of his denial of being the killer. The elements of perjury differ from the elements of murder, so autrefois acquit did not assist the defendant, but the High Court unanimously held that the perjury prosecution was an abuse of process and it was stayed.

While issue estoppel has no place in criminal law in Australia (Rogers v R [1994] HCA 42; (1994) 181 CLR 251), the UK (R v Humphrys [1977] AC1) and New Zealand (R v Davis [1982] 1 NZLR 584 (CA)), this is not so in Canada: R v Mahalingan [2008] SCC 63 (14 November 2008), discussed here on 15 November 2008. In that case the issue was whether evidence that the defendant had attempted to interfere with a witness, a charge on which he had been acquitted, was admissible at his trial for aggravated assault. It should be obvious that, unlike in Carroll, proof of the second charge did not entail contradicting the acquittal, but the majority of the Supreme Court of Canada held that issue estoppel applied because it included findings of reasonable doubt on an issue. It seems clear that the Canadian approach, committed as it is to the use of the res judicata-based estoppels, is having its limitations reduced by a fiction that a finding of a reasonable doubt is a determination of an issue.

Trial fairness and conviction appeals

Stimulating holiday reading! My new paper on what a "fair trial" means and on reform of the criteria for allowing appeals against conviction! Sit up straight and click on the link under draft papers on this page. Also available at NZLawyer (4 February 2011), but the former - my pdf file - is easier to read because it has footnotes, whereas the latter incorporates the footnotes into the main text.

Judging or calculating?

Some general thoughts on sentencing guideline judgments are prompted by the High Court of Australia's decision this week in Hili v R; Jones v R [2010] HCA 45 (8 December 2010):

• The point of first reference for sentencing must be the relevant legislation. A court cannot issue guidelines that are inconsistent with statutory provisions.
  

• If a guideline indicates a "norm" for the starting point for sentence determination, can that go beyond being merely descriptive of what earlier cases have decided? What authority does the court have to set "norms" for the future?
  

• Without statutory authority to lay down norms for the future, or to describe even in the most general terms the circumstances in which those norms can be departed from, what binding force can these future "norms" have?
  

• Plainly, the purpose of sentencing guidelines is to promote consistency between cases. But cases are never exactly the same, and a sentencing judge should exercise judgment about what is appropriate for an individual offender. The task of counsel for each side is to bring to the court's attention cases that it asks to be used as precedents, and this is an exercise in describing what has happened in the past. That is the way the law usually works.
  

• Consistency between cases requires consideration of the reasons why previous sentences were imposed. This is not revealed by an exercise in tabulation of seriousness of offence and level of sentence, because that misses the vital ingredient of individual circumstance and other matters required by statute to be taken into account.
  

• So, first look at the legislation, then look at what has been done in other cases – especially to identify unifying principles – remembering that what has been done in the past does not necessarily set the limits on what can now be appropriate.
  

• Whether a sentence under appeal was manifestly inadequate or excessive does not depend on an intuitive reaction based on earlier decisions, but rather is a question that has to be answered in the light of all the matters relevant to fixing sentence, including the circumstances of the offending and the circumstances of the offender, and the sentences imposed in the most comparable cases.
  

• It is possible (as Heydon J said in his concurring but partly differently reasoned judgment, although the point must be uncontroversial) that there may be more than one "correct" sentence in a given case. A sentencing judge must be guided by his or her perception of what is appropriate.
  

Those points encourage us to doubt the status of so-called guideline judgments. Examples in New Zealand: establishing sentencing bands for drug offending based on the quantity of drug involved: R v Fatu [2006] 2 NZLR 72, R v Terewi [1999] 3 NZLR 62; categories for burglary based on the kind of offender: Senior v Police (2000) 18 CRNZ 340; bands of sentences for violent offending based on listed aggravating features: R v Taueki [2005] 3 NZLR 372; and – illustrating a level of analysis that has become, I respectfully suggest, grotesque – bands of sexual offending based on the physical details and grouped into "rape bands" and unlawful sexual connection or "USC bands" (!): R v AM (CA27/2009) [2010] NZCA 114, [2010] 2 NZLR 750. These cases are prescriptive and may well be found to contravene the principles in the Sentencing Act 2002. Their underlying assumption is that judges are unable to exercise discretion judicially. These cases are, perhaps some might argue, an overly eager attempt by the appellate court to placate pressure groups whose only thoughts about sentencing are retributive. They are, of course, honest endeavours to promote uniformity in sentencing, but, especially when read in the light of the Supreme Court's rejection of a formulaic assessment of the amount of discount appropriate for pleas of guilty, they may have gone wrong in neglecting the legislated approach; see Hessell v R [2010] NZSC 135 (discussed here 16 November 2010).