The New Test for Dishonesty in Criminal Law—Lessons From the Courts of Equity?

Introduction

In criminal law, the issue of dishonesty in certain property offences ¹ is considered a fundamental ingredient ² in finding criminal liability. However, the concept of dishonesty is only partially defined in statute, with s 2 of the Theft Act 1968 providing examples where a person is not acting dishonestly. Namely when he is acting in a belief that he has a right in law, ³ a belief that the other person would consent to his taking of the property ⁴ or that he is acting under a belief that the owner of the property could not be found by taking reasonable steps. ⁵ The statute fails to define the concept of dishonesty, leaving it to the criminal courts to provide their own definition. Elliott suggests that the word ‘dishonestly’ should not, in theory, be a difficult concept to define, however, it has been a case of the judges making a rod for their own back, with anyone trying to pin down ‘dishonesty’ in English case law soon finding that he is aiming at a constantly moving target. ⁶ Until recently the accepted test for dishonesty was set out in R v Ghosh, ⁷ Lord Lane CJ providing a two-stage test where the jury must first of all decide whether ‘according to the ordinary standards of a reasonable and honest people what was done was dishonest’. ⁸ If they believe that this is the case they must then apply a subjective test to determine ‘whether the defendant himself realised what he was doing was by those standards dishonest’. ⁹

A contrast between the criminal and civil law approaches to dishonesty emerged, by virtue of cases such as Royal Brunei Airlines v Tan, ¹⁰ Barlow Clowes v Eurotrust ¹¹ and Abou-Rahman v Abacha ¹² where an objective approach to dishonesty was favoured by the latter. ¹³ Civil law has not been without its own problems in this area however, as the judiciary struggled to come to terms to what extent the role of the defendant’s subjectivity should play within the role of the equitable tort of dishonest assistance in relation to a breach of trust. The recent decision in Ivey v Genting Casinos ¹⁴ provided the Supreme Court an opportunity to clarify the civil law test for dishonesty, and in doing so, the Court made obiter statements criticising the criminal approach set out in Ghosh. It is argued that due to the unanimous nature of the Supreme Court’s decision, the reasoning of Ivey v Genting Casinos is to be adopted by the criminal law and has created a uniform test for dishonesty in the civil and criminal courts. ¹⁵

This article will consider the confusion encountered when balancing the objective test with the subjective knowledge of the defendant which has historically maligned the equitable tort of dishonest assistance and ask whether the decision in Ivey v Genting Casinos has provided a satisfactory resolution to this problem. With an apparent unification of the civil and criminal tests for dishonesty, it is important to ask what lessons can be learnt from the civil law when applying an objective test for dishonesty in criminal law going forward.

The Ghosh Test

Prior to Ghosh, the case of Feely ¹⁶ had set the criminal law test for dishonesty purely on objective terms. Lawton LJ finding that the jury were entitled to adjudicate whether ‘the man who takes money from a till intending to put it back and genuinely believing on reasonable grounds that he will be able to do so’ ¹⁷ was acting honestly based on ‘the current standards of ordinary decent people’. ¹⁸

Ghosh ¹⁹ concerned a surgeon acting as a locum consultant at a hospital who falsified claims in order to obtain money that he was owed for other work that he had carried out previously. The Court of Appeal considered whether he had the requisite mens rea of dishonesty under the offence of obtaining property by deception contrary to s 15 of the Theft Act 1968. In rejecting Feely, Lord Lane CJ argued that if Parliament, in its use of the term ‘dishonesty’ in the Act, intended to describe a state of mind, then ‘the knowledge and belief of the accused are at the root of the problem’. ²⁰ A man visiting from another country where public transport is free, travelling on a bus without paying in this country, would be unfairly dishonest if judged objectively ²¹ ; his lordship concluding that ‘Parliament cannot have intended to catch dishonest conduct…. to which no moral obloquy could possible attach’. ²² The prospect of an unadorned objective test was resisted on the basis that ‘if the mind of the accused is honest, it cannot be deemed dishonest merely because members of the jury would have regarded it as dishonest to embark on that course of conduct’. ²³

Griew ²⁴ points out that the objective limb in Ghosh deviates from that in Feely with Lord Lane replacing the term ‘ordinary decent people’ in favour of ‘honest people’. ²⁵ Although this may be a matter of mere linguistics, it was not necessarily helpful in maintaining clarity in this regard. Regardless, he question’s Lord Lane’s interpretation of the Feely test on the basis that his lordship seems to do away with a holistic view of the conduct of the accused with Griew arguing that his should be done ‘in the context of the state of mind in which he did it’. ²⁶ The implication is that the Ghosh direction considers the dishonesty of the actions alone, independent of the defendant’s frame of mind at the time. As Griew puts it:

The question is not…whether the employee who takes money from a till…is dishonest in doing so and therefore guilty of theft; the question is rather he is guilty of theft, because dishonest, when he ‘takes money from the till intending to put it back and genuinely believing…that he will be able to do so’. ²⁷

A simple subjective test, on the other hand, is problematic when dealing with circumstances such as those in R v Gilks ³³ where the accused believed that when dealing with a bookmaker he was entitled to keep any winnings paid to him in error and that there was nothing dishonest in doing so. Although Williams suggests that the objection to the approach is ‘the judges’ supposition that the defendant was entitled as a matter of law to set his own standards’, ³⁴ ‘subjectivism of this degree gives subjectivism a bad name’. ³⁵ He reasons that the ‘subjective approach to criminal liability, properly understood, looks to the defendant’s intention and to the facts as he believed them to be, not to his system of values’. ³⁶ Professor Smith agrees, adding that the rule tends to abandon all standards other than that of the accused himself, in the determination of his responsibility. ³⁷

The introduction of the hybrid, two-stage test of Ghosh seemed a compromise on behalf of Lord Lane CJ to mitigate cases such as Gilks. By including reference to the ordinary standards of reasonable and honest people in the first limb of the Ghosh test, his lordship was confident that:

Robin Hood or those ardent anti-vivisectionists who remove animals from vivisection laboratories are acting dishonest, even though they may consider themselves to be morally justified in doing what they do, because they know that the ordinary people would consider these actions to be dishonest. ³⁸

The outcome in Ghosh removes the strident nature of the objective test while also preventing a thief’s charter provided by a purely subjective test but in doing so, created confusion in the form of the second limb of the test. ⁴¹ Halpin highlights that the appeal of the Ghosh test is that it appears to strike an effective compromise but ‘[u]pon further examination…we find within it not a stable compromise but a continuing tension between the subjective and objective approaches…’. ⁴² Mellisaris labels the test as being ‘vague and indeterminate’. ⁴³ Campbell warns that ‘no one should be seduced into thinking that it is a test of pure social fact…it is a partially idealised test with a necessary component of moral evaluation which will vary from jury to jury’. ⁴⁴ Glover, in agreeing with earlier criticisms by Griew, ⁴⁵ suggests that as a ‘consequence of the nebulous nature of the Ghosh test, and reliance on the assumption of jurors’ innate knowledge of dishonesty, the law appears uncertain and unpredictable’ ⁴⁶ and therefore contravenes the rule of law. ⁴⁷ It is suggested that the test’s second limb is not necessary as juries should take into account the defendant’s circumstances while applying the first limb anyway as a standard of dishonesty is not considered in vacuo. ⁴⁸

Spencer takes exception to the retention of a subjective limb of the test suggesting that this allows a defendant to advance a defence of mistake of law by arguing that he believed society would have tolerated his behaviour, ⁴⁹ concluding that for ‘the courts to take their criminal law from the Clapham omnibus is one thing; to take it from the man accused of stealing is quite another’. ⁵⁰ Williams notes that ‘Lord Lane may have considered his judgement as a rescue operation, but if so, it is a rescue that still leaves this heroine in considerable peril’. ⁵¹ Wasik, supported a two-stage approach suggesting that if a jury were to consider the standards of ordinary people, it would not be contradictory in requiring them to take into account whether the accused believed his conduct was honest or not. ⁵² In its report on Fraud, ⁵³ The Law Commission concluded that ‘[m]any years after its adaption, the Ghosh test remains, in practice, unproblematic’. ⁵⁴ Indeed, Spencer had argued at the time that ‘sooner or later, the question of dishonesty is bound to make its appearance in the House of Lords’, ⁵⁵ where the slate could be wiped clean and a new definition for dishonesty be provided. ⁵⁶ Yet, the Ghosh test remained the accepted test for the next 35 years until the Ivey decision swept it aside.

Dishonesty in Civil Law—Dishonest Assistance

In Barnes v Addy, ⁵⁷ Lord Selborne recognised that strangers to a trust can be personally liable to account as a constructive trustee to beneficiaries of a trust for any loss caused if they assist a breach of trust ‘with knowledge in a dishonest and fraudulent design on the part of the trustee’. ⁵⁸ This concept of ‘knowing assistance’ moved towards ‘dishonest assistance’ in Royal Brunei Airlines v Tan ⁵⁹ where, when considering dishonesty in this context, Lord Nicholls purported to put forward an objective test for dishonesty stating that ‘dishonestly…means simply not acting as an honest person would in the circumstances. This is an objective standard’. ⁶⁰

Despite this clear statement, confusion creeped into the law due to his subsequent comments referencing a subjective element, suggesting that

Honesty…does have a strong subjective element in that it is a description of a type of conduct assessed in the light of what a person actually knew at the time, as distinct from what a reasonable person would have known or appreciated. ⁶¹

Yet, Lord Hutton in Twinsectra v Yardley ⁶³ disagreed. In analysing the possible approaches to civil dishonesty, he identified three possible options, the subjective approach (Robin Hood test), the objective approach and a combined test similar to that set out in Ghosh. ⁶⁴ His interpretation of the Tan test was this must be a version of the combined (Ghosh) test. He cited the statement from Lord Nicholls that ‘[u]ltimately…an honest person should have little difficulty in knowing whether a proposed transaction…would offend the normally accepted standards of honest conduct’, ⁶⁵ highlighting the use of the word ‘knowing’ as being crucial in this context. His lordship stating that the use of the word would be ‘superfluous if the defendant did not have to be aware that what he was doing would offend the normally accepted standards of honest conduct’. ⁶⁶ Hudson describes Lord Hutton’s approach as an awkward ‘combination of an alteration of [the] Tan test and yet a purported approval of it by the House of Lords’ ⁶⁷ adding that by emphasising the role of knowledge in the test is ‘unfortunate because the test for dishonesty was developed precisely to move away from tests of knowledge’. ⁶⁸ Lord Millet however, dissenting, argued that Lord Nicholls had purposely avoided adopting the Ghosh approach ⁶⁹ and had instead adopted an objective test which would be applied after considering the defendant’s experience, intelligence and knowledge of the circumstances at the time. ⁷⁰ This, however, did not make it necessary that ‘he should actually have appreciated that he was acting dishonestly; it is sufficient that he was’. ⁷¹ Although Lord Millet was unable to persuade the rest of the Court in this case, his interpretation of Lord Nicholls’ words is preferred in academia over the much maligned approach taken by Lord Hutton. ⁷² As Ryan puts it, Lord Millett’s approach ‘arguably reflected more accurately the approach…in Royal Brunei, who had…expressly rejected any requirement of consciousness on the part of the defendant that his conduct would be considered dishonest by ordinary persons’. ⁷³

In Barlow Clowes v Eurotrust, ⁷⁴ the Privy Council seemed to reaffirm Lord Nicholls’ test for dishonesty in Tan. ⁷⁵ In explaining away the distortion created by Twinsectra, Lord Hoffman suggested that he had not departed from the test in Tan and that:

The reference to ‘what he knows would offend normally accepted standards of honest conduct’ meant only that his knowledge of the transaction had to be such as to render his participation contrary to normally acceptable standards of honest conduct. It did not require that he should have had reflections about what those normally acceptable standards were. ⁷⁶

Lord Hoffman’s, apparent, reinstatement of the test from Tan was endorsed by the Court of Appeal in Abou-Rahman v Abacha ⁸² but in doing so, the test was inadvertently distorted. LJJ Rix ⁸³ and Pill ⁸⁴ purported the test to include subjective knowledge on one’s dishonesty. Lady Justice Arden, in her analysis, declared that the decision in Barlow Clowes v Eurotrust ‘represented the law of England and Wales’ ⁸⁵ ; however, she, just as Lord Hutton had done previously, placed weight into Lord Nicholls’ obiter statement regarding the knowledge of the defendant. Her ladyship concluded that although the test was predominantly objective ‘there are also subjective aspects of dishonesty’. ⁸⁶ Halliwell and Prochaska sum up the frustrations when they state that ‘it is a great pity that the fundamental principles that [Lord Nicholls] adopted have been relegated to an inordinate degree’ ⁸⁷ adding that ‘[i]t seems yet again that semantic distinctions have denigrated the substantive test’. ⁸⁸

The Ivey Test for Dishonesty

In Ivey v Genting Casinos, ⁸⁹ concerning a gambler who cheated at Punto Banco Baccarat, the Supreme Court considered the role of dishonesty under s 42 of the Gambling Act 2005. Although the court concluded that dishonesty was not a necessary ingredient for cheating, Lord Hughes took the opportunity to flesh out the legal concept of dishonesty in English law. His Lordship was highly critical of the criminal law test for dishonesty set out in Ghosh stating that it benefited defendants with warped standards of honesty; too much reliance is unnecessarily placed on the defendant’s state of mind; it provides a confusing test for jurors to understand; it has resulted in a divergence between dishonesty in civil and criminal law and that the court in Ghosh were not required to make such a ruling. ⁹⁰ This adds to academic criticism of the test where it is argued that the second limb of the Ghosh test complicates criminal trials and can result in inconsistency due to additional grounds that can be contested. ⁹¹ This is particularly problematic if faced with jurors, required to apply their own standards of honesty, whose standards may not live up to those of ‘ordinary decent people’ creating the danger of asking jurors to apply higher standards than they themselves attain, resulting in hypocrisy. ⁹² Griew is also unconvinced that the Ghosh test rids the law of the undesired ‘Robin Hood’ defence where a defendant may claim to be ignorant that the conduct was dishonest by ordinary standards ⁹³ and it allows mistakes as to the law to be a valid defence. ⁹⁴

In Ivey v Genting Casinos, Lord Hughes added to this criticism, stating that the ‘principle objection to the second leg of the Ghosh test is that the less a defendant’s standards conform to what society in general expects, the less likely he is to be held criminally responsible for his behaviour’. ⁹⁵ His lordship arguing that ‘[t]here is no reason why the law should excuse those who make a mistake about what contemporary standards of honesty are…’ ⁹⁶ regardless of the context. Although Lord Hughes acknowledged Lord Lane’s basis for the second limb of the Ghosh test being that criminal responsibility for dishonesty must be based on the accused’s actual state of mind, he rejected the presumption that the objective approach to dishonesty would preclude this. His lordship argued that ‘[w]hat is objectively judged is the standard of behaviour, given any known actual state of mind of the actor as to the facts’. ⁹⁷

Lord Hughes expressed preference for the civil test of dishonesty set out in Royal Brunei Airlines v Tan and adopted in Barlow Clowes v Eurotrust, reasoning that ‘there can be no logical or principled basis for the meaning of dishonesty…to differ according to whether it arises in a civil action or a criminal prosecution’. ⁹⁸ Lord Hughes purported to adopt the civil test created by Lord Nicholls in Tan but explained it in the following terms:

When dishonesty is in question the fact-finding tribunal must first ascertain (subjectively) the actual state of the individual’s [genuine] knowledge or belief as to the facts…When once his actual state of mind as to knowledge or belief as to facts is established, the question [of dishonesty] is to be determined 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. ⁹⁹

Although Ivey was dealing with a civil appeal, Dyson and Jarvis argue that ‘the real reason the case of Ivey found its way to the UKSC was nothing to do with cheating at cards and everything to do with a desire to dispense with part of the Ghosh test for dishonesty’. ¹⁰² Ivey now represents the civil law position in terms of dishonesty. Although, in technical terms, Ghosh still remains the leading criminal law precedent, it seems logical, following the same justifications which allowed the Privy Council decisions in Tan and Barlow Clowes to overrule the Court of Appeal in Twinsectra, that Ivey will be adopted in future by the criminal courts too. As Laird puts it:

It is trite law that the only part of a judgment that is binding is the ratio. However, the tone of the Supreme Court’s judgment strongly suggests that it intends for its obiter analysis of Ghosh to be treated as binding by lower courts. ¹⁰³

Conclusion

Having changed the test for dishonesty, Clough suggests that this deals with the common criticism of Ghosh in that the second limb could create some absurd jury decisions. ¹⁰⁹ The new Ivey approach leaving ‘less room for manoeuvre’. ¹¹⁰ Likewise, Galli suggests that the ‘judgement is likely to be welcome for practitioners and jurors alike, but not those looking to rely on the distorted test for determination of dishonesty in order to circumvent the purpose of the law’, ¹¹¹ arguing that the Ghosh test allowed the defendant’s belief ‘to undermine an objective standard’ ¹¹² and the change removes that ‘vulnerability’ ¹¹³ by eliminating the subjective limb of the test. Virgo ¹¹⁴ contests this, suggesting that ‘there is no evidence from the cases that juries found the Ghosh direction difficult to apply’. ¹¹⁵

The mode of the change of law in Ivey also attracts widespread criticism. Laird expresses surprise that not only Ghosh was rejected but more so in that this took place in a civil case. ¹¹⁶ He points out that Ghosh was not even considered during the appellate history of the Ivey case and as such, ‘Ghosh and its effect upon the criminal law was not subject to detailed scrutiny at any stage of the proceedings in Ivey’ ¹¹⁷ adding that this ‘perhaps explains some of the problematic omissions evident in the Supreme Court’s judgement’. ¹¹⁸ First is the presumption by Lord Hughes, in his judgment, that an ‘unprincipled divergence’ ¹¹⁹ between the civil and criminal tests was problematic. Dyson and Jarvis brazenly suggest that the ‘implication that there could be a principled divergence between civil and criminal concepts is welcome’. ¹²⁰ They add that ‘it is not clear that the function of dishonesty in the criminal law is the same as the function of it in the civil law…’ ¹²¹ and that the civil judiciary can deal with complexity whereas a lay juries require simplicity. ¹²² Spencer ¹²³ makes similar arguments stating that: ‘As unsuccessful defendants in criminal proceedings face consequences far worse than do unsuccessful parties to a civil action, a more generous interpretation of the dishonesty requirement in their favour can hardly be described as “an unprincipled divergence”’. ¹²⁴ Virgo adds that:

Different tests of dishonesty could be justified because civil law dishonesty determines unacceptable conduct in order to impose liability, whereas dishonesty in the criminal law is concerned with identifying culpability, which requires consideration of the defendant’s mental state. The effect of Ivey is to treat dishonesty in the criminal law as a mechanism for assessing conduct rather than culpability, albeit that the defendant’s knowledge or belief about the facts is relevant to this objective assessment. ¹²⁵

This change in direction must also be considered in the light of cases such as R v Hinks ¹³¹ which considered the actus reus of theft and whether the appellant had appropriated £60,000 and a television set when she encouraged her 53-year-old male friend, Mr Dolphin, who was described as naive, gullible and of limited intelligence ¹³² ; to donate the gifts to her. Rose LJ confirmed that there was a distinction between the two separate ingredients of appropriation and dishonesty and that ‘appropriation can occur even though the owner has consented to the property being taken’. ¹³³ The result of this was that the entire issue of whether Ms Hinks had committed theft therefore rested on whether the jury could be satisfied that she had acted dishonestly when she persuaded Mr Dolphin to give her the gifts. As suggested by Beatson and Simester, ¹³⁴ in order for this to have been a civil wrong, the transaction must have been induced by a misrepresentation, duress or undue influence. ¹³⁵ But despite the absence of any of these inducements, as Beatson and Simester put it, ‘[i]t appears…that such despicable conduct, though with no civil consequences, may constitute a crime’. ¹³⁶ In essence, the offence of theft hangs on the dishonesty element, and, as a direct consequence of Hinks and Ivey, there is no longer a need to the defendant to either take another’s property adversely nor are they required to realise what they are doing is dishonest. Spencer, who had previously been critical of the Ghosh test of dishonesty admits that ‘…36 years later, and on the other side of Gomez and Hinks, I confess that I have changed my mind’ ¹³⁷ adding that ‘…the innocence of anyone who genuinely believes his conduct to be proper by the ordinary standards of honest and reasonable people can be seen as an important limit; and rejecting it extends the offence of theft yet further’. ¹³⁸ It is hard not to agree with Virgo’s concerns when he suggests that theft is now:

[A] crime which requires neither proof of harm nor subjective fault. Together with Hinks, Ivey has resulted in an unacceptable expansion of the criminal jurisdiction, one which is inconsistent with the civil jurisdiction and so constitutes an unprincipled divergence between criminal and civil law. ¹³⁹

It has already been observed that when the civil courts have been faced with a simple objective test, they have struggled to maintain this standard and in the cases of Twinsectra, Barlow Clowes v Eurotrust and Abou-Rahman v Abacha a subjective element crept back into the respective judgments. The case of R v Hayes, ¹⁴³ decided under the Ghosh test and prior to the change brought about by Ivey, is of particular interest. Here, a market trader was convicted of conspiracy to defraud when he manipulated the London Inter-Bank Offered Rate (LIBOR), to his advantage. It was considered whether, when applying the objective limb of the Ghosh test, the jury could consider the general ethos of the banking system and evidence of market practice. In rejecting this, the Court of Appeal suggested that this would ‘gravely affect the proper conduct of business’. ¹⁴⁴ They did, however, concede that this evidence would be ‘plainly relevant to the second subjective limb’. ¹⁴⁵

However, in Hussein v FCA, ¹⁴⁶ the Ivey test was applied when considering whether a trader had acted dishonestly when he influenced his employers LIBOR submissions to their advantage. The case was based around communication made by H to the bank’s LIBOR submitter and by providing information on his individual trading position this had improperly influenced the submissions. H argued that the conversations were merely about internal hedging opportunities to protect his clients from interest rate fluctuations. He argued that he did not believe the conduct to be improper and believed it to be good practice at the bank during this period. Judge Herrington, accepted that:

as a consequence of the test now formulated in Ivey…. the subjective element must be on what Mr Hussain knew about the definition of LIBOR…. in other words what did Mr Hussein genuinely believe were the factors that could be taken into account in determining the objective LIBOR rate…In that context it must also be taken into account what Mr Hussein believed as regards to how the information he provided…would be taken into account in determining UBS’s LIBOR submissions. ¹⁴⁷