Centre for Criminal Law


Ivey v Genting Casinos – What it does (and what it does not)

24 January 2018

By Mark Dsouza


In 1982, the Court of Appeal in R v Ghosh [1982] QB 1053 took it upon itself to explain how a jury should go about determining whether a person had acted dishonestly for the purposes of the Theft Act, 1968. It said the jury should ask itself two questions viz.,

1. Was what the defendant did dishonest according to the ordinary standards of reasonable and honest people? And if so,

2. Did the defendant realise that reasonable and honest people would regard what he did as dishonest?

If both questions were answered in the affirmative, the defendant could be said to have acted dishonestly.

This clearly was not a ‘definition’ of the term dishonesty – it is far too self-referential to be a definition. It relied on the jury bringing to the table an intuited sense of what it is to be dishonest in order to apply both limbs of the test. After all, as the Criminal Law Revision Committee said in 1966, “Dishonesty is something which laymen can easily recognise when they see it”.

But despite not being terribly illuminating, in the absence of any competing suggestions, over the next 35 years, the Ghosh test came to be used across the length and breadth of the criminal law, wherever questions of dishonesty had to be answered.

The civil law, in the meantime, went its own way with dishonesty. In a plethora of cases including Royal Brunei Airlines v. Tan [1995] 2 AC 378 and Barlow Clowes International v Eurotrust International [2005] UKPC 37 the civil test for dishonesty was explained in the following terms:

“Although a dishonest state of mind is a subjective mental state, the standard by which the law determines whether it is dishonest is objective. If by ordinary standards a defendant’s mental state would be characterised as dishonest, it is irrelevant that the defendant judges by different standards.”

Then, 30 years after the decision in Ghosh, Mr Ivey walked into a London casino, and proceeded, in less than 24 hours of play, to lay the foundations for a radical change to the criminal law’s conception of dishonesty.

The issue in Ivey

Mr Ivey knew that certain brands of playing cards have a very slightly asymmetrical pattern on their backs, and saw that this casino was using one such brand. So, with the help of an accomplice, over a series of games, Mr Ivey persuaded croupier to indulge his ‘luck-inducing superstitions’, by turning some (good) cards along the shorter edge and other cards along the longer edge. Over a period of time, the deck that the croupier was using became so arranged that Mr Ivey could, by looking at the back of a card, discern whether it was likely to be a ‘good’ card or a ‘bad’ card for the game he was playing. He then proceeded to bet larger sums on subsequent games, and carried on, until the casino decided to use a new set of cards. By that time, he’d won £7.7 million.

However, he was refused payment after the casino pored over video footage of all his games and finally figured out how Mr Ivey had gamed the casino. So Mr Ivey filed a civil suit against the casino demanding to be paid, and the casino resisted on the basis that he had cheated. To decide on the soundness of the casino’s allegation, it was necessary to consider whether Mr Ivey had been dishonest.

The Supreme Court’s decision

This civil case made its way through the court system and finally, in 2017, landed up before a 5 judge bench of the Supreme Court. Although this was a civil case, the Supreme Court, in Ivey v Genting Casinos [2017] UKSC 67 decided that there was no logical or principled basis for the civil and criminal law to have different tests for dishonesty [para 63]. It therefore resolved to set the record on dishonesty straight once and for all in respect of both, the civil, and the criminal law.

The Supreme Court chose to apply what was previously the civil law test for dishonesty (taken from Royal Brunei Airlines and Barlow Clowes International) across the board. Although the Supreme Court identified a few problems (some arguably more pressing than others) with the Ghosh test, its principal objection to the Ghosh test was that it gave rise to the counter-intuitive result 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” [para 58].

What Ivey does not do

In the immediate aftermath of the Ivey judgment, it was widely reported that the Supreme Court had dropped the second limb of the Ghosh test. This is plainly an oversimplification. The principal problem that the Supreme Court identified with the Ghosh test did indeed arise because of the second limb thereof, but the Supreme Court’s solution was far subtler than merely amputating the said second limb. Instead, it adopted the test that already applied in civil law contexts and explained it in these terms:

“…the fact-finding tribunal [or in the case of a criminal trial, the jury] must first ascertain… the actual [subjective] state of the individual’s knowledge or belief as to the facts. The reasonableness or otherwise of his belief is a matter of evidence (often in practice determinative) going to whether he held the belief, but it is not an additional requirement that his belief must be reasonable; the question is whether it is genuinely held. When once his actual state of mind as to knowledge or belief as to facts is established, the question whether his conduct was honest or dishonest is to be determined by the fact-finder by applying the (objective) standards of ordinary decent people. There is no requirement that the defendant must appreciate that what he has done is, by those standards, dishonest” [para 74].

And what it does

For a person wondering what practical effect this restatement of the test for dishonesty has in criminal cases, the main thing to note is what the Supreme Court means when it requires the jury to ascertain the defendant’s actual state of knowledge or belief as to ‘the facts’. It appears that what the Supreme Court means by ‘the facts’ was not quite the same as what the Court of Appeal in Ghosh had in mind in relation to the first, objective limb of what became known as the Ghosh test. This emerges from a careful reading of para 60 of the Ivey judgment, where the Supreme Court considers the following example drawn from Ghosh:

Foreigner: A man who comes from a country where public transport is free visits Britain. On the first day of his visit, he travels on a bus and gets off without paying. He never had any intention of paying, because he assumed public transport was also free in Britain. His mind is clearly honest; but his conduct, judged objectively by what he has done, is dishonest.

The Court of Appeal in Ghosh thought that because the foreigner would not have realised that reasonable and honest people would regard what he did as dishonest, he should not be found to be dishonest in law. In other words, it believed that the foreigner would be saved by what became the second limb of the test it laid down, but not by the first limb. Recall that in setting up the example, the Court of Appeal made it clear that the foreigner’s ‘conduct, judged objectively by what he has done, is dishonest.’

The Supreme Court however insists that the same result would have followed even under the first limb of the Ghosh test, because, “in order to determine the honesty or otherwise of a person’s conduct, one must ask what he knew or believed about the facts affecting the area of activity in which he was engaging. In order to decide whether this visitor was dishonest by the standards of ordinary people, it would be necessary to establish his own actual state of knowledge of how public transport works. Because he genuinely believes that public transport is free, there is nothing objectively dishonest about his not paying on the bus.”

Even if the Supreme Court is correct about the potential for reading the first limb of the Ghosh test in this way, this was clearly not how the Court of Appeal in Ghosh itself read it. It seems to have had in mind a purely objective assessment of the conduct of the defendant, in which no account was taken of what the defendant knew or believed to be permissible modes of behaviour. That said, the Supreme Court’s analysis does reveal its own approach to performing the first task that it says that the jury must perform when deciding on questions of dishonesty.

In the Supreme Court’s view, when identifying the defendant’s ‘knowledge or belief as to… facts’, the jury must take note also of the defendant’s knowledge and beliefs as to the normative standards applicable in a society, including apparently, legal normative standards, with the presumable exception of those imported by the requirement not to be dishonest itself. In other words, the defendant’s mistaken belief about the content of the law (apart from the law on dishonesty) is permitted to work in her favour in deciding whether she was dishonest.

This is certainly consistent with s2(1)(a) of the Theft Act 1968 as well. But, in adopting this view of ‘the facts’ to be taken note of, has the Supreme Court just smuggled the second limb of the Ghosh test into the first task that it says a jury must perform when deciding on questions of dishonesty? Not quite. There are at least two cases that would be decided differently under the Ivey test and the Ghosh test.

The first one, identified by the Supreme Court itself, related to the facts of R v Gilks [1972] 1 WLR 1341. In that case, the defendant went to a bookmaker to claim his winnings on a series of bets. To his delight, he was overpaid by about £106, because the bookmaker mistakenly thought that the defendant had bet on a different (winning) horse in one race, than the one he had actually backed. The defendant realised the mistake of course, but kept the money anyway. When he was asked why, he explained that in his opinion, although it would be dishonest to pocket such an overpayment from his grocer, bookmakers were different. If the bookmaker made such a mistake, he said, there was nothing dishonest about pocketing the overpayment.

The Supreme Court noted that Mr Gilks’ statement, at face value, was a statement not only about his own subjective standards of honesty, but also about what he perceived to be the general standard of honesty. In other words, if we believed Mr Gilks, we would have to accept that he subjectively did not realise that reasonable and honest people would regard what he did as dishonest, since he subjectively thought that reasonable and honest people would agree with him that bookmakers were ‘fair game’. So if the Ghosh test applied, the second limb of that test would exonerate Mr Gilks of dishonesty, essentially because of how warped his own standards of honesty (as well as views about everyone else’s standards of honesty) were.

On the other hand, under the Ivey test, the jury first would ascertain what Mr Gilks knew about the situation viz., that he had actually won only £10.62, and was being paid £117.25 because the bookmaker had misread his bet. It would then apply the (objective) standards of ordinary decent people to decide whether his conduct (i.e. keeping the extra £106.63) was dishonest. A properly instructed jury would be very likely to conclude on these facts that Mr Gilks was dishonest. This, no doubt, is a preferable outcome.

The second case that would be decided differently under Ghosh and Ivey is a more speculative one, involving a special kind of Rogue. This rogue performs some conduct thinking (correctly, as it happens) that she is acting dishonestly. The rogue might nevertheless also harbour the (incorrect) belief that everyone else (or more specifically, honest and reasonable people) would not see her actions as dishonest. Under the Ghosh test, the rogue would not be dishonest, despite her subjective realisation to the contrary. She would however be found to have acted dishonestly under the Ivey test. Admittedly, this is an unlikely set of facts, but experience with the criminal law seems to suggest that it is only a matter of time before unlikely sets of facts arise. But Ivey now has us forearmed.

In summary then, after Ivey, a person like Mr Gilks, who harbours incorrect normative beliefs about what is dishonest behaviour, is no longer excluded by the test for dishonesty. Neither is the rogue who has the correct set of normative beliefs about what is dishonest behaviour, and simply chooses to act contrary to them (while also having an incorrect belief about everyone else’s normative standards of honesty). On the other hand, the foreigner who does something that appears to be dishonest because he harbours incorrect beliefs on matters other than what constitutes dishonest behaviour, continues (rightly) to be excluded by the test for dishonesty. These seem to be positive developments in the law of dishonesty.

Other points of interest

Two interesting side points remain. Firstly, whereas in R v Hinks [2001] 2 AC 241, the House of Lords was quite happy for the civil and criminal law relating to the property rights to diverge, the Supreme Court in Ivey appears to have taken a far less permissive stance. It suggests that where the same concept is relevant both to the civil and the criminal law, the same test ought to apply unless there is a good ‘logical or principled basis’ for the divergence [paras 57(4) and 63].

And secondly, even after Ivey, we are no closer to a definition of dishonesty. But perhaps that is, as Hart put it, the ‘predicament… of the man who says, “I can recognise an elephant when I see one but I cannot define it”’. And if all the jury has to do is to recognise the elephant in this metaphor, then why bother with a definition?

Postscript: But what about Mr Ivey and his £7.7 million? Well, to cut a long story short, the definition of dishonesty adopted by the Supreme Court brought home to Mr Ivey one of the abiding rules of gambling: the house always wins.