WHEN the ambit of the criminal law is narrowed judicially, what impact should this have on convictions previously secured under the disavowed, broader rules? This question was considered by the Court of Appeal in Johnson and Others [2016] EWCA Crim 1613, [2017] 1 Cr. App. R. 12.

The Court was faced with several appeals based on the decision in Jogee and Ruddock [2016] UKSC 8; [2016] 2 W.L.R. 681 (Jogee – noted Dyson [2016] C.L.J. 196). Before Jogee was decided in February 2016, D1 could be held liable as an accessory for a foreseen collateral crime B (e.g. murder) committed by D2 in the course of committing an agreed crime A (e.g. burglary). In Jogee, it was decided that this “parasitic accessorial liability” (PAL) was mistakenly introduced in 1984 by Chan Wing-Siu [1985] A.C. 168 (discussed in Stark [2016] C.L.J. 550). The correct position, it was decided, was that D1 could be liable for the murder committed by D2 only if D1 had intentionally assisted or encouraged D2 intentionally to cause at least grievous bodily harm (GBH). Foresight that D2 may intentionally cause GBH was no more than evidence of D1’s intention to encourage or assist D2’s offending.

In Johnson, the Court distinguished between two main categories of appellants (more complex circumstances will be ignored here for reasons of space). First, defendants who managed to appeal following the decision in Jogee within 28 days of their own convictions will succeed if those convictions are rendered “unsafe” by Jogee (Criminal Appeal Act 1968, ss. 2, 18(2)). Johnson confirms that safety will be compromised where a direction in accordance with Jogee could realistically have made a difference to the jury's decision.

Secondly, defendants wishing to appeal out of time (i.e. more than 28 days after their convictions) must demonstrate additionally that not granting leave to appeal would result in “substantial injustice” or “substantial injury” (Hawkins [1997] 1 Cr. App. R. 234, 240). It is uncontroversial, doctrinally, that a change in the criminal law is not itself sufficient to constitute “substantial injustice”. Johnson indicates that, post Jogee, the relevant additional factors will be: (i) “the strength of the case advanced that the change in [Jogee] would, in fact, have made a difference” at trial; and (ii) whether the defendant was guilty of “other, though less serious, criminal conduct” (at [21]). This second criterion is potentially misleading. First, in PAL cases, the underlying conspiracy to commit crime A will always have been present. Secondly, in murder cases, the second criterion might suggest that if, post Jogee, the defendant would not have been found liable for murder, but would necessarily have been found liable for manslaughter, he might not demonstrate “substantial injustice”. In considering the appeal in Hall (one of the cases heard with Johnson), however, the Court explained that a defendant who presents a “sufficiently strong” case that he would not have been convicted of murder will have suffered a “substantial injustice”, even if he would – on the facts that were found by the jury – necessarily be liable for manslaughter under Jogee. In such circumstances, the defendant's murder conviction would be substituted for a manslaughter conviction (at [184], [191]). In short, the Court's second criterion seems superfluous, at least in murder cases: the matter of “substantial injustice” appears to turn exclusively on whether the defendant can present a “sufficiently strong” argument that a Jogee-compliant direction would have resulted in his being acquitted of murder.

Arguably, further superfluity can be found in the Court's treatment of what happens after “substantial injustice” is made out. The question of whether the conviction is safe will apparently then have to be addressed separately. The Court nevertheless envisaged it as unlikely that, if “substantial injustice” was present, the conviction would be safe (at [23]). Ultimately, in out-of-time appeals concerning murder convictions, all emphasis will therefore be placed on the court's assessment of the strength of the argument that Jogee would have made a difference to the defendant's liability.

Despite the nuanced distinction between in- and out-of-time appeals, Johnson indicates that all appellants will encounter difficulties in showing that Jogee could have made any difference to their liability. One reason for this is that many accessorial liability cases did not employ PAL. In Hore, one of the appeals heard with Johnson, the question left for the jury was whether the defendant had been a party to an agreement intentionally to cause GBH to the victim, whom he had deliberately lured to the scene of the attack. The word “foresight” appeared in the trial judge's direction simply in relation to which crime the defendant thought he was intentionally helping the principals to commit. This was, the Court recognised, a hopeless appeal. A mere reference to foresight in a judicial direction is thus an insufficient ground for appealing based on Jogee.

Those appeals heard in Johnson that engaged the two-crime analysis characteristic of PAL, and where the jury was directed in terms of foresight of crime B, were scuppered by the Court's willingness to use D1’s foresight to find a conditional intention on his part to assist or encourage D2 intentionally to cause (at least) GBH, if such violence proved necessary. Jogee is clear that such conditional intention will suffice for accessorial liability for murder. If such a conditional intention is found on the facts, then it will not be possible to argue that manslaughter was a more appropriate verdict.

A good example of the Court's broad approach to conditional intention is the appeal in Hall. That the defendant “foresaw that [D2] would attack [V] … with intent to cause really serious bodily injury” was found to be consistent with a finding of conditional intention to assist or encourage the intentional causing of GBH (at [189]). Here (despite the Court's vague reference to “the circumstances” of the case), conditional intention looks suspiciously like foresight of crime B combined with a decision to continue with the original plan to commit crime A, namely PAL. If conditional intention to assist or encourage offending can easily be found in such cases, the impact of Jogee on past convictions will be limited: few will be unsafe, and few instances of “substantial injustice” will be found, because the jury will (having applied Chan Wing-Siu, etc.) have been sure that the defendant foresaw crime B's possible commission in the course of crime A's commission. Johnson may also suggest that, in the future, convictions for murder could legitimately be returned routinely – on the post-Jogee basis of conditional intention – in cases where the defendant foresaw the possibility of GBH being caused intentionally, and continued with an underlying criminal plan nonetheless. This will be particularly the case where the defendant knew that the principal was armed (e.g. at [21]). In short, Jogee may have changed little.

This is not to say that accessorial liability for murder will always, on appeal, be found where the intentional causing of GBH is foreseen as an offshoot of another agreed crime. The Court suggested in Johnson that Jogee most plausibly makes a difference to cases where crime A was one “not involving intended violence or use of force” whereas crime B was (at [21]). Such cases, although much discussed in the academic literature, make up a small number of reported cases, and – presumably – cases of accessorial liability overall. Thus, although Jogee is not neutered by the approach to conditional intention in Johnson, it is tamed, unless and until conditional intention becomes a more demanding standard.

It is worth noting that delay in bringing an appeal following a change in the law makes “substantial injustice” more difficult to establish (Bellinger [2005] 2 Cr. App. R. 29), and that the Criminal Cases Review Commission must, when contemplating a referral, take into consideration the Court of Appeal's conservative approach to granting leave out of time (Criminal Appeal Act 1968, s. 16C). Accordingly, the Court of Appeal will probably not be kept busy for long following Jogee. Indeed, the avoidance of a tide of appeals based on a change in the law is what motivates the courts' approach to “substantial injustice”. The practical need to balance securing individual justice for defendants and “finality and certainty in the administration of criminal justice” has been stressed repeatedly (Cottrell [2008] 1 Cr. App. R. 107, [42]). Johnson also prizes the ability of the appellate courts to change (or “correct”) the criminal law without fear of overwhelming the appellate system (at [18]). These practical concerns (which may be alleged to miss the point, particularly when the claim is that the law was wrongly applied since 1984) motivated the Supreme Court and Privy Council in Jogee to sound a warning shot across the bows of potential defendants (at [100]). Johnson fires another. (Defendants in Northern Ireland have received their own: Skinner et al. [2016] NICA 40.)

The broad approach to conditional intention in Johnson allowed the Court to avoid real engagement with the injustice at the heart of its bifurcated approach. Regardless of one's views on whether Jogee was correct to kill off PAL, the difference in sentencing (and labelling) between murder and manslaughter is vast. The sole concern in murder cases (whether in in-time or in out-of-time appeals) should be whether Jogee might plausibly have resulted in a conviction for manslaughter. This unified approach would reach a more appropriate balance between finality and individual justice. The Court's approach to conditional intention also allowed it to avoid a connected issue: whether the statements regarding manslaughter in Jogee were correct. If the defendant intended to assist or encourage an attack causing actual bodily harm (or a non-violent offence), and the principal killed the victim in an attack intended to cause GBH (or worse), did the defendant intentionally assist or encourage the act that caused death? If not, on what basis is the defendant liable for manslaughter (see Simester (2017) 133 L.Q.R. 73, 86)? The difference between murder and manslaughter is marked, but it is nothing compared to the difference between murder and no liability for homicide.