Lunn, R v

On 3 December 2015 in the Crown Court at Southwark (before HHJ Goymer and a Jury), the appellant was convicted by unanimous verdicts of four counts of cheating the Revenue (counts 1-4). This was a retrial. At the first trial, which had concluded in January 2014, the Jury had been unable to reach verdicts on counts 1-4 and acquitted the appellant on two further counts of cheating the Revenue (counts 5 and 6).

On 11 January 2016, the appellant was sentenced to 4 years’ imprisonment on count 1, and 12 months on counts 2-4, concurrent to each other but consecutive to the 4-year term on count 1. The total term was 5 years’ imprisonment.

He appeals against conviction and sentenceby leave of the Single Judge.

The appellant ran an accountancy business, both as a self-employed sole trader ‘(trading as Christopher Lunn & Company) and as the sole director and shareholder of a limited company (Christopher Lunn & Company Limited). The two firms were run as one business. For most of the period covered by the indictment the appellant was the only qualified accountant in the business. He hired a number of people with limited or no previous experience in preparing accounts and tax returns; and training was generally given to these people either by the appellant or by someone he had trained.

The prosecution case was that he had acted dishonestly with the result that the Revenue was deprived of tax to which it was entitled.

Count 1 related to accountancy fees charged by the firms to its clients. The accountancy fees were tax-deductible expenses; and the prosecution contended that the firms adopted the dishonest practice of entering standard and non-standard accountancy fees in the accounts and tax returns of clients that were higher than those charged to the clients, so that the figure for accountancy fees presented to the Revenue was falsely overstated and the tax payable by the clients was reduced, with a loss to the Revenue. Although the individual amounts were relatively small, it was a system that was applied over a number of years, across a large client base (which included members of staff) and the overall loss was significant.

The defence was that the appellant was not aware that false invoices representing inflated charges were being produced by his staff; and that in any event he had not been dishonest. It was also contended that he only acted as agent for the clients and it was they who benefitted from the overstatement of accountancy charges and not the appellant or his business. Furthermore, to the extent that figures for accountancy fees were inaccurate, there was no net loss to the Revenue because the full allowance which was available to the clients for the use of their home as an office had not been claimed. If the appropriate allowance had been claimed, it would have offset any overstatement of accountancy fees. These matters were set out in appellant’s defence statement dated 11 February 2013. It will be necessary to return later in this judgment to the last point, referred to in argument as the ‘no loss’ point and also (as we prefer to describe it) as the ‘home as office point’.

Counts 2 and 3 concerned what was said to be the dishonest manipulation of the accounts of two companies: Sno White Ltd (count 2) and Rain White Ltd (count 3). These were originally owned by clients. The appellant took control of the companies at a point where they had ceased to trade, with accumulated trading losses. The prosecution alleged that these trading losses were offset against fictional profits arising from fictional trades carried out with his own business which supposedly purchased services from these companies. The nature of the cheat was that the appellant included the costs of these services in his accounts as tax-deductible expenses, thereby reducing his firms’ tax bills. On count 2 the total loss to the Revenue was said to be £20,200 income tax and £10,100 corporation tax; and on count 3, approximately £39,850 in corporation tax.

The defence in relation to counts 2 and 3 was that the appellant acquired these companies because they had losses that could properly be used to reduce the profits and tax liabilities of his own and other associated businesses. As far as he was concerned, they were legitimate tax-saving schemes. The trade on each was similar to that of the previous businesses and, in any event, he did not act dishonestly.

Count 4 related to what was said to be the dishonest manipulation by the appellant of his firms’ accounts and those of a company called Carte Blanche Design Ltd, a company owned by his wife, by including false entries in his firms’ tax returns relating to services provided by Carte Blanche Design Ltd in order to reduce their tax liability. The total loss was said to be approximately £30,278.80.

The defence in relation to count 4 was that Carte Blanche Design Ltd was paid a reasonable sum for work that was legitimately done. There was no dishonesty and no cheat.

Count 1 was in the following terms.

Statement of Offence

Cheating Her Majesty, the Queen and the Public Revenue, contrary to Common Law.

Particulars of Offence

[The appellant] between 24 March 2003 and 31 December 2011 in connection with Christopher Lunn & Company and Christopher Lunn & Company Ltd cheated Her Majesty the Queen and the Public Revenue of income tax and corporation tax by dishonestly submitting or causing to be submitted tax returns and accounts on behalf of clients which were misleading in that they falsely overstated the amount charged for the accountancy services provided.

A number of issues were raised both at trial and on appeal in relation to the way in which count 1 was framed. First, it was said to be duplicitous. Secondly, because of the long period and the number of clients covered by the indictment, it was said to be contrary to the Criminal Procedure Rules (‘CPR’) and the Criminal Practice Direction (‘CPD’) relating to the drafting of indictments. Thirdly, and for similar reasons, it was contended to give rise to a trial process that was unmanageable and unfair.

The period covered by the count was 8 years and 9 months; and it was on this basis that the defence challenged the indictment at the retrial as it had at the first trial.

On 15 September 2015, the defence made an application that count 1 should either be quashed, or alternatively stayed as an abuse of process.

The Judge noted that it was similar to an application made in the first trial, and refused the application for the reasons he had given then. However, he noted that he had required the prosecution in the first trial to select specific transactions on which it relied, so that the trial would be manageable and the Jury could focus on the issues, and would keep the matter under review.

In his ruling in the first trial the Judge had voiced justifiable displeasure that the application had been made on the first day of the trial, rather than at an earlier case management hearing, but recognised that it was a matter he had to deal with.

The Judge summarised the arguments on each side. The defence argued that count 1 was defective either for duplicity or for lack of particularity, and that it was faced with an impossible position because there were so many transactions relied on and so many different issues that arose. It would be difficult for the Judge to sum up; and if there were a conviction it would be impossible to know on what basis the Jury had convicted. The lack of particularity made it unfair for the defence and constituted an abuse of the Court’s process.

The prosecution contended that the framing of count 1 did not give rise to difficulty. The prosecution was alleging a course of conduct by the appellant when acting on behalf of his clients. It was a form of systemic dishonesty which had operated throughout the indictment period whose effect was to defraud the Revenue.

In his ruling the Judge identified three types of ‘duplicity’. First, where, on the face of the count, more than one offence was charged. Second, where, although the count appeared good on its face, the evidence established that more than one offence was committed, sometimes referred to as ‘divergence or departure’. Third, where the Judge, in the interests of justice, requires the prosecution to amend a count or elect on which, among many, counts to proceed.

The Judge identified the complaint as being about lack of particularity (the third category) and referred to rule 14.2 of the 2010 Criminal Procedure Rules and the relevant Practice Direction. He reminded himself that, even where the framing of a count was consistent with the Rules and the Practice Direction, the Court must be vigilant to ensure a fair trial. The defendant would have to know the case he had to meet, and the Court would have to understand from the Jury’s verdict, at least in general terms, the basis on which they had convicted.

The Judge considered that of the four elements set out in the CPD (which we set out below). Three of these applied. He referred to the case of Litanzios (1999) Crim LR 667 as support for the proposition that a count charging a cheat ‘over a substantial period of time’ was not objectionable where it related to a course of conduct, providing the details of the cheat were made clear. The longer the period, the more essential that the conduct was particularised. In the Judge’s view the real issue on the application was whether count 1 was so lacking in particularity that the defendant could not properly and fairly answer the allegations. The Judge expressed himself ‘wholly unpersuaded’ that there was any injustice to the defendant or an abuse of the Court’s process in the form of the count. Nevertheless, the prosecution would not be permitted to broaden the evidential scope and the matter would be kept under review.

The offence of Cheating the Public Revenue derives from the common law misdemeanour of cheating. Historically, the offence was broadly defined by reference to ‘deceitful practices’ which either defrauded, or endeavoured to defraud, another by dishonest means. For a history and analysis of the offence, see Hudson [1956] 2 QB 252, Archbold 2017 §25-409 and Smith and Hogan Criminal Law 14th ed. §24.5.

Two linked points may be noted. First, to the extent that the distinction is relevant to identifying the actus reus, the offence of Cheating the Revenue can be either a ‘conduct offence’ or a ‘result offence’. Secondly, in the former case the prosecution does not need to prove an actual loss, see for example Hunt [1994] STC 819 at 826-827 and Attorney-General’s Ref No.8 of 2012 [2012] EWCA Crim 1730 at [46]. In the present case the prosecution took on the burden of proving a net loss to the Revenue and it is not clear why it did so. It may be that it felt that a loss to the Revenue would support its case on intent and dishonesty, or that it would inform the sentencing process if there were a conviction. In any event, the Judge directed the Jury that the prosecution bore the burden of proving a net loss to the Revenue in the light of ‘the home as office’ point.

So far as the arguments on duplicity and abuse of process are concerned, we consider that Judge’s ruling was correct and that Mr Cox QC was right not to pursue the points as such.

Duplicity is directed to the form of the indictment and count 1 did not charge more than one offence: it charged a continuing and systemic cheat. A stay of proceedings is generally directed to where it is necessary to protect the integrity of the criminal justice system or where it is impossible for the defendant to have a fair trial, see Archbold 2017 §4-75. In the present case the Judge exercised his case management powers with a view to ensuring that the appellant was not prejudiced in his defence and that his trial was fair. We consider below the criticisms of his trial management decisions.

Mr Cox was on firmer ground in his criticism of the period (8 years and 9 months) and the practical effect of the number of transactions covered by count 1. He submitted that the count was objectionable because the appellant did not know the case he had to meet and the period covered by the charge, and was, in any event, contrary to principles of elementary fairness, see for example, Archbold 2017 §1-237.

For the prosecution, Ms Johnston QC submitted that the number of routine overstatements of accountancy fees could not properly be reflected in counts reflecting the cheats in respect of individual clients. Such an indictment would have run to many hundreds of counts and would have resulted in the overloading of the indictment contrary to good practice.

Charging and the case management issues that arise from allegations of multiple offending have given rise to much discussion since the decision of this court in Canavan, Kidd & Shaw [1998] 1 Cr App R 79.

In Barton v. DPP [2001] EWHC Admin 223, a defendant had been charged on an information alleging the theft of £1,338.23 over a period of years. The Divisional Court (Kennedy LJ and Astill J) reviewed the authorities on ‘continuous offences’; and concluded that, although the individual appropriations were each capable of being separately identified, it was permissible to charge the whole course of conduct as a continuous offence because the defendant put forward the same defence in relation to all the takings.

The issues were the subject of a Law Commission Report: ‘Effective Prosecution of Multiple Offending’, Law Commission No.277 CM 5609, presented to Parliament in October 2002. The Report highlighted the problems of repeat offending with too many individual offences to be accommodated in a single trial, and referred to the ‘pragmatic arrangement’ of charging offences treated as specimens of a wider range of offending that was ‘thrown into disarray’ by the decision in Canavan, Kidd and Shaw. Paragraph 4 of the Report’s executive summary identified two fundamental principles: first, that a defendant should only be sentenced for that which he has admitted, or which has been proved following a trial in which the evidence can be examined; and secondly, that it should be possible to sentence for the totality of an individual’s offending; and a defendant should not escape appropriate punishment because the procedure cannot accommodate it.

The subject has continued to give rise to difficulty, see for example Tovey [2005] EWCA Crim 530, Grout [2011] 1 Cr. App R 38, and the discussion in Archbold 2017 §5-91/94.

At this stage of the analysis the issue is primarily one of practicality and fairness, rather than sentencing, and this is addressed in the Criminal Procedure Rules.

The rule which applied at the material time was CPR rule 14.2(2). It is now rule 10.2(2) and, for convenience we shall refer to the most recent version of the rules and practice direction. Rule 10.2(2) provides:

More than one incident of the commission of an offence may be included in a count if those incidents taken together amount to a course of conduct having regard to the time, place or purpose of commission.

CPD II §10A.11 (previously PD §14.A.10) is in the following terms.

Crim PR 10.2(2) allows a single count to allege more than one incident of the commission of an offence in certain circumstances. Each incident must be of the same offence. The circumstances in which such a count may be appropriate include, but are not limited to, the following:

(a)

the victim on each occasion was the same, or there was no identifiable individual victim as, for example, in a case of the unlawful importation of controlled drugs or money laundering;

(b)

the alleged incidents involved a marked degree of repetition in the method employed or in their location, or both;

(c)

the alleged incidents took place over a clearly defined period, typically (but not necessarily) no more than about a year;

(d)

in any event, the defence is such as to apply to every alleged incident. Where what is in issue differs in relation to different issues, a single ‘multiple incidents’ counts will not be appropriate (though it may be appropriate to use two or more such counts according to the circumstances and to the issues raised by the defence).

Further guidance is provided by the observations of this Court in Hartley [2012] 1 Cr App R 7, Hughes LJ (Vice President) at [22].

Where specific incidents are capable of identification … then ordinarily we would expect the indictment to contain a count referable and identifiably referable to that event so that the jury can determine it. That is subject to not, if there are hundreds of them, overloading the indictment with more counts than the jury can be expected to determine. Generally, it is necessary for those who are framing indictments to pay attention to any issues flagged up by what the defendant has said either in interview with the police or later in a defence statement. Ordinarily we would suggest where there is simply a complaint of a course of conduct over a period of months, often years, more than a single count for each period is usually appropriate, although one per year may well suffice if the alleged period is extended. But the overall principle is simply that regard must be had in an intelligent way to the possible views of the case at which a jury might arrive and to the position of the judge in due course should there be convictions. If thought is given to those questions we have little doubt that it will normally be possible to frame an indictment in a manner which enables the sentencing to be realistic and complies with the strict rules of law as set out in R v Canavan.

In the present case, the prosecution case was summarised in draft spread sheets which were said to show either (a) the charging of no fee for accountancy services and a claim for the costs of those services in the relevant accounts or claims presented to the Revenue, or (b) a claim for the costs of the services which was significantly higher than the amount actually charged and paid for those services. The draft spreadsheets identified the clients or members of staff and the relevant figures (including the arithmetical differences) in a readily comprehensible form. This was neither an uncommon nor an objectionable way of dealing with a charge of fraud or cheat.

The alleged transactions taken together plainly amounted to a course of conduct within the meaning of CPR Rule 10.2(2) and was compliant with that rule. The victim was the Revenue and the transactions involved a marked degree of repetition in the method used. It would have been wholly impractical to reflect the alleged criminality by charging individual cheats based on individual clients and their accounts and tax returns over the number of years, since such an indictment would have run to many hundreds of counts. Nevertheless, we accept that a long period and large number of transactions covered by a single count may give rise to particular trial management and sentencing issues. We address these below.

So far as trial management was concerned, we cannot see that there would have been any great advantage in fragmenting the indicted period into counts covering a single year. The trial management issues in this type of case (and in this particular case) would be as likely to occur if there had been nine counts, divided up into what might have been arbitrary periods, as they would in a single count, see for example, Litanzios (above). We are fortified in this conclusion by the qualifying words in the practice direction, ‘not limited to’, to the further qualification in (c) ‘not necessarily’ and to the word ‘ordinarily’ in the ante-penultimate sentence in [22] of Hartley (above).

Nor are we persuaded that defences advanced at trial should necessarily determine the form of the indictment, although they plainly may, depending on the circumstances.

As to the periods covered by counts for this type of continuing offence, we do not think that it is either possible or desirable to set out any general statement of principle, beyond the observation that it is plainly sensible for consideration to be given as to the form of the indictment in good time before the trial begins so that the case can be properly managed in accordance with the principles that we have described above.

We will return later in this judgment to the sentencing issues, simply noting at this stage what was said by this Court in Dosanjh [2014] 2 Cr App R (S) 25 at [33] about the charge of Cheating the Public Revenue.

It is recognised as the appropriate charge for the small number of most serious revenue frauds, where the statutory offences will not adequately reflect the criminality involved and where a sentence at large is more appropriate to one subject to statutory restrictions. These are not ‘ordinary’ cases.

It follows from the above, that we have concluded that the Judge was right to reject the appellant’s initial application as to the form and particularity of count 1; and that he was also right to make clear that the prosecution should not be permitted to broaden the evidential scope of the count and that the whole issue would be kept under review. The Judge was bound to apply the overriding objective set out in CPR Part 1, and in this ruling he plainly did so.

The next procedural development occurred on the first day of the trial (17 September 2015), in relation to spreadsheets which were served by the prosecution. The Judge ruled that these could be placed in the Jury bundles as working documents that had not been agreed; and directed that the prosecution would have to prove the matters summarised in the schedules and, if it did not, the material would be removed. There is no current complaint about that ruling. In fraud cases, schedules summarising parts of the evidence are properly deployed so as to reduce the burden of the Jury in dealing with what would otherwise be a large number of documents. The trial was an EPE (electronic presentation of evidence) trial and the prosecution illustrated many of the underlying documents on screen. As the defence did not agree any of the entries on any of the schedules, it was necessary to show the jury that the entries were accurate and the Jury was shown the underlying evidence on screen. In fact, none of the figures were disputed.

Following this, the prosecution called a large number of witnesses (both clients of the firms and members of staff).

On 30 October 2014, the defence applied for further clarification of the charge in count 1 and, in particular, what was described as ‘the focus of the Crown’s case’.

Mr Cox pointed out that there were 400 potential transactions in the schedules and argued that this was an unmanageable number for the defence and the Jury to deal with. He applied for the prosecution to identify which transactions it relied on and a ruling that it be bound by that selection. His arguments were essentially directed at sensible case management and fairness.

The Judge indicated that, if the indictment had charged 400 or even 40 counts, he would have directed the prosecution to limit the charges in the interest of trial management. However, he declined to rule on the defence application because the prosecution had accepted that it would have to limit the number of transactions to a dozen or so, and give voluntary particulars in relation to these.

On 4 November, the defence made further submissions in relation to count 1. At this point the prosecution had confined its case to transactions relating to 15 particular individuals (11 clients and 4 members of staff). The defence complained that this number of transactions was itself unmanageable. Due to the long period covered by the indictment the jury would have to consider 99 transactions and even this reduced number was unmanageable.

The Judge rejected the complaint, ruling that the prosecution’s proposal to rely on any transactions associated with the 15 named individuals was a fair and reasonable balance that did justice to both sides. He declined to direct that the prosecution should further confine the number of transactions that the defence would have to deal with. The Jury would be directed that it was not necessary for the prosecution to prove that the appellant took part in every transaction and that it would be sufficient if the Jury was sure that he took part in three, provided it was sure of all the other elements of the offence and was agreed on the same three.

When it came to the summing-up the Judge directed the jury in line with this indication (at p.19B).

Now in order to assist you, as far as Count 1 is concerned, in your task, the prosecution has focused on a relatively small number of transactions involving 15 clients or staff members, as listed in the particulars annexed to Count 1.

You are entitled to look at the whole of the evidence; not just these in isolation, when you decide if the prosecution has proved the case. You can use the other transactions that involve others in order to reach your conclusions on whether the prosecution has proved knowledge, participation, dishonesty and all the other things that must be proved. What you cannot do is to convict him only on the other transactions that go outside the 15. You must be sure that the prosecution has proved its case on transactions within the list of 15 named clients or staff members, and because of the large number of transactions involved it is impossible to deal with each and every one. It would be unfair to convict the defendant by relying on any about which he was not specifically asked when he was in the witness box and this is the reason for this direction. If it is not necessary to prove that the defendant took part in all of the transactions relied upon by the prosecution, how many will amount to a system? It is sufficient that he took part in just three transactions involving any of the 15 individuals named. Of course, you must also be sure that in doing so he intended to cause loss to the Revenue, that he did cause loss to the Revenue and was acting dishonestly. But where it is that number of three you must all agree that it is the same three at least. It will not do if each of you thinks it was a different combination of three. In other words, it must be at least three and the same three. That is what we might call the bottom line. There must be a hard core of the same three which are common ground.

Mr Cox did not specifically criticise this direction, which was supplemented by a written summary for the assistance of the Jury. His complaint was directed to the prior decision to allow the prosecution to rely on so many transactions. He submitted that even reducing the transactions to 15 individuals and 99 transactions imposed too great a burden on the defence and the jury.

We do not accept this criticism. We have looked at the two-page spreadsheet and the appellant’s evidence in relation to the particular transactions; and we can see no real basis for saying that the appellant was unfairly disadvantaged by the number of transactions relied on. The defence knew from the first trial that the prosecution might be confined to reliance on transactions involving a limited number of clients and staff, and the appellant had sufficient time to prepare what he would say in respect of this evidence.

The prosecution case was relatively straightforward. In relation to each transaction there was a figure for the amount charged for accountancy fees in a particular accounting period and a higher figure representing the same work in the accounts or tax returns.

The figures set out in the schedules were not substantially in issue.

The defence case was of general effect. The appellant gave evidence that he had no interest in administrative matters, and was unaware that false invoices were being produced. He explained that he was away from the office much of the time and that, where a standard fee was charged, there would be a subsequent year adjustment to take into account the work actually done. In relation to those accounts which he dealt with personally, he told the Jury that there was nothing wrong or dishonest about the way in which the accountancy fees were dealt with, because there was no intent to cheat the Revenue, and in fact no loss had been caused to the Revenue. His evidence was that any over-claim of accountancy fees was always matched by and in many cases exceeded by what could have been claimed for use of the clients’ use of their home as their office.

In fact, the schedules showed that, contrary to the appellant’s assertion, there were no subsequent year adjustments of what was said to be a standard fee. We note that, if the counts had been confined to a single year, this point might have been more difficult for the Jury to identify.

The next ground of appeal relates to, first, the Judge’s direction on the appellant’s participation in, and knowledge of, the cheat; and secondly, to his direction on loss. Mr Cox submitted that the directions in both these respects were insufficient and inadequate.

As to the first point, the Judge gave written directions to the Jury, setting out the four elements of cheating the Revenue of which it had to be sure, namely: that the appellant, (1) was personally involved in the activity, either taking part himself or actively directing or encouraging others with the intention that the cheat be committed, (2) intended to cause loss to the Revenue, (3) caused loss to the Revenue by his actions, and (4) was acting dishonestly.

We should note that the prosecution accepted that the appellant had not personally prepared all the accounts and tax returns submitted to the Revenue. However, it adduced evidence from which it invited the Jury to conclude that, even in those cases where the appellant did not personally prepare accounts and tax returns, he knew of the approach adopted by the business, including the routine inclusion of overstated accountancy fee deductions.

In evidence the appellant explained how the routine overstating of accountancy fees had developed since about early 2006 and that he had adopted it himself. This was evidence from which the Jury could conclude that he not only knew of the practice but that he endorsed it. His defence necessarily involved him having knowledge of this practice because he had known that some staff were not adjusting the fees in subsequent years and that he had thought there was no cheat or loss to the Revenue because of the under-claim implicit in the home as office point.

The prosecution case went rather further, but it could at least point out that on the appellant’s own case the home as office point was only credible if he knew of the routine overstating of accountancy fees at the time.

Part of the Judge’s development of the direction (at p.18H) was in the following terms:

Now the prosecution’s case on Count 1 is that the defendant was involved in a system. The prosecution does not have to prove that he was involved in every single act, or even in the majority of them. What the prosecution does have to prove is that he was involved in a sufficient number to amount to a system, and when I say involved or take part I will tell you exactly what that means.

The Judge directed the Jury that in those transactions in which he was not personally involved, the prosecution had to show that he knew of what was going on, encouraged and approved of it, and took part in it, intending that the Revenue should be cheated.

He returned later to this aspect of the case (at p.86F).

You are, of course, concerned … with what the defendant himself knew. So whether the staff knew or didn’t know doesn’t really matter, because what you must be sure about is that the defendant knew about this, he intended to be part of it, to encourage it and to encourage the commission of the offence.

In our view, the summing up viewed overall, taken with written directions and his summary of the evidence of the 15 selected clients and staff in which he indicated where the appellant had been directly involved and where his staff had been involved (p.90-128), clearly conveyed to the Jury the importance of the prosecution having to make it sure of the appellant’s personal involvement in the cheat, with the requisite intent and dishonesty.

We would add that it was not necessary to go through each tax return for each tax year since this information was set out in the schedules and the defence was effectively the same for each tax year involved. Nor was it necessary to remind the Jury of each specific item in the schedules on which the appellant had been cross-examined. It was a matter for the Jury to decide in the light of what they had heard and the Judge’s summing up whether the prosecution had disproved the defence. Accordingly, we reject this ground of appeal.

So far as the second point is concerned, there appear to have been two elements to the defence argument. First, it was said that the fact that clients had subsequently settled enquiries into their tax affairs, and made additional payments of tax properly owed to the Revenue, was not evidence of loss. This matter was left to the Jury in terms which we consider generous to the appellant (at p.24D-25H, 60E-G and 87C-88B); with the Judge making it clear that it was for the prosecution to disprove the appellant’s assertion that there had been no loss. Secondly, there was reliance on the ‘home as office point’. This defence was also left to the Jury (at p.139B-134F), with the Judge directing the Jury that the issue was important because it affected three of the matters of which they had to be sure: whether loss was intended to be caused, whether loss was in fact caused and whether the appellant was acting dishonestly. In our view this direction does not give rise to proper grounds of complaint.

The appellant faced the forensic difficulty that, on the prosecution case the home as office point had only been raised some time after his arrest and his first interview under caution. The prosecution was able to make the point that it was a retrospective justification for his dishonesty.

Count 2 was drafted in terms which give rise to the same grounds of challenge to the conviction on counts 2 and 3, and it is convenient to focus on this count.

Statement of Offence

Cheating Her Majesty, the Queen and the Public Revenue, contrary to Common Law.

Particulars of Offence

[The appellant] between 1 January 2006 and 31 December 2011 in connection with Christopher Lunn & Company, Christopher Lunn & Company Ltd and Sno White Ltd cheated Her Majesty the Queen and the Public Revenue of income tax and corporation tax by dishonestly submitting or causing to be submitted tax returns and accounts:

(i)

for Sno White Ltd which were misleading in that they falsely brought forward losses of £37,764 to reduce taxable profits which were of a different trade;

(ii)

for Sno White Ltd which were misleading in that they falsely overstated the turnover by showing sales of £50,500 which had not taken place;

(iii)

for Christopher Lunn & Company which were misleading in that they falsely claimed as costs an amount of £50,500;

(iv)

for Christopher Lunn & Company Ltd which were misleading in that they falsely claimed as costs an amount of £50,500.

Mr Cox submitted first, that the count was duplicitous and uncertain, in that it was unclear whether the prosecution was relying on an accumulation of the enumerated particulars, or whether these were relied on as alternative ways in which it was said that the offence had been committed. His second point was that, when it came to the summing-up the Judge’s directions on law and fact were inadequate.

So far as the first point is concerned, the counts charged a fraudulent course of conduct: particulars (i) and (ii) dealing with the tax returns and accounts of Sno White Ltd, and particulars (iii) and (iv) dealing with the tax returns and accounts of the appellant’s firms. They were, as Ms Johnston put it, ‘two sides of the same coin’. On the prosecution case the appellant had (i) impermissibly applied the legitimate losses of Sno White Ltd to relieve future profits of a different trade, and (ii) reported false sales and turnover figures purporting to reflect sales and services to the appellant’s firms. Particulars (iii) and (iv) described the way in which the ‘fictitious’ sales were accounted for in the accounts and tax returns of the appellant’s firms. The figures for the expenses claimed were based on equivalent false figures in the accounts of Sno White Ltd; but the same costs had been claimed in the firm’s accounts for 2007 (iii) and the company’s accounts for 2008 (iv). The prosecution case was that this was all part of the same cheat.

The primary defence was that this was all part of a legitimate tax saving scheme which could have been properly argued if the Revenue taken issue with it.

In our judgment count 2 was sufficiently clear in its terms. Subject to Mr Cox’s second point, the issue for the Jury was whether they were sure that it was not a legitimate and honest scheme viewed overall.

We have already summarised the Judge’s general direction on the offence of Cheating the Revenue. In relation to the specific charge under count 2, Mr Cox submitted that the summing-up was perfunctory and confusing. One of the matters of specific complaint was what is said to be the Judge’s failure to direct the Jury, in relation to Sno White Ltd and the appellant’s firms, as to ‘whether there was continuity of trading’ and ‘whether it was the same business’.

We do not accept this criticism. The Jury had a large number of Agreed Facts (running to 34 pages). These included (at §4.1-5) the circumstances in which the trading losses of one trade can legitimately be carried forward and applied to relieve profits of the same trade, but not those of a different trade. Although there were a number of further issues relating (for example) to work which may or may not have been done by or through Sno White Ltd and the double counting issues addressed by particulars (iii) and (iv), the relevant issue so far as this ground is concerned, was whether the nature of Sno White’s trade under its previous ownership (about which the Jury had heard evidence) was the same as that said to have been conducted under the appellant’s stewardship. In our view there was nothing perfunctory or confusing about the Judge’s direction. His task was to explain the issue to the Jury in terms which they would understand, and he did so. He made clear that it was for the prosecution to disprove the appellant’s case that there was continuity of trade: the appellant having accepted in evidence that he knew it was wrong to carry forward and use losses in respect of different trades.

Accordingly, we reject the challenge to the conviction on counts 2 and 3.

Count 4 was a further charge of Cheating the Revenue which related to a graphic design business named Carte Blanche, later incorporated as Carte Blanche Design Ltd. The business was run by the appellant’s wife, and it was accepted that it provided some services to the appellant’s firms. The prosecution case was that the tax returns and accounts of Carte Blanche Design and the appellant’s firms were misleading in that they exaggerated and thereby falsely stated the costs of providing work in relation to the appellant’s firms’ website.

At the close of the prosecution case, the defence made submissions of no case to answer in relation to each of counts 2-4. On this appeal, Mr Cox submitted that the Judge should (at least) have acceded to the submission in relation to count 4. His argument was that the prosecution had accepted that Carte Blanche Design Ltd had done some work for the firms. The defence submitted that, in the absence of expert evidence as to the value of that work, there was no basis on which the Jury could properly conclude that the amount attributed to it in the accounts of the firm (some £129,840) was not a reasonable reflection of the work. If the case were to proceed it would have the effect of reversing the burden of proof and requiring the defence to justify the sums claimed as costs, which was contrary to principle.

The prosecution submitted that it was not a suitable issue for expert evidence, since there was insufficient underlying material (contracts, instructions, invoices, evidence of payments) and no indication of what had been done by Carte Blanche Design Ltd and when it had been done. The prosecution relied on inferences that could properly be drawn by the Jury from the lack of this material evidence, the positive evidence of two witnesses who had been involved in designing and building the website, other evidence of the limited services which had been provided and the very close involvement of the appellant in preparing the accounts of Carte Blanche Design Ltd. The prosecution case was that the value was very considerably less than the purported cost (£129,840) of the Carte Blanche Design Ltd services on ‘website design and computer consultancy’.

In our view the prosecution submission was well-founded, and the Judge was right to reject the submission of no case to answer. There was plainly evidence upon which a reasonable jury properly directed could conclude, on the basis of the prosecution evidence, that the appellant had cheated the Revenue as charged in count 4.

There is a further complaint that the summing-up was perfunctory, inadequate and misleading. We do not agree.

The Judge had already provided a written direction on the law and dealt with the facts in relation to count 4 on the second day of the summing-up. He described the factual issues as being the nature and value of the services provided by Carte Blanche and whether the appellant had acted dishonestly. He then reminded the Jury of the evidence in relation to these matters over the following 9 pages of the transcript.

Accordingly, we also reject the challenge to the conviction on count 4.

It follows that the appellant’s appeal against his conviction on counts 1-4 is dismissed.

The appellant was aged 70 at the date of sentence and was of previous good character.

The Judge found that throughout the relevant period he was the sole principal of the firm, the most highly qualified person within it and the sole driving force behind it. It was a relatively small business with less than 100 staff; and the Judge was satisfied from the evidence in the trial that there was little if anything that escaped his notice.

In terms of placing the offences within the Sentencing Council’s Definitive Guideline for Revenue Fraud, the defence acknowledged that the offences came within the high culpability category (category A). In the Judge’s view there were a number of aggravating features which increased the culpability within that category: the appellant had played a leading role, involving others in the operation of the cheat through pressure or influence; he had abused a position of responsibility; the offending had been carried out over a sustained period of time; and he had attempted to conceal evidence and blame others. The mitigating features were his age, the state of his health, and his previous good character. These features, however, were far outweighed by the aggravating features.

The Judge identified the real issue between the prosecution and the defence on count 1 as being the extent to which it was right to draw inferences about the extent and duration of the offending and the scale of the loss to the Revenue. He noted that the appellant was responsible for instituting and directing a scheme to overstate accountancy fees and, in doing so, he had boosted his firms’ reputation to his personal benefit. This should be reflected in the sentence.

The Judge accepted the prosecution’s contention that counts 2 and 3 fell together into harm category 6 (loss or intended loss of £20,000-£100,000). These two offences represented a blatant misuse of provisions in the tax legislation that enabled trading losses to be carried over if the businesses were the same. Count 4 came within harm category 7 (less than £20,000).

The Judge then addressed the argument that has been repeated on behalf of the appellant in relation to count 1 on this appeal, namely: that since the Jury had been directed that it need only be satisfied that the appellant took part in three transactions, the Judge was bound to pass a sentence based on the average loss to the Revenue from three transactions. The defence calculated that the average loss on each transaction was £237.87, and on 3 transactions was £713.62. On this basis the offence fell within harm category 7. The starting point for category 7A was 18 months’ imprisonment (based on a loss of £12,500) and a range of 36 weeks to 3 years. In view of the small amount of the loss, the defence submitted that the sentence should be measured in terms of no more than a few weeks.

The Judge noted that the prosecution had argued that the loss ran into millions of pounds on the basis of the figures set out in the initial schedules, but considered that this approach would be unfair to the defendant. His task was to interpret the Jury’s verdict. As a matter of principle, sentences should be passed on the basis that was most favourable to a defendant if the evidence allowed such a possibility, but judges were not obliged to do so if satisfied beyond reasonable doubt about the extent of the offending.

On this basis the Judge concluded that the offence charge in count 1 should be sentenced on the basis of harm category 5: loss to the Revenue of £100,000-£500,000, with a starting point based on £300,000. The starting point for category 5A was a term of 5 years custody, with a range of 3 to 6 years. As we have noted above, the Judge sentenced the appellant to a term of 4 years on count 1; and to 12 months on each of counts 2, 3 and 4, concurrent to each other but consecutive to the sentence on count 1.

In the grounds of appeal, a number of points were directed to the sentence on count 1, which Mr Cox developed in his oral submissions.

First, he submitted that the starting point on count 1 was wrong in principle. The Judge had failed to adopt the appropriate basis for sentence as set in R v Canavan, Kidd & Shaw (above). Due to the way count 1 was framed and the nature of the Judge’s direction to the Jury, the verdict had to be interpreted as a finding that the appellant was guilty of 3 of the many transactions relied on by the prosecution. Beyond this, it was impossible to determine the extent of the offending for the purpose of sentence. The Judge should therefore have sentenced on the basis of 3 transactions, and passed a sentence at the lowest end of the range in category 7A.

Secondly, he argued that the Judge had erred in principle by adopting as the basis for his sentence, a schedule entitled ‘Prosecution Response to Approach on Sentencing’, produced on the day before the sentencing hearing, which had been advanced without either proper notice or any rational foundation. As we have already noted, the previous approach of the prosecution had been that the losses ran into millions of pounds. The new document advanced an alternative approach based on a loss of £186,000.

Thirdly, Mr Cox submitted that the Judge had misled himself as to the position of the appellant and the suitability of category 5 of the Guidelines.

It has been clear at least since Canavan, Kidd & Shaw (above) that a defendant cannot be sentenced for conduct which has not formed, expressly or by necessary implication, the subject of charges laid and proved against him.

Where a defendant is convicted on an indictment charging him with offences said to be representative of other similar criminal offences committed by him, it is inconsistent with principle that the court should take into account such other offences so as to increase the sentence if the defendant does not admit the commission of other offences and does not ask the court to take them into consideration. Nor does [Statute] legitimate the practice for unindicted, unadmitted offences.

In the case of A [2015] 2 Cr App R (S) p.115 the prosecution had charged a number of ‘multiple incident’ counts of rape and assault by penetration. The jury were directed that they had to be sure that the incidents complained of had happened more than once. In the Court of Appeal, Fulford LJ set out three distinct possible approaches, of which two are relevant here: (1) to include a count or counts in the indictment pursuant CPR r.14.2(2) where the incidents taken together form a course of conduct having regard to time, place or purpose of commission, or (2) to include sufficient counts to enable the judge to impose a sentence which sufficiently represents what happened, but without overloading the indictment. The difficulty in the case of A was that there was no means of interpreting the jury’s verdict. At [47] the Court said this:

In our judgment, the central answer to this problem is to be identified in the purpose underpinning multiple counts: it is to enable the prosecution to reflect the defendant’s alleged criminality when the offences are so similar and numerous that it is inappropriate to indict each occasions, or a large number of different occasions in separate charges. This provision allows the prosecution to reflect the offending in these circumstances to a single count rather than a number of specimen counts.

For the reasons we have already given, we consider that it was appropriate in the circumstances of this case to charge the offence of cheating as a single count alleging a course of criminal conduct. It was neither a sample count nor a specimen count. The Jury’s verdict was that the appellant was guilty of Cheating the Revenue, not that he had cheated the Revenue in relation to 3 transactions. Once that verdict had been returned, it was for the Judge to sentence on a conventional assessment of the culpability and harm of the offending; and, in doing so, he was not confined to sentence on the basis of 3 transactions. In our view the Judge’s approach to sentencing, including his approach to making assumptions in favour of the defendant, was plainly correct, see for example, Efionayi (1995) 16 Cr App R (S) 380.

Although we have considered Mr Cox’s second and third complaint about the sentence on count 1, we are not persuaded that there was any error of approach by the Judge nor any underlying unfairness in the way the prosecution characterised the offence. The prosecution had submitted that the harm was category 3 (loss to the Revenue of £2 to £10 million, with a starting point of £5 million). It subsequently advanced an alternative calculation based on basic rate tax of 20% applied to the amount by which the accountancy fees were overstated. This figure for tax was based solely on the schedules produced during the evidence which had been called or read. If the defence had been placed in real difficulty by the timing of the prosecution alternative case, and we are not persuaded that they were, they should have asked for an adjournment to deal with the matter.

Since we reject the appellant’s criticism of the sentence on count 1, it follows that the appeal against sentence must also be dismissed.