Case No. EWCA-Crim-2366

wn CourtHHJ TestarRoyal Courts of JusticeStrand, London, WC2A 2LL Before : LADY JUSTICE HALLETT VICE PRESIDENT OF THE CACD MR JUSTICE HAMBLENandHHJ GOSS, RECORDER OF NEWCASTLE. Between : Miss CP Montgomery QC and Mr. A MacDonald for the 1st AppellantRH Christie QC and Mr. Simon Baker for the 2nd AppellantMr. S Hammond for the 3rd AppellantMr. J Waddington for the RespondentHearing date : 20th November 2013Judgment The Vice President of the Court of Appeal Criminal Division Lady Justice Hallett DBE: Background 1.On 14 June 2012 the three appellants were convicted in the Crown Court at Southwark of a major Value Added Tax (“VAT”) fraud involving a total loss to the public purse of £39 million. The mechanism of the fraud was a Missing Trader Intra Community (‘MTIC’), with carbon credits as the purportedly traded commodity. The prime organisers were said to be the appellant Sandeep Dosanjh and his second cousin Pardeep Dosanjh (who fled the jurisdiction before arrest). Gill and Chahal were variously described as “organisers under the Dosanjh family” and “lieutenants”.2.On 18 June 2012 the trial judge HHJ Testar sentenced each of them for the offence of conspiring to cheat the public revenue. Dosanjh received a sentence of 15 years imprisonment, Gill 11 years’ imprisonment and Chahal 9 years’ imprisonment. They were each disqualified from acting as directors for 12 years. The co-accused Dhanvinder Singh Basra, Sandeep Harry, Pritpal Singh and Kernjit Gill Dhillon were all acquitted by the jury. 3.The prosecution case against the appellants was that they were involved in manipulation of the EU Emissions Trade Scheme. The scheme was set up pursuant to the Kyoto Protocol to help reduce ‘Green House Gases’ and it regulated trade in carbon allowances or ‘credits’. A ‘carbon credit’ or European Union Allowance (‘EUA’) was the right to emit one tonne of carbon into the atmosphere. Polluting companies were issued with a number of carbon credits to cover a particular period, but could sell surplus credits or buy more credits on the market. 4.The appellants were involved in the running of companies that formed two artificial ‘trading chains’ through which the fraud operated. At the bottom end of each chain was a ‘missing trader’, a company that defaulted on its VAT liability. There were then two ‘buffer’ companies in each chain. At the apex of each chain was a company owned and operated by the appellant Sandeep Dosanjh, called KO Brokers Ltd. 5.KO Brokers Ltd was incorporated in 2006 and registered for VAT in 2008 but had little legitimate business activity prior to and after the fraud. The fraudulent trading took place over a total of 69 days between January to May 2009 and the total VAT owed to the UK Treasury by the missing traders was approximately £39 million (41,039,261 Euros). 6.The first chain involved carbon credits being acquired on a zero-rated basis from other EU traders by Mak & Co UK Ltd which sold them on to Helios Technical Services Ltd. Helios sold them onto Infiniti UK Ltd and Infiniti sold them on to KO Brokers Ltd. KO Brokers sold them onto the open market to ‘blue-chip’ companies such as BP, Shell and Gazprom at a profit. 7.Gill was involved with both Infiniti and Mak. There was little legitimate business activity in either company before the start of the fraud. Chahal set up a replacement for Mak in case it was needed. The first trade through chain one was on 20 January 2009 and the final trade was on 8 April 2009. A total of 19,478,000 carbon credits were traded during this period. In total a sum of 240,214,116 Euros (of which 31,332,276 were VAT) was paid on invoices into this chain by KO Brokers. Mak, as a ‘missing trader’, defaulted on the VAT due in the sum of 31,244,175 Euros. 8.The second chain involved credits being acquired on a zero-rated basis from other EU traders by Swift Enterprises Ltd that sold them on to Allianz Group Ltd. Chahal supervised the operation of Swift. Gill introduced one of his companies AGH Associates Ltd into the Swift chain, as a ‘buffer’, to be run by his wife. Chahal was a director of Allianz that sold credits on to AGH. They, in turn, sold them on to KO Brokers Ltd, which sold them onto blue-chip companies. 9.The first trade through chain two was on 2 April 2009 and the final trade was on 6 May 2009. The trading ended at this point because AGH Associates Ltd lost their registration for VAT. A total of 5,283,000 credits were traded during this period. In total, the sum of 77,964,745 Euros (of which 10,169,315 was VAT) was paid on invoices into the chain by KO Brokers. Swift was a ‘missing trader’ and defaulted on the VAT due in the sum of 9,795,086 Euros. 10.The proceeds were swiftly moved offshore into ‘banking platforms’, in particular, commercial banks in Hong Kong, Australia and New Zealand. These operated in a manner described as being ‘analogous to a solicitor’s client account’. All the money went into one account but there were internal ledgers used to allocate it between the ‘sub-accounts’ of the traders. This meant the true nature of the transfers was effectively disguised and difficult to detect.11.All three appellants were of previous good character and were each arrested on 19 August 2009. Dosanjh was 31 years of age. He gave a prepared statement in interview denying any involvement in a conspiracy to defraud and then declined to answer further questions. Of those charged, he took the lion’s share of the profits which amounted to approximately £6,600,000 (spending a million pounds in cash on one property in London and on a Rolls Royce). At a hearing on 16 October 2013 the judge made an order for £12,887,685 to be confiscated and paid within six months. The default sentence was set at 10 years imprisonment. 12.Navdeep Gill was 33. He too declined to answer questions in interview. His personal benefit was calculated at £309,000. A bundle of 8 character letters were produced on his behalf. 13.Ranjot Chahal was 36. He answered no comment to all questions, but gave a prepared statement denying any involvement. His benefit was calculated at £40,000 half of which was paid to HM Revenue and Customs Grounds of Appeal against Sentence 14.Miss Clare Montgomery QC appeared for Dosanjh and advances two grounds of appeal the first of which was common to all three appellants. Ground 1: wrong in principle. 15.Miss Montgomery insists that a “significant and growing discrepancy” has become apparent in penalties for common law and statutory offences. Unless this “anomaly” is properly addressed by the court, a prosecutor may simply ignore any restrictions imposed by statute and choose to charge an offence where no time limits apply or the penalty is at large. This potentially offends the principles in R v Rimmington; R v Goldstein [2006] 1 AC 459. 16.As the law stands, there are four categories of offence, all of which may apply to the same fraudulent conduct: i)Common law conspiracy to defraud where sentence was originally at large but which now carries a maximum of 10 years imprisonment, see s.12 of the Criminal Justice Act 1987. ii)Common law conspiracy to cheat which was abolished by s.32(1)(a) of the Theft Act 1968 ‘except as regards offences relating to the public revenue’. Sentence remains at large.iii)The fraudulent evasion of VAT contrary to s.72 of the Value Added Tax Act 1994, which carries a maximum sentence on indictment of seven years’ imprisonment.iv)A statutory fraud offence, carrying a maximum sentence of ten years’ imprisonment. 17.Here, the Appellants were convicted of conspiracy to commit the common law offence of cheating the public revenue. Had they been convicted of statutory fraud or VAT offences they would have been liable to maximum penalties of ten and seven years respectively. Miss Montgomery’s theme was that it is wrong in principle to pass a sentence on a common law conspiracy that is longer than the maximum penalty available for the equivalent statutory offence or the cognate common law charge of conspiracy to defraud. 18.In support of this assertion, Miss Montgomery placed heavy reliance upon the decisions in Rimmington and Goldstein. The appellants Rimmington and Goldstein were each charged with the common law offence of public nuisance for sending out racially offensive material in Rimmington’s case and salt in Goldstein’s case. The House of Lords considered the ingredients of the offence of public nuisance and the extent to which it is still known to the common law in the light of statutory developments. 19.Lord Bingham of Cornhill, in the leading speech, reviewed its history. He noted that conduct formerly chargeable as the crime of public nuisance has become the subject of express statutory provision. Where that is so, the general rule should be that conduct which might amount to an offence under both common law and statute should be charged under statute. At paragraph 30 he gave his reasons: “Where Parliament has defined the ingredients of an offence, perhaps stipulating what shall and shall not be a defence, and has prescribed a mode of trial and a maximum penalty, it must ordinarily be proper that conduct falling within that definition should be prosecuted for the statutory offence and not for a common law offence which may or may not provide the same defences and for which the potential penalty is unlimited. … It cannot in the ordinary way be a reason for resorting to the common law offence that the prosecutor is freed from mandatory time limits or restrictions on penalty. It must rather be assumed that Parliament imposed the restrictions which it did having considered and weighed up what the protection of the public reasonably demanded. I would not go to the length of holding that conduct may never be lawfully prosecuted as a generally-expressed common law crime where it falls within the terms of a specific statutory provision, but good practice and respect for the primacy of statute do in my judgment require that conduct falling within the terms of a specific statutory provision should be prosecuted under that provision unless there is good reason for doing otherwise.”20.Miss Montgomery accepted it does not follow that it is necessarily wrong per se for a prosecutor to charge a common law offence where a statutory offence is available. Lord Bingham’s concern related to cases where the decision to do so creates a disadvantage to the defendant, for example by avoiding a statutory time limit or a maximum penalty. However, there must be ‘good reason’ for charging the common law offence. In the context of public nuisance, Lord Bingham considered that ‘the circumstances in which, in future, there can properly be resort to the common law crime of public nuisance will be relatively rare.’ (See paragraph 31). 21.The principles in Rimmington were recently applied in R v Dady [2013] EWHC 475 (QB) in which Coulson J considered a prosecution application to prefer a voluntary bill of indictment. The defendant was alleged to have operated a website that facilitated the illegal downloading of football matches. The draft bill contained three counts: conspiracy to defraud at common law, an offence under s.107(2A) of the Copyright, Designs and Patents Act 1988 and an offence under s.297 of the 1988 Act. Coulson J held that the second count was not properly available on the evidence. At paragraph 21 he considered the relationship between counts 1 and 3 and observed: “I consider that s.297 catches precisely the criminal conduct which is now alleged against Mr Dady. He was providing his subscribers, for a fee, with a means to get round the encryption. For present purposes, I am prepared to accept that this amounted to a conspiracy to breach s.297. In those circumstances, it seems that his offending should have been charged as such. The matters of difficulty created by s.3 of the Criminal Law Act 1977, such as the six month time limit and the need for permission from the DPP, are procedural safeguards which were plainly regarded by Parliament as important. In my judgment, it would be wrong to allow the Crown now to ignore these safeguards, and to charge what would otherwise be a precise statutory offence under the wide common law offence of conspiracy to defraud. In addition, of course, to allow the Crown to prefer this voluntary Bill would be contrary to Lord Bingham’s guidance in R v Rimmington, because it would be depriving the defendant of the protection of being charged with an offence which was summary only.”22.Coulson J was also a member of the Courts Martial Appeal Court in R v Armstrong [2012] EWCA Crim 83. The appellant had been charged with four counts contrary s.69 of the Army Act 1955 (conduct to the prejudice of military discipline). The alleged conduct could also have formed the basis of substantive charges under the Criminal Law: under the Firearms Act 1968, the Misuse of Drugs Act 1971 and the Official Secrets Act 1989. At paragraphs 20 and 21, the Court noted that: “Beyond recording the position of the Crown, it is neither necessary nor desirable that this Court should comment on the scope of comment under s.69, save to say that where conduct constitutes an offence under the ordinary criminal law, it must be charged as such save in wholly exceptional circumstances. As Mr Mably rightly submitted [for the Attorney General], there can ordinarily be no justification for using s.69 in such circumstances and it would be outside the lawful exercise of the prosecutor’s discretion; its use might well circumvent the statutory sentencing regime imposed by Parliament on the courts either to the detriment of the defendant by enabling a harsher sentence to be passed or to the detriment of the public interest in preventing the court passing a sentence within the range specified by Parliament and imposing ancillary orders.”23.Miss Montgomery acknowledged that there is at least one decision of the Court of Appeal in an appeal against conviction for revenue fraud which appears to contradict her argument (namely Mavji [1987] 2 All ER 758) and a whole line of sentencing decisions which undoubtedly contradict it. Nothing daunted, she argued the decision in Mavji, in so far as it conflicts with Rimmington, has been overruled and the subsequent sentencing cases were decided per incuriam (the point having been overlooked). 24.In Mavji the Court of Appeal considered an appeal against conviction for conspiracy to cheat the revenue by way of the non-payment of VAT. Giving the judgment of the court Michael Davies J observed at page 761A-B: “It was submitted on behalf of the appellant that it would be anomalous if the common law and statutory offences stood side by side, with no more having to be proved to establish the former than the latter, and, in the former case, without limit of penalty. We see no anomaly. In our judgment s.38(1) of the 1972 Act was a ‘catch-all’ provision directed specifically to punishing evasion of value added tax when that tax was newly introduced and has no bearing on the general principles of cheating in relation to the public revenue, particularly as there is no counterpart to s.38(1) in the statutory provisions relating to other taxes.” 25.Similar clear statements appear in the line of sentencing decisions. We shall refer to just three. In R v Ward [2005] EWCA Crim 1926, the appellant was convicted of conspiracy to cheat, and argued that the judge ought to have had regard to the 7 year maximum for the statutory VAT offence. The Court rejected this argument at paragraph 16: “It seems to us that the position in cases such as this, where the allegation is that many millions, not merely one million, has been lost to the country’s revenues is that it is entirely appropriate for the court to approach the matter on the basis that a conspiracy to cheat is the appropriate charge and that entitles it to conclude that a sentence in excess of the statutory maximum for the single substantive offence would be available and proper. This court made it clear in Dosanjh and repeated it in Czyzewski that that was a permissible course to take. The judge in the present case was accordingly entitled, as a matter of principle, to conclude that he was not constrained by the statutory maximum for the substantive offences. He had, however, as he did, to take it into account as assisting him in coming to a conclusion as to where he should place the appropriate sentence.”26.In Bright [2008] EWCA Crim 462; [2008] 2 Cr App R(S) 102 the appellant was convicted of conspiring to commit an insurance fraud on evidence which would have supported an indictment for conspiracies to trade fraudulently. The ultimate losses were likely to be £1 billion and 1000 employees lost their jobs. The court, over which the then Sir Igor Judge P presided, held that the sentencing judge was not obliged to sentence on the basis that the lower maximum sentence for conspiracy to trade fraudulently applied. 27.In R v Randhawa [2012] EWCA Crim the Court upheld sentences of 15 and 14 years for appellants who had had an organising role in an MTIC fraud. The court noted the distinction between the sentencing regimes applicable to common law cheating and statutory fraud and observed at para 38: “There are no sentencing guidelines applicable to this case. The guidelines for statutory offences of fraud (carrying a maximum of 10 years’ imprisonment) do not extend to offences of cheating, or conspiring to cheat, the public revenue, for which the penalty is at large. Such offences are reserved for the most serious cases, where a sentence in excess of the statutory maximum for other offences may be appropriate.” 28.Miss Montgomery suggested that the effect of these and other similar observations (for example in the Sentencing Guidelines Council’s Definitive Guideline on Fraud) would mean that a serious case of cheating the public revenue is necessarily more serious than the most serious case of statutory fraud or conspiracy to defraud. Upholding a 15 year sentence in a case of conspiracy to cheat is tantamount to saying that an offender’s conduct is so serious that it calls for a 50% longer sentence than any conceivable case of fraud. She posed the rhetorical question: surely Kallakis (see Kallakis 2013 EWCA Crim 709) who was involved in the UK’s largest ever mortgage fraud (over £740 million advanced) required a more severe punishment than the offenders here?29.It is for Parliament to decide upon maximum and minimum level of sentences, and Miss Montgomery claimed there is a real possibility of courts encroaching on their territory. Where Parliament has set maximum sentences for particular conduct, it is not for the courts and the executive to decide that those sentences are not enough, and that the statutory limits should be evaded. This is precisely the objectionable course of action that Lord Bingham identified in Rimmington. Conclusions on Ground 1 30.We can dispose of this ground relatively swiftly. As Mr Waddington for the Crown, observed, it is contrary to a consistent line of Court of Appeal authority going back over 25 years. We see no reason to dissent from the proposition in the Sentencing Guidelines Council’s Definitive Guideline on Fraud to the effect that “the common law offence of cheating the public revenue is generally reserved for the most serious and unusual offences and where a sentence in excess of the statutory maximum would be proper”. In our judgment that statement coincides with the law and best practice. We are not surprised, therefore, that the SGC’s successor, the Sentencing Council, has adopted a similar approach in its recently published draft guideline. 31.There is no inconsistency with the decision in Rimmington. Both approaches are based on the will of Parliament. In Rimmington the House held that where Parliament has created a statutory offence and defined its ingredients, (possibly providing for time limits on prosecutions, defences and a maximum penalty), it is not for a prosecutor to thwart Parliament’s clearly expressed intention by charging a common law offence to which none of the restrictions apply. 32.Here, however, Parliament has created statutory offences of fraud and conspiracy to defraud to which maximum penalties apply but it has also expressly retained the common law offence of cheating the revenue. The offence was singled out from the general abolition of common law dishonesty offences by section 32(1)(a) of the Theft Act 1968. Further, despite subsequent reviews of the offences of fraud, Parliament has left not only the offence in existence but the penalty at large. This is entirely consistent with the general approach over decades to major frauds on the revenue. They have always been treated as offences of particular seriousness. 33.Thus, we are entirely confident that as far as Parliament is concerned, the offence of conspiracy to cheat the public revenue retains its established and clearly understood role in the prosecution of revenue cases. It is used to supplement the statutory framework and is recognised as the appropriate charge for the small number of the most serious revenue frauds, where the statutory offences will not adequately reflect the criminality involved and where a sentence at large is more appropriate than one subject to statutory restrictions. These are not ‘ordinary’ cases. 34.Lord Bingham’s assertion of what would “ordinarily” be appropriate is premised on the assumption “that Parliament imposed the restrictions which it did having considered and weighed up what the protection of the public reasonably demanded”. It is reasonably to be assumed that Parliament has deliberately left the common law offence of cheating the revenue untouched by statutory changes in this area because it recognised that it was appropriate to do so for the protection of the public, for all the reasons set out in the authorities. That was clearly Parliament’s intent at the time of the Theft Act and there is no reason to believe that their intent has changed. In any event, Lord Bingham recognised that his general approach may not apply where there was good reason for charging the common law offence. In this case, we agree with Mr Waddington, there is “good reason”. Ground 2 Dosanjh: manifestly excessive 35.In the event that her primary argument fails, Miss Montgomery submits that, even if the courts are, in principle, entitled to pass sentences in common law cases which are greater than the statutory maxima for comparable offences, a sentence of fifteen years is manifestly excessive on the facts of his case. 36.Pardeep Dosanjh played a more important role in the conspiracy. A sentence of 15 years for the Appellant would imply a sentence of 17 years or more for him. A sentence of that length should be reserved for cases where an even larger sum of money is involved. Further, she argued it cannot be right that the Appellant received a sentence that was 50% longer than could be passed for any statutory fraud, however grave. 37.The sentence also looks excessive when compared with previous decisions of this court and with the draft guideline from the Sentencing Council. For a category 2 offence (which this would be) and for an organiser (which the appellant was) the range of 8-13 years is proposed with a starting point of 10 years. The starting point is based on the fraud being worth over £30 million pounds. A sentencing range of 10 to 17 years is reserved for frauds of over £50 million. Miss Montgomery did not suggest the draft Guideline is binding but drew to our attention the fact that it is based on a thorough review of current sentencing practice. On that basis, it does provide a useful guide.38.The submissions of both Gill and Chahal repeat and depend to some extent on the success of Miss Montgomery’s submissions. If the benchmark of 15 years for Dosanjh is reduced for any reason they would hope to take advantage of a similar reduction. Ground 3 Gill: disparity 39.Additionally Mr Richard Christie QC for Gill sought to persuade us that the judge wrongly attributed to Gill a greater degree of involvement and culpability than he did to Chahal. He took us to the judge’s sentencing remarks in which the judge observed that there was “very clear and cogent evidence that both Mr Gill and Mr Chahal had an involvement which went over and beyond the managing of the buffer companies that they fronted.” 40.However, there was no evidence to suggest Gill had any involvement in the operation of the missing trader by way of issuing invoices, transferring EUAs on the register or controlling monies through their bank accounts. 41.Without wishing to make the situation for Chahal worse, Mr Christie suggested the evidence against him was more extensive than against Gill, for example Pritpal Singh was put under very considerable pressure by Chahal and the Dosanjh family to become involved. Gill, on the other hand, may have facilitated the introduction of others like his wife and Himat Singh to the running of the companies but he used no pressure. They were each willing participants. On that basis the judge was wrong to reflect the involvement of others as an aggravating feature. 42.Mr Christie addressed the glaring difference between the two cases of Gill and Chahal (the benefit they received) in this way: much of the difference in monies each man received is explicable by the fact that Infiniti traded from 20 January to 8 April 2009 and Allianz only traded from 2 April to 6 May 2009. The longer trading period meant that Infiniti had time to trade a higher volume of EUAs, which necessarily meant a greater income. Whilst Infiniti traded 19,478,000 EUAs, Allianz only traded 5,283,000 EUAs. Chahal intended to obtain far more but was stopped from doing so. 43.Furthermore, the Crown’s case (accepted by the judge for sentencing purposes) was that chain two was ended by the involvement of Her Majesty’s Revenue and Customs and the de-registration of AGH. Hence, but for the involvement of HMRC, chain two would have continued for far longer and Allianz’s receipts from the fraud would have matched (if not exceeded) those of Infiniti. 44.In any event, the judge recognised that “when one is sentencing for fraud, it is always impossible to know exactly how much people have ‘made’ out of a fraud”. The mere fact that Allianz received a lesser sum than Infiniti received, does not mean that Mr Chahal received (or was going to receive) less than Mr Gill. Ground 4 Gill: manifestly excessive 45.Mr Christie questioned whether the judge placed undue reliance on Attorney Generals Reference (88, 89, 90 & 91 of 2006) (Meehan & Others)