The Author


About this blog

Welcome to the 5PB Blog where you can find comments and thoughts on current legal topics.

Examining the criminal test for dishonesty in the wake of Ivey v Genting Casinos [2017]


It is quite rare that the single judgment of a court can be said to completely reformulate a whole area of the criminal law; it is rarer still that a civil judgment is the cause of such a paradigm shift. However, the decision of the Supreme Court in Ivey v Genting Casinos (UK) Ltd t/a Crockfords [2017] UKSC 67 is being hailed as such a case, following the Court's conclusion that "there are convincing grounds for holding that the second leg of the test propounded in Ghosh does not correctly represent the law and that directions based upon it ought no longer to be given" [74]. 


The case concerned a claim brought by Phillip Ivey, a professional gambler, against Crockfords, a casino. In 2012, Ivey won £7.7m at Crockfords playing Punto Banco, a variant of baccarat. Ivey, with the help of an accomplice, exploited a technique known as "edge-sorting", which, in simple terms, allows an eagle-eyed player to identify cards on the basis of sub-millimetric differences between the patterns printed on the reverse side of the deck. This technique, if used correctly, can alter the odds significantly in the player's favour; indeed, Ivey managed to reduce the house edge by some 8%, and enjoyed two hugely successful sessions of Punto Banco at the expense of the casino. 

Crockfords, having realised that Ivey's success had been predicated upon his use of edge-sorting, refused to pay out his £7.7m in winnings on the basis that the use of edge-sorting amounted to cheating. Both parties agreed that there was an implied term in the gambling contract that neither party would cheat. Furthermore, Ivey freely admitted having used the edge-sorting technique. What Ivey insisted, however, was that edge-sorting constituted legitimate gamesmanship, and that because he did not subjectively possess any dishonest intention, his play could not amount to cheating. 

Ghosh Dishonesty 

In criminal proceedings, the tribunal of fact must be satisfied of two criteria if they are to find an accused to be dishonest. This test, which was originally propounded in R v Ghosh [1982] EWCA Crim 2, requires the tribunal of fact to ask two questions: 

1). Whether the conduct complained of was dishonest by the objective standards of ordinary reasonable and honest people (the objective limb); and 

2). If the answer to the first question is yes, the tribunal of fact must then ask whether the defendant himself realised that what he was doing was, by those standards, dishonest (the subjective limb). 

Only if both of these limbs are satisfied can the tribunal of fact find the accused to be "dishonest". 

Background to the Appeal 

The High Court 

At first instance, Mitting J. ruled in favour of Crockfords ([2014] EWHC 3394 (QB)). Although the judge found on the facts that Ivey honestly did not believe that his conduct amounted to cheating (and that he was therefore not subjectively dishonest), he was satisfied that, in a civil context, the standard of whether conduct amounted to cheating was objective, and that he was "unconvinced" that the civil concept of cheating required dishonesty [45]. 

The Court of Appeal 

Ivey appealed to the Court of Appeal which, by a majority of 2-1, dismissed his appeal ([2016] EWCA Civ 1093). 

Arden L.J. disagreed with Mitting J.'s conclusion that there is a "civil" concept of cheating, and held that the definition of "cheating" for the purposes of the implied term in a gambling contract must be the same as the definition of "cheating" for the purposes of the criminal offence under section 42 of the Gambling Act 2005 [60]. However, Arden L.J. rejected Ivey's argument that cheating required an element of dishonesty, stating that she "did not consider that dishonesty is a necessary ingredient of the criminal offence of cheating"[48], and ultimately held that the use of edge-sorting did constitute cheating. 

Tomlinson L.J. also dismissed the appeal, but chose to do so without reference to the criminal offence of cheating under section 42 of the Gambling Act 2005; he stated that "we should not on this civil appeal espouse a meaning of section 42 of the Gambling Act 2005 which might prove insufficiently certain in its application in the criminal jurisdiction" [104]. 

Sharp L.J., dissenting, was of the opinion that dishonesty is an essential element of cheating, and that the Ghosh test was the appropriate test for determining dishonesty in such a situation: "there is no difficulty in determining the correct test to be applied by the trier of fact in any case where it is necessary to give a dishonesty direction. It is that identified in R v Ghosh," [138]. 

Ivey proceeded to appeal to the Supreme Court. The Supreme Court noted that the case "raises questions about[1]: 

1). The meaning of the concept of cheating in gambling; 

2). The relevance to it of dishonesty; and 

3). The proper test for dishonesty if it is an essential element of cheating." 

The Judgment of the Supreme Court 

In respect of the first and second questions raised, the Supreme Court held that dishonesty is not an essential element of cheating [48]. Lord Hughes held that dishonesty "cannot be regarded as a concept which would bring to the assessment of behaviour a clarity or certainty which would be lacking if the jury were left to say whether the behaviour under examination amounted to cheating or did not," [48], and agreed with Mitting J.'s conclusion that the test for whether conduct amounts to cheating is objective [48]. 

However, it is the Supreme Court's analysis of the proper test for dishonesty that is most remarkable. Lord Hughes - with whom the four other Justices agreed, albeit without giving reasoned judgments of their own - highlighted what he described as the "serious problem" [57]with the two-limb Ghosh test, focusing particularly on the fact that "the less the defendant's standards conform to what society in general expects, the less likely he is to be held criminally responsible for his behaviour," [58]. Lord Hughes also argued that juries struggle with the application of Ghosh, in that they "often find it puzzling and difficult to apply," [57(3)]. 

Lord Hughes concluded his analysis of Ghosh dishonesty by articulating that "the second leg of the test propounded in Ghosh does not correctly represent the law", and opined that the test for dishonesty in criminal proceedings should be the same as the test used in civil proceedings, as set out by Lord Nicholls in Royal Brunei Airlines Sdn Bhd v Tan and Lord Hoffman in Barlow Cloves v Eurotrust International Ltd [74]: 

1). The tribunal of fact must first ascertain (subjectively) the actual state of the accused's knowledge or belief as to the facts. The reasonableness of his belief is a matter of evidence going to whether he held the belief, but it is not an additional requirement that his belief must be reasonable. 

2). Once the accused's statement of mind as to knowledge or belief as to facts is established, the question of whether his conduct was dishonest is to be determined by the tribunal of fact, by applying the (objective) standard of ordinary decent people. 

Crucially, Lord Hughes explicitly stated that there is "no requirement that the defendant must appreciate that what he has done is, by those [objective] standards, dishonest," [74]. 

The Current State of Dishonesty in Criminal Proceedings? 

The judgment of the Supreme Court has already been the subject of much debate, not just in respect of the merits of the decision and the accompanying analysis, but also as to whether the Court has in fact overruled the Ghosh test for dishonesty. Although the prevailing view among the legal press is that Ivey has overruled Ghosh, this author's view is that however emphatic Lord Hughes' analysis was, the Court's consideration of the Ghosh test did not form a necessary part of its conclusion, and that Lord Hughes' rejection of Ghosh dishonesty was therefore only obiter

To address this point fully, it is necessary to examine the statutory and common law authorities regarding a court's jurisdiction, and which judicial comments are capable of forming the ratio of a decision. Under section 40(5) of the Constitutional Reform Act 2005, the Supreme Court is empowered to "determine any question necessary to be determined for the purposes of doing justice to an appeal to it under any enactment." As for what constitutes the ratio for a decision, Buxton L.J's judgment in R (Kadhim) v Brent London Borough Council Housing Benefit Review Board [2001] provides helpful guidance: "The ratio decidendi of a case is any rule of law expressly or impliedly treated by the judge as a necessary step in reaching his conclusion, having regard to the line of reasoning adopted by him." 

Therefore, questions that are not necessary for the determination of appeal, however interesting, are not questions which can form part of the ratio of the Court's judgment, and therefore cannot form the basis of any precedent set by the Court. 

In light of the above, can the Supreme Court's analysis and rejection of Ghosh dishonesty form part of the ratio for its decision, and therefore bind lower courts? It is submitted that it cannot. In Ivey, both the High Court and the Court of Appeal correctly identified the core issues as being whether the use of edge-sorting constituted "cheating", and whether "cheating" required any element of dishonesty. Neither party appears to have contended that the Ghosh test for dishonesty was incorrect. The High Court did not consider it necessary to resolve the issue, and nor did the Court of Appeal; the core issues were identified by Arden L.J. as being those mentioned above [([2016] EWCA Civ 1093: 1, 37]. 

Before the Supreme Court, the appeal was properly disposed of by reference to these core issues. Lord Hughes held that edge-sorting did constitute cheating, and that dishonesty was not an essential element of cheating. The Supreme Court needed to go no further to dispose of the appeal. Indeed, Lord Hughes even conceded that the consideration of the appropriate test for dishonesty was not necessary to dispose of the appeal: "[Mitting J.] did not get to the question of dishonesty and did not need to do so" [75]. In this respect, Lord Hughes was correct: the reformulation of a doctrine of criminal law, in a civil appeal, in relation to a point which neither party had contested and which was not necessary for the determination of the appeal, went beyond the scope of the issues of which the Supreme Court was seized. It cannot be said that the adequacy of Ghosh dishonesty was a "question necessary to be determined for the purposes of doing justice to [the] appeal", and Lord Hughes' analysis of Ghosh dishonesty therefore cannot be said to constitute "a necessary step in reaching his conclusion". 

There are other subtle indications that the Supreme Court did not intend to overrule Ghosh dishonesty. For example, had the Supreme Court intended to overrule the Ghosh test for dishonesty, one might expect one of the four other Justices to provide a reasoned judgment in support - especially when the Court's mandate for overruling Ghosh dishonesty was already limited. It is surprising that not one of the other Justices did so. 

Furthermore, a look at the particular words chosen by Lord Hughes lends further credence to the argument that the Court did not intend to overrule Ghosh dishonesty. Lord Hughes stopped short of expressly holding that the Ghosh test for dishonesty should no longer be applied; at paragraph 74, he concludes his analysis passively by stating "These several considerations provide convincing grounds for holding that the second leg of the test propounded in Ghosh does not correctly represent the law and that directions based upon it ought no longer to be given." Although this point is certainly not conclusive in and of itself, in light of the above it corroborates the view that the Supreme Court did not intend to overrule Ghosh dishonesty at this time, and that it was simply laying the groundwork to overrule it at a later date, in the context of a more appropriate case. 

Either way, the Supreme Court's judgment puts practitioners and the judiciary in an uncomfortable position regarding dishonesty-based offences. If the prevailing view is correct, and the Ghosh test for dishonesty has been overruled, then certain dishonesty-based offences will need to be viewed through a different prism. However, even if Lord Hughes' analysis was merely obiter, then the issue will inevitably be brought before the appellate courts in the coming weeks and months to be determined properly. 

The Impact of the New Test on Dishonesty 

Presuming that the Supreme Court's decision is binding, what impact will this have on the test for dishonesty in criminal proceedings? Conceptually, it will represent a significant departure from the current position; the abolition of the second, subjective limb in Ghosh and the increased focus on objective standards of honesty will shift the criminal law's emphasis away from penalising a culpable state of mind, and towards penalising a culpable course of conduct. While previously the law only sought to penalise individuals who appreciated that their conduct fell below objective standards of honesty, Ivey will punish those whose conduct falls below the objective standard of honesty, irrespective of whether the individual appreciated this standard. 

In theory, the new test makes it easier for prosecutors to prove dishonesty, as they will no longer have to prove that not only was the accused dishonest, but that he also appreciated that his conduct was objectively dishonest. 

In practice, however, the change may not be as remarkable as first suspected. First, "dishonesty" as an issue does not come before the tribunal of fact very frequently, even in the context of dishonesty-based offences. For example, in theft and robbery trials, it is very rare for a defendant to mount a defence based on the fact that their conduct was not dishonest, and a new Ivey test for dishonesty will have very little practical impact in those cases. Where such a test might have more of an influence, however, is in relation to fraud and deception-based offences, where the event in question is often admitted, but a defence is mounted on the basis that the defendant's conduct was not dishonest. 

Second, Lord Hughes' test does not do away with all considerations of the accused's subjective thoughts, merely the question of whether the accused genuinely believed that his conduct was honest. The tribunal of fact will still have to ascertain the accused's subjective knowledge and belief of the facts; this analysis will now just take place under the first limb of the test. Interestingly, it is distinctly possible that juries will struggle with this fusion of objective standards of dishonesty and the defendant's subjective knowledge and belief of facts, which is particularly ironic given the Supreme Court's concern that juries find the Ghosh test "puzzling and difficult to apply," [57(3)]. 

Third, it is unlikely that Lord Hughes' reformulated test for dishonesty will lead to an increase in the number of convictions of "morally depraved" defendants who genuinely believed that their conduct was in line with objective standards of honesty, and who may have theoretically escaped liability under Ghosh. Theoretically speaking, it may be the case that the more morally depraved an individual is, the more likely he is to get an acquittal under the Ghosh test - but it also the case that the more morally depraved an individual is, the less likely it is that the tribunal of fact will believe that he genuinely believed his conduct to be honest. Lord Hughes' objection to Ghosh on the basis that it allows morally depraved individuals to escape conviction is simply not borne out in practice, and the abolition of the subjective limb of Ghosh is unlikely to have any tangible impact on this. 

Concluding Thoughts 

The fact that the Supreme Court has ostensibly overruled Ghosh and proposed a move towards a more objective test of dishonesty is of clear conceptual significance, and it is made all the more interesting by the fact that such a seismic shift in the criminal law has taken place against the backdrop of a civil case. However, in practice it is possible that the decision in Ivey v Genting Casinos may not be as revolutionary as first thought. 


By Chris Jenkins who is currently undergoing pupillage at 5pb.

Posted by 5PB on 03 November 2017 at 15:59