The Law Commission

PART VI
CODIFYING THE DOUBLE JEOPARDY RULE

6.1      In CP 156 we proposed that the rule against double jeopardy and its exceptions should be put into statutory form. Those respondents who commented on this proposal were generally in favour, on the grounds that this would ensure greater clarity and (the CPS noted) compatibility with the ECHR.
6.2      Our proposal was for the codification of both the principles that, together, make up the "rule against double jeopardy" – that is, not only

(1) the "autrefois" rule, that a person who has previously been acquitted or convicted of an offence may under no circumstances^[1] be prosecuted again for the same offence, but also

(2) the principle laid down by the House of Lords in Connelly v DPP,^[2] that only in special circumstances may a person who has previously been acquitted or convicted of an offence be prosecuted for any offence based on the same or substantially the same facts.^[3]

That, however, was on the basis that the exceptions to the two principles would be identical. Now that we have decided (in the light of the consultation process) to place more emphasis on the value of finality, and to recommend a new evidence exception much narrower than we had originally proposed, the case for consolidating both principles into one seems weaker. We think it clear that the autrefois rule, at least, should be put in statutory form. We recommend the codification of the autrefois rule and its exceptions.

(Recommendation 19)

6.3      We return later to the question whether the Connelly principle should also be codified.^[4]

The concepts of acquittal and conviction

6.4      The statutory autrefois rule will need to make use of the concepts of acquittal and conviction, because it is only when a person has been acquitted or convicted that the rule will apply. It would be possible simply to use these concepts in the senses in which they are used in the present law. However, in Part IX of CP 156 we considered whether they should be made narrower or wider than they are now.

Postponing the point at which the prosecution can drop the case without losing the right to re-open it

6.5      The CPS has power to discontinue a prosecution without the defendant being acquitted, so that a subsequent prosecution for the same offence is not precluded by the autrefois rule. Once the case is committed for trial, however (or, where the case is sent directly to the Crown Court because it cannot be tried by magistrates, once the indictment is lodged), this power no longer exists. From that point on, the Crown can abandon the case only by formally offering no evidence.^[5] If a jury has been sworn, the judge will then direct it to acquit; if not, the judge will direct a verdict of acquittal to be recorded, and by statute that verdict has the same effect as if the defendant had been acquitted by a jury.^[6] In either case, therefore, the autrefois rule will apply to prevent the case being reopened.
6.6      In CP 156 we discussed whether the point at which the prosecution may be dropped, without losing it for ever, should be postponed to some later stage. This might be done by extending the CPS's power to discontinue by an administrative procedure, without the need for any judicial determination of the case at all. The CPS argued for such an extension. Alternatively the case might still come to court, but when the Crown offers no evidence the case might be disposed of by an order which does not amount to an acquittal for the purpose of the autrefois rule. Such an order might be similar to that commonly made where the Crown accepts a plea of guilty to a lesser offence than that charged – namely that the count or counts not proceeded with should lie on the file, not to be proceeded with without the consent of the court or the Court of Appeal.
6.7      It is arguable that a defendant has not been in jeopardy until the trial has at least commenced, and that, should proceedings be halted before that time, they should not be regarded as conclusively determined, and it should still be possible to try the defendant. This was the view put forward by the CPS. Our answer can be found in paragraph 9.6 of CP 156:

In our view, … a defendant is entitled to expect accurate assessment of his or her case, and sound decision-making by the prosecutor, at a reasonably early stage. Committal (or preferment,^[7] where there is no committal) is the earliest reasonable stage, given the realities of the criminal justice system. The prosecutor must have some opportunity to abandon the case without penalty, because the proceedings are usually initiated by the police before the prosecutor sees the file. But the decision whether to commit requires the prosecutor to make a positive decision about the case. It therefore seems reasonable that the protection afforded by the double jeopardy rule should begin at that point.

6.8      In any event it is debatable whether this issue is best regarded as an aspect of double jeopardy at all. The double jeopardy rule is designed to ensure that, once proceedings have been finally disposed of, they cannot be resurrected. The rule does not apply if the proceedings come to an end without the defendant being finally pronounced either guilty or not guilty. That must inevitably continue to be the case: it is the essence of the double jeopardy rule. That being so, arguably the real question is what sort of order should be made where the Crown wishes to offer no evidence (that is, whether the defendant should have to be acquitted, or whether it should be possible to bring the case to an end without an acquittal) rather than whether the order that is made should count as an acquittal. This is an important issue,^[8] but, in our view, it raises wider questions of criminal procedure, and a report on double jeopardy is not the context in which to tackle it. We therefore make no recommendation.

Acquittal or conviction in another jurisdiction

6.9      At the moment, it seems that the autrefois plea would preclude a second trial in England and Wales after a previous trial in a foreign jurisdiction.^[9] We pointed out in CP 156 that in this respect the rule might be relaxed without infringing the ECHR.^[10] We offered four possibilities, namely that the rule against double jeopardy should apply where there was a previous acquittal or conviction (1) anywhere; (2) anywhere except in one of a number of countries expressly excluded for this purpose; (3) in an EU state only; or (4) in England and Wales only. Of these, we provisionally proposed option (1).
6.10      Respondents overwhelmingly, and in many cases strongly, agreed. The need for comity was a common theme. P W Ferguson pointed out that Scots law follows option (1). The DTI favoured option (1) because any change would pre-empt Schengen and that would cause problems. The Northern Ireland judiciary felt that the options other than (1) would confuse our various extradition arrangements. Professor David Feldman argued that, although English courts will not enforce foreign courts' criminal judgments in civil disputes, there is a different concern in recognising foreign acquittals and convictions for double jeopardy purposes. There was a general view that at the root of this issue is the basic principle that a second trial for the same crime is unjust and oppressive, and that that fact is not changed by the location of the first trial.
6.11      The CPS agreed with this general principle, but favoured a list of countries whose verdicts would bind the English courts (in effect, option 2) because this was the proper way to ensure that only the decisions of competent courts would prevent second trials here. The CPS thought it a matter of policy for the Government, not the courts, to determine which foreign courts were competent. The problem with the CPS's view is that it involves casting general aspersions at governmental level on countries with which we have friendly diplomatic relations, when it would be better to focus on whether a particular trial abroad was so flawed that its verdict need not be considered final by an English court.
6.12      The CPS response does, however, point to a serious concern, because a blanket rule that prior foreign proceedings will always preclude an English trial could work manifest injustice. There is a considerable danger in option (1) of having to respect the outcomes of sham trials in corrupt and illegitimate regimes. Take the case of a very wealthy businessman who perpetrates a fraud on thousands of English pensioners, stealing £50 million. He travels to a state where he has for many years cultivated influential connections. He arranges to be tried there for the fraud and is acquitted (or convicted and given an absolute discharge on the grounds that he is an honourable man) because the nation's president ordered it, or because the judge dismissed the case in return for an honorarium. If the businessman returned to England, it may be thought to be an affront to justice were it not possible for him to be tried. Again, if the fraudster were foreign, and were tried in his own state and acquitted because his activities were not criminal under the law of that state, or because they were conducted in England and therefore outside the territorial jurisdiction of the courts of that state, it would be unreasonable to bar English proceedings against him. In another example, a foreign national commits a series of assaults and flees to his own country of citizenship where he is tried and acquitted on the grounds that that state has a policy of acquitting its nationals for offences committed abroad. Surely, he too should be open to prosecution. With certain exceptions, English criminal jurisdiction is territorial, and so applies to offences committed within England and Wales only. Where an English court has jurisdiction over an offence because it was committed here, it would be more than unfortunate if the possibility of trying the suspect were ousted by a show trial somewhere else.
6.13      There may be concern that it would be wrong in principle for English courts to investigate the propriety of proceedings before courts in other jurisdictions. However, this is exactly what English courts are required to do in private international law cases, where they may not recognise judgments delivered by foreign courts in breach of "substantial justice",^[11] or by fraud, or where to do so would be against public policy.^[12] If it is right in principle for an English court to refuse to recognise a foreign judgment in a civil matter, we think it should be equally acceptable for it to refuse to do so in a criminal matter, where the consequences may be far more serious. English courts are, of course, extremely reluctant to impugn the decisions of foreign courts, and will strain to maintain comity. The power to ignore a foreign verdict would undoubtedly be very sparingly used, but we believe that it is desirable in the interests of justice. This is all the more true as the ECHR would apparently permit English law to refuse to recognise any foreign verdicts for the purposes of the domestic double jeopardy rule. There is also a close analogy with the tainted acquittal procedure in denying finality to foreign proceedings which were fundamentally defective.
6.14      We do not think that it would be sufficient, however, simply to give the courts a power to disregard foreign verdicts where they think it appropriate, with no indication of the considerations to which they should have regard in deciding whether to do so. For this purpose a useful precedent is provided by Article 20(3) of the Rome Statute of the International Criminal Court. This allows a second trial, despite a prior one on the same facts, where the first trial was in a different jurisdiction and was designed to shield the suspect from prosecution elsewhere, or was not independent, impartial or consistent with an intent to bring the perpetrator to justice. In the light of the overwhelming international consensus on this point, we think it would be appropriate to adopt similar criteria in English law.
6.15      We recommend that the autrefois rule should apply wherever the previous acquittal or conviction occurred, but an English court should be permitted to disregard an acquittal or conviction in another jurisdiction where it is satisfied that it is in the interests of justice to do so; and, in determining whether it is so satisfied, the court should be required to have regard to whether it appears that the foreign proceedings

(1) were held for the purpose of shielding the defendant from criminal responsibility for offences within the jurisdiction of the English court,

(2) were not conducted independently or impartially in accordance with the minimum requirements of due process and fairness, or

(3) were conducted in a manner which, in the circumstances, was inconsistent with an intent to bring the defendant to justice, together with any other considerations which appear to the court to be relevant.

(Recommendation 20)

6.16      There is a further, and quite different, situation in which it may be justifiable to allow a prosecution in England and Wales despite a prior acquittal in another country: namely where the defendant was acquitted on the ground that the alleged offence, if committed, was committed outside the jurisdiction of the foreign court. In this case there is no question of impugning the integrity of the foreign proceedings. Where a foreign court dismisses a case on the basis that the foreign legal system has insufficient interest in or connection with the alleged offence, we think it should still be open to the English courts to proceed against the defendant. Indeed (assuming that the offence is alleged to have been committed within the territorial jurisdiction of the English courts), it may be said that the very basis on which the foreign court has dismissed the case is that the matter is one for the English courts, and it would be illogical for an English court to decline to hear the case on the basis that a foreign court thought it should be heard by an English court.
6.17      This difficulty would not arise if, under the relevant foreign law, the foreign court's dismissal of the case were not a final determination but a preliminary decision as to admissibility. In that case there would be no final determination of the case, and therefore no possibility of an autrefois plea. If, on the other hand, the foreign ruling were a final one according to the foreign law, that would activate the autrefois rule, and prevent an English court from trying the case. We do not think it would be satisfactory that the applicability of the autrefois rule should depend whether, under the particular foreign law in question, the court's decision on the issue of jurisdiction was regarded as final. English law must have its own position on whether want of jurisdiction is final, and cannot answer that differently according to the country in which the abortive proceedings were taken. We therefore believe that a ruling of lack of jurisdiction by a foreign court should not count as a final determination of the proceedings for the purposes of the autrefois rule. We note that, according to Archbold, only a verdict of a foreign court of competent jurisdiction will activate the autrefois rule.^[13] Presumably, therefore, an acquittal for want of jurisdiction cannot have that effect, even under the present law.
6.18      The conclusion that the foreign court does not have "jurisdiction" to try an offence allegedly committed outside the foreign country may be reached in one of two ways. The first is that the court literally has no jurisdiction to try the case at all. The second is that the court does have jurisdiction to try the case, because it accepts jurisdiction in respect of the defendant such as on the basis of his or her presence within the jurisdiction, but the defendant is not guilty as charged because, as a matter of substantive criminal law, the conduct complained of is not an offence under the law of the country in question if it occurs outside the boundaries of that country. English law adopts the second analysis.^[14] There may well be countries whose law would adopt the first. We believe, however, that it should make no difference which analysis the foreign law adopts. In neither case should there be any bar to subsequent proceedings in the English courts.
6.19      There may be circumstances in which it would be inappropriate to try a defendant who has been prosecuted abroad but acquitted for want of jurisdiction: for example, the defendant may have spent several years on remand awaiting the foreign court's decision. In such a case, however, it would be open to the English court to stay the proceedings as an abuse of process.
6.20      We recommend that an acquittal by a foreign court should not be regarded as an acquittal for the purposes of the autrefois rule if it appears to have been based solely on the fact that the alleged offence, if committed, was committed outside the territorial jurisdiction of that court.

(Recommendation 21)

6.21      The last issue to be addressed in this section is whether the double jeopardy protection granted by virtue of a foreign acquittal should be subject to the same exceptions as an acquittal in England and Wales. It would be exceptionally difficult to apply the tainted acquittal procedure to foreign proceedings, and we believe that this possibility should therefore be discarded (except insofar as a foreign verdict could be ignored under the considerations discussed above). Our proposed new exception for fresh evidence in murder cases, however, would apply to acquittals in any jurisdiction, because it would be inconsistent to give a foreign verdict greater finality than one of our own. We recommend that the tainted acquittal procedure should not apply to acquittals outside England and Wales.

(Recommendation 22)

Extending the concept of an acquittal

6.22      In CP 156, we set out the main ways in which criminal proceedings can end without the defendant being convicted and sentenced, but without a final acquittal, with the result that the defendant can be prosecuted again for the same offence – namely:

(1) the discharge of the jury, without a verdict being obtained;

(2) the quashing of an indictment following a motion to quash;^[15]

(3) the discharge of the defendant at the conclusion of committal proceedings;^[16]

(4) the dismissal of an information on the non-appearance of the prosecutor;^[17]

(5) the dismissal of an information which is too faulty for the defendant to have been in jeopardy on it;^[18]

(6) the discontinuance of proceedings under section 23 or 23A of the Prosecution of Offences Act 1985;

(7) the withdrawal of a charge in the magistrates' court before the defendant has pleaded to it;^[19]

(8) the entering of a nolle prosequi;^[20] and

(9) an order that a count (or an indictment) lie on the file, not to be proceeded with without the consent of the court or the Court of Appeal.^[21]

6.23      The question we posed was whether any of these situations should count as a final acquittal or conviction so as to activate the double jeopardy rule in the future. Few respondents expressed views on this question but the clear majority were against any change. In none of these situations does there appear to be a strong case for protecting the defendant against subsequent prosecution. One respondent suggested that where a jury fails to agree, this should count as an acquittal; but, as we pointed out in CP 156,^[22] this would mean that in order to secure immunity from further prosecution a defendant would only need to persuade three jurors out of twelve to hold out against a conviction. The Stipendiary Magistrates' Council suggested that adjournments sine die should be treated as acquittals; but, if a magistrates' court wishes to close a case finally, it can dismiss the information. A decision to adjourn indicates an intention that the matter is not to be regarded as permanently closed. We recommend no change in this respect.

Conviction without sentence

6.24      In CP 156 we pointed out that, according to the Privy Council's decision in Richards,^[23] for the purposes of the autrefois rule there is no conviction until sentence is passed.^[24] We argued that, where for some reason sentence cannot be passed by the trial judge (for example, because the judge dies, becomes ill or resigns), the situation is analogous to the case where the jury cannot agree, and that the autrefois rule should not apply, so that the defendant can be retried and (if convicted again) sentenced. However, the Stipendiary Magistrates' Council pointed out that where such exceptional circumstances arise in a magistrates' court, sentence may be passed by a differently constituted bench. It is difficult to see why the position should be different in the Crown Court.
6.25      The only authority to the contrary of which we are aware is an obiter dictum in Richards. Giving the judgment of their Lordships, Lord Bridge of Harwich said:

Where a defendant is tried before judge and jury, both have their roles to play and together they constitute the court of trial. If, in any case following trial and conviction by the jury, the judge were to die before passing sentence, there would be no court seized of the case by which sentence could be passed. The defendant, it seems to their Lordships, would in those circumstances have to be rearraigned before another court and if he again pleaded not guilty would have to be retried.^[25]

6.26      This reasoning does not accord with everyday practice in the Crown Court. We are aware of many occasions where, for good reasons of convenience and justice, a defendant may be sentenced by a judge other than the one who presided when the defendant pleaded guilty or was found guilty – for example where there are successive trials before different judges, perhaps at different venues, but it is sensible for the defendant to be sentenced on the same occasion for all outstanding matters by the judge presiding over the last trial in time. We can see no reason why the position should be any different where the judge dies after conviction but before sentence. If this is right then Lord Bridge's argument does not amount to a good reason for withholding the protection of the autrefois rule until sentence is passed.
6.27      It is noticeable that the case of Richards is not cited in either of the standard textbooks^[26] as authority for the proposition stated. Indeed, neither of them even deals with the question of the death of the judge between verdict and sentence.^[27] We believe that the dictum is at variance with the principle that a defendant should if possible be sentenced in respect of all outstanding matters at once, and with the fact that it is common for a defendant to be convicted at one trial but sentenced by the judge who hears a second trial. There is no reason, in law or in principle, why a convicted defendant should have to go through another trial on the same issue.
6.28      We recommend that, for the purposes of the autrefois rule, a conviction should be defined as including the giving of a verdict of guilty by a jury or a finding by a magistrates' court that an information is proved, whether or not sentence is passed.

(Recommendation 23)

Offences taken into consideration

6.29      Where a defendant is convicted, and asks for other offences to be taken into consideration ("TIC") in sentencing, for the purposes of the autrefois rule there is no conviction for those other offences. A few respondents (including Liberty and the Stipendiary Magistrates' Council) argued that a subsequent prosecution for those offences ought to be barred by the autrefois rule. Indeed Steyn J (as he then was) has described the present rule as "extraordinary".^[28]
6.30      We agree. The autrefois rule applies in a situation in which the prosecution offers no evidence. This may be for reasons which may be practical only, and do not involve the prosecution resiling from its belief in the guilt of the defendant. On the other hand the defendant, when asking for other offences to be dealt with as TICs, has to sign a form and confirm in court that he or she admits these offences even though never formally charged with them. Thus, the defendant goes further in acknowledging the rightness of the outcome than the prosecution need do in offering no evidence. It would be anomalous, therefore, were that outcome to be regarded as less final, for double jeopardy purposes, than where the prosecution offers no evidence.
6.31      Moreover, it is conceivable that TICs might constitute convictions for the purposes of the ECHR rule against double jeopardy. The Strasbourg Court has said (albeit in the context of Article 5, rather than Article 4 of Protocol 7) that

the word "conviction" … has to be understood as signifying both a "finding of guilt" after "it has been established in accordance with the law that there has been an offence" and the imposition of a penalty or other measure involving deprivation of liberty.^[29]

TICs involve a finding of guilt (by the defendant's admission) in accordance with law (there is a clear procedure) in respect of which a penalty is imposed (in that it is taken into consideration in the sentence). If TICs are within this definition and the definition is applicable to double jeopardy cases, any retrial of an offence previously the subject of a TIC would be a breach of the ECHR except where, virtually inconceivably, the trial took place on the permitted grounds of fundamental defect or new evidence.

6.32      We therefore conclude that TICs should in general be treated as convictions for the purposes of the autrefois rule. There is, however, one situation in which this would produce an unsatisfactory outcome. That is where a person has been convicted of an offence and on sentence asks for other offences to be TICed. If the conviction is quashed on appeal, it would seem wrong that the prosecution should be barred from proceeding in respect of the TICs. It is most unlikely that this would amount to double jeopardy for ECHR purposes because the TICs would not be a final disposal of the offences in question until the defendant's appeal in respect of the conviction was disposed of.
6.33      We recommend that, for the purposes of the autrefois rule, a conviction should be defined as including the taking of an offence into consideration in sentencing a person for another offence, unless the conviction for the latter offence is quashed on appeal.

(Recommendation 24)

Foreign proceedings

6.34      It may be that in some other jurisdictions the termination of proceedings in some of the circumstances listed at paragraph 6.22 above would be regarded, under the foreign law in question, as a final acquittal. Where this is so, our understanding is that the autrefois rule would preclude any attempt to prosecute the matter in England and Wales, even though a termination of English proceedings in similar circumstances would not count as a final acquittal in English law. We make no recommendation for any change in this respect. A defendant who has once secured what counts as a final acquittal under the law of the country where it occurs is in general entitled to assume that that is the end of the matter (subject to the exceptions recognised by the ECHR). It should not be possible to circumvent the rule against double jeopardy by taking further proceedings in another country. Our recommendation in relation to acquittals for want of jurisdiction^[30] is an exception to this principle, justified by the peculiar characteristics of such an acquittal.

The Connelly principle

6.35      In Part II, we explained that the protection against double jeopardy afforded by the autrefois rule is complemented by the wider and more flexible principle laid down by the House of Lords in Connelly v DPP^[31] and confirmed by the Court of Appeal in Beedie,^[32] namely that a person who has previously been acquitted or convicted of an offence may not be prosecuted for any offence based on the same or substantially the same facts unless there are special circumstances^[33] (which it seems may include the emergence of new evidence).^[34]
6.36      In CP 156 we concluded that the autrefois and Connelly rules should be restated as a single statutory rule against double jeopardy. The same exceptions (including our proposed exception for new evidence) would apply to both. This seemed to be logical in the light of our proposals that the new evidence exception should have a relatively wide reach. Indeed we relied on the apparent existence of a new evidence exception to the rule in Connelly as an argument of principle in support for a new evidence exception to the autrefois rule.
6.37      Our proposal attracted a lot of support, but there were significant voices in dissent – notably the CPS, which pointed out the difficulties where indictments had been severed, or where the prosecution had knowingly adopted two indictments and the defence had acquiesced in the first trial on such an indictment and had not applied for joinder.
6.38      Our provisional proposal, moreover, was for the reach of the new evidence exception to be widely drawn. We now accept, in the light of the response to CP 156, that we underestimated the importance of the autrefois rule by focusing solely on the effect on the individual, whereas there is a community interest in the individual not being subject to or at risk of oppression by the state making repeated attempts to convict on the same facts. As we have explained, this has persuaded us to recommend that the new evidence exception to the autrefois rule should have only a very limited reach.
6.39      The combination of these two factors has caused us to look again at Connelly. Plainly, if there is already a "special circumstance" exception to the Connelly principle which may encompass new evidence (though it is not clear how firmly established that exception is), then including the Connelly principle in a single statutory rule against double jeopardy which has a new evidence exception with a very limited reach would significantly change the balance of that rule as between prosecution and defence.
6.40      Having re-examined the Connelly principle, we have been reminded that it was not originally intended to be confined to situations of "double jeopardy" per se (that is, successive prosecutions based on the same or substantially the same facts), but concerned the wider question of the conscionability of successive prosecutions being brought where the charges could all have been dealt with in the one trial. In his speech in Connelly (which was treated in Beedie as forming the ratio of the House of Lords' decision) Lord Devlin argued that a prosecution should be stayed, in the absence of special circumstances, if the offence charged is one which could have been included in an indictment previously preferred against the same defendant. It was oppressive to bring two successive prosecutions where the matters alleged could have been dealt with in one. Counts may be joined in the same indictment if they "are founded on the same facts", or if they "form or are a part of a series of offences of the same or similar character".^[35] It is clear that, as stated by Lord Devlin, the principle concerns the rules for joinder, and applies to both limbs. It is worth quoting a sizeable section of his speech:

As a general rule a judge should stay an indictment (that is, order that it remain on the file not to be proceeded with) when he is satisfied that the charges therein are founded on the same facts as the charges in a previous indictment on which the accused has been tried, or form or are a part of a series of offences of the same or a similar character as the offences charged in the previous indictment. He will do this because as a general rule it is oppressive to an accused for the prosecution not to use rule 3^[36] where it can properly be used. But a second trial on the same or similar facts is not always and necessarily oppressive, and there may in a particular case be special circumstances which make it just and convenient in that case. The judge must then, in all the circumstances of the particular case, exercise his discretion as to whether or not he applies the general rule. Without attempting a comprehensive definition, it may be useful to indicate the sort of thing that would, I think, clearly amount to a special circumstance. Under section 5(3) of the Act a judge has a complete discretion to order separate trials of offences charged in one indictment. It must, therefore, follow that where the case is one in which, if the offences in the second indictment had been included in the first, the judge would have ordered a separate trial of them, he will in his discretion allow the second indictment to be proceeded with. A fortiori, where the accused has himself obtained an order for a separate trial under section 5(3). Moreover, I do not think that it is obligatory on the prosecution, in order to be on the safe side, to put into an indictment all the charges that might conceivably come within rule 3, leaving it to the defence to apply for separation. If the prosecution considers that there ought to be two or more trials, it can make its choice plain by preferring two or more indictments. In many cases this may be to the advantage of the defence. If the defence accepts the choice without complaint and avails itself of any advantages that may flow from it, I should regard that as a special circumstance; for where the defence considers that a single trial of two indictments is desirable, it can apply to the judge for an order in the form made by Glyn-Jones J in R v Smith.^[37]

6.41      Lord Devlin clearly intended that his principle should apply not only where the second prosecution relates to essentially the same conduct (that is, where it would involve double jeopardy) but also where it relates to completely different conduct which could have been included in the same indictment. In CP 156 we pointed out that Lord Devlin's formulation of the principle would preclude a prosecution for burglary in July which could have been combined with an earlier charge of burglary in June; but we added: "Even if the rule is really this wide (which seems doubtful) this sort of case is not an example of double jeopardy, and we do not deal with it in this paper."^[38]
6.42      We are still in doubt as to the true width of the principle. Our doubts are based on the fact that the principle itself lay dormant for 20 years before being resurrected in Beedie – it had been thought that the true ratio of Connelly lay in the other speeches. Beedie itself was a double jeopardy case, and the judgment does not indicate that the court saw the principle as extending beyond double jeopardy. We are not aware of a single case not involving double jeopardy in which Lord Devlin's principle has been applied. We have not considered whether it ought to apply to such cases. We did not consult on the point. In these circumstances we clearly cannot recommend, in a report on double jeopardy, the codification of a rule which is applicable to cases not involving double jeopardy, which (though there is high authority for it) has never to our knowledge been applied, and which, if widely applied, might well cause great difficulty for prosecutors.
6.43      It follows that we cannot recommend codifying the whole of the Connelly principle. At most we could recommend codifying that part of it which concerns double jeopardy, namely the principle that a prosecution should be stayed (in the absence of special circumstances) if it is based on the same or substantially the same facts as a charge of which the defendant has previously been acquitted or convicted. We have, however, concluded that we should make no such recommendation, for two reasons.
6.44      First, we cannot simply disregard the fact that Lord Devlin's principle, which Beedie treats as stating the present law, was in fact a general principle against the unjustifiable proliferation of proceedings, and not solely a rule against double jeopardy. Although its role as a safeguard against double jeopardy is now recognised, whereas its wider implications are not, we think it would be inappropriate to codify that part which has so far received recognition while leaving the remainder to be developed (or ignored) by the courts. What is needed is a proper examination of the merits and drawbacks of Lord Devlin's principle, in the terms in which he stated it. Until such an examination is carried out, we think it would be wrong to anticipate it by legislating on one aspect of the principle.
6.45      Secondly, we are in any event not entirely convinced that Lord Devlin's principle is sufficiently flexible to do justice in every case to which it would apply. This may be illustrated by the well-known case of Kevin Maxwell and others. Kevin Maxwell and three other defendants were acquitted in 1996 on charges of conspiracy to defraud the trustees and beneficiaries of the Maxwell Group's pension scheme by dealing with shares belonging to the pension fund in a way which created such a risk of loss as to be dishonest. The Crown then sought to prosecute Kevin Maxwell and other defendants for conspiracy to defraud various banks by giving them, as security for certain loans, shares owned by a subsidiary of the Maxwell Group. The Crown's case, as explained by counsel in resisting an application to stay these counts,^[39] was that the defendants had agreed to induce the banks to accept these shares as security by falsely asserting that the borrower was the beneficial owner of the shares. It was no part of the Crown's case that the acceptance of the shares as security involved an unacceptable risk to the banks through the illiquidity of the group. These counts had been severed from the original indictment in order to keep the length and complexity of the first trial within manageable bounds. However, Buckley J ruled that they should be stayed as an abuse of process.^[40]
6.46      In reaching this decision Buckley J did not purport to apply the principle laid down by Lord Devlin in Connelly: it was not until Beedie was decided in 1998 that that principle was accepted as an accurate statement of the law, even in relation to double jeopardy. Instead, he appears to have dealt with the case under the ordinary principles of abuse of process, exercising a large degree of judicial discretion. In particular, he took into account various factors in addition to the similarity between the new counts and those already disposed of, such as the publicity to which the defendants had been subjected, the distress which they and their families had already suffered, and the fact that the first trial had ended in all the defendants being acquitted. It is possible that, by focusing only on whether the new counts arose out of substantially the same facts and whether, if so, there were "special circumstances" justifying the second trial, Lord Devlin's principle might have made it harder for Buckley J to reach what he regarded as a just conclusion. Arguably the courts should have more room for manoeuvre in such cases than the Connelly principle allows.
6.47      That is not to say that the Connelly principle should be abandoned, and such cases left to the general law of abuse of process. Such a change would involve a substantial reduction in the protection against double jeopardy that the Connelly principle presently affords. This is because under the Connelly principle the prosecution is required to show that the proceedings are justified, whereas under the abuse of process regime the burden is on the defendant to show that the proceedings are not justified. Rather, we suspect that a balance needs to be struck between abandoning the principle altogether and retaining it as it stands. We believe, however, that the striking of this balance would be more appropriately done within the context of an examination of the law of abuse of process than in a report on double jeopardy.
6.48      One possible argument for seeking to codify the double jeopardy aspect of the Connelly principle is that it arguably represents the requirements of the ECHR. Unfortunately it is impossible to be sure what the ECHR does require, given that the two decided cases are directly contradictory. Gradinger v Austria^[41] suggests that a second prosecution for a different offence "based on the same conduct" is a breach of Article 4 of Protocol 7; Oliveira v Switzerland^[42] suggests that it is not, and that Article 4 therefore provides no greater protection than the autrefois rule. If Oliveira is right, the Connelly principle is not needed in order to ensure compliance. Even if Gradinger is right, the criterion of whether the second charge is "based on the same conduct" appears to be essentially the same as Lord Devlin's "substantially the same facts". It is therefore open to the courts to ensure compliance in the same way that they have until now, namely by applying either the Connelly principle or the general law of abuse of process. We do not believe that the possibility of courts ignoring the Connelly principle in such a way as to infringe Article 4 (even assuming that Article 4 goes as far as Gradinger suggests) is sufficient reason for us to recommend its codification. We therefore make no such recommendation.
Ý
Ü   Þ

Note 1    With the exception of the tainted acquittal procedure.    [Back]

Note 2    [1964] AC 1254; see paras 2.16 – 2.19 above.    [Back]

Note 3    We also envisaged that, if (contrary to the proposal at para 8.40 of CP 156) the rule in Sambasivam v Public Prosecutor, Federation of Malaya [1950] AC 458 were to be retained in some form, that rule too should form part of the codified rule against double jeopardy. We understood Sambasivam as deciding that in proceedings for any offence the prosecution may not adduce evidence that the defendant was in fact guilty of an offence of which he or she has previously been acquitted. In Z [2000] 2 AC 483, however, the House of Lords accepted our arguments against such a rule, and held that no such rule exists. It is therefore unnecessary for us to make any recommendation on the point. See paras 2.22 – 2.28 above.    [Back]

Note 4    See paras 6.35 – 6.48 below.    [Back]

Note 5    Unless the Attorney-General enters a nolle prosequi, which does not count as an acquittal.    [Back]

Note 6    Criminal Justice Act 1967, s 17.    [Back]

Note 7    Sc of the indictment. (Footnote added)    [Back]

Note 8    We were sent an interesting judgment on this tricky issue delivered by Judge John Samuels QC in the case of Johnston (William Martin) and others (Blackfriars Crown Court, 20 October 2000) for which we are grateful.    [Back]

Note 9    Aughet (1919) 13 Cr App R 101, confirmed, obiter, in Treacy v DPP [1971] AC 537, per Lord Diplock. See CP 156, para 9.10.    [Back]

Note 10    Protocol 7, Art 4(1) applies only to “proceedings under the jurisdiction of the same State”. See para 3.7 above.    [Back]

Note 11    See Adams v Cape Industries plc [1990] Ch 433.    [Back]

Note 12    See Jaffey on the Conflict of Laws (1997) pp 166–170;Habib Bank v Mian Aftab Ahmed, The Times 2 November 2000, per Carnwath J.    [Back]

Note 13    Archbold 2001, para 4–130.    [Back]

Note 14    Treacy v DPP [1971] AC 537, 559, per Lord Diplock; DPP v Stonehouse [1978] AC 55, 90, per Lord Keith.    [Back]

Note 15    Newland [1988] QB 402.    [Back]

Note 16    R v Manchester City Stipendiary Magistrate, ex p Snelson [1977] 1 WLR 911.    [Back]

Note 17    R v Bennet and Bond, ex p Bennet (1908) 72 JP 362.    [Back]

Note 18    Dabhade [1993] QB 329.    [Back]

Note 19    R v Grays JJ, ex p Low [1990] QB 54.    [Back]

Note 20    Ridpath (1713) 10 Mod 152.    [Back]

Note 21    We also included a tenth case, namely the taking of an offence into consideration when sentencing for another offence. As to this, see paras 6.29 – 6.33 below.    [Back]

Note 22    Para 9.18.    [Back]

Note 23    [1993] AC 217.    [Back]

Note 24    Para 9.20.    [Back]

Note 25    [1993] AC 217, 226.    [Back]

Note 26    Archbold and Blackstone.    [Back]

Note 27    It is also interesting that counsel for the defendant, who unsuccessfully argued that sentence could be passed by another judge without the need for a retrial, was Peter Thornton QC of the English bar; his opponent was the DPP for Jamaica who, it may be assumed, was unable to assist the court on practice in England.    [Back]

Note 28    Howard (1991) 92 Cr App R 223, 227. See also Nicholson (1948) 32 Cr App R 127.    [Back]

Note 29    Van Droogenbroeck v Belgium (1982) 4 EHRR 443, 454. See also B v Austria (1991) 13 EHRR 20 for confirmation and further analysis of this definition.    [Back]

Note 30    See para 6.20 above.    [Back]

Note 31    [1964] AC 1254.    [Back]

Note 32    [1998] QB 356.    [Back]

Note 33    This principle appears to subsume the older and narrower rule in Elrington (1861) 1 B & S 688; 121 ER 170, that a person who has been acquitted or convicted of an offence may not later be charged with a more serious offence arising out of the same facts: see para 2.20 above. However, neither Elrington nor Connelly precludes a second prosecution for a more serious offence where the facts constituting that offence were not in existence at the time of the earlier acquittal or conviction, eg where D is convicted of assault and the victim subsequently dies from the injuries sustained.    [Back]

Note 34    Attorney-General for Gibraltar v Leoni, Criminal Appeal No 4 of 1998, judgment given 19 March 1999, unreported; see para 2.18 above.    [Back]

Note 35    Indictment Rules 1971, r 9, which is substantially the same as its predecessor, Indictment Rules 1915, r 3.    [Back]

Note 36    This refers to Indictment Rules 1915, r 3; see n 35 above. (footnote added)    [Back]

Note 37    [1958] 1 WLR 312. A High Court judge sitting at first instance preferred a voluntary bill to effect the joinder of two separate indictments. It is not clear from the report whose application it was that the two be joined. (Footnote added)    [Back]

Note 38    Para 2.21, n 43.    [Back]

Note 39    There was disagreement between counsel as to whether this had always been the Crown’s case on these counts, or whether it was presented in this way as a deliberate device for the purpose of defeating the application to stay the proceedings, by artificially playing down the similarity between the two sets of charges.    [Back]

Note 40    Unreported ruling, 19 September 1996.    [Back]

Note 41    A328-C (1995); see para 3.11 above.    [Back]

Note 42    1998-V p 1990; see para 3.13 above.    [Back]