[2016] UKFTT 831 (TC)

[image removed]

TC05556

 

Appeal number: TC/2014/04182         

 

INCOME TAX – discovery assessment – claim to relief ‘in’ self-assessment meant self-assessment insufficient– whether claim made carelessly in view of advice given – yes – whether HMRC newly discovered insufficiency after Supreme  Court decision in Cotter – yes-  appeal dismissed

 

 

FIRST-TIER TRIBUNAL

TAX CHAMBER

 

 

 

	RICHARD ATHERTON	Appellant
	- and -
	THE COMMISSIONERS FOR HER MAJESTY’S	Respondents
	REVENUE & CUSTOMS

 

 

 

TRIBUNAL:	JUDGE Barbara Mosedale
	G Noel Barrett LL.B

 

 

 

Sitting in public at the Royal Courts of Justice, Strand, London on 5-6 May, 6 July, 26 and 29 September 2016

 

 

Mr Keith Gordon, Counsel,  for the Appellant

 

Ms K Balmer, Counsel, instructed by the General Counsel and Solicitor to HM Revenue and Customs, for the Respondents

 

© CROWN COPYRIGHT 2016

DECISION

 

 

The scheme

1.       Sometime between 12 and 30 January 2009, the appellant entered into a tax planning scheme called Romangate marketed by NT Advisers LLP (‘NTA’).  The workings of the scheme were not explained to us but it was agreed it resulted in a manufactured employment loss arising in tax year 2008/09.

2.       The appellant submitted his tax return for tax year 2007/8 on 30 January 2009 and in that return, he claimed relief for the employment losses generated in January 2009 (within tax year 2008/9) against his income (some £5 million) for 2007/08, reducing his tax liability from just over £2million to nil for that tax year. 

3.       The Tribunal was not called on to decide whether or not the scheme would have been effective to generate a tax loss as the law stood at the time the appellant entered into the arrangements.  This was because s 68(1) Finance Act 2009 inserted new section s 128(5A) into the Income Tax Act 2007 and made the scheme ineffective because it removed relief for employment losses generated from arrangements whose objective was tax avoidance:   the legislation was retrospective in effect, capturing anything done on or after 12 January 2009, and therefore capturing the appellant’s arrangements.

4.       It is now accepted by the appellant that, due to this retrospective legislation, he was not entitled to the claimed relief in any tax year.  His position is nevertheless that the assessment raised by HMRC to recover the unpaid tax was procedurally wrong and therefore ineffective.

What did the tax return show?

5.       The case largely turns on how Mr Atherton (or his advisers) completed his tax return.  There were three entries relevant to the appeal, as follows:

(1)   Box 3:  The standard tax return pages for 2007/08 submitted by the appellant included ‘Additional Information’ pages.  The Additional information pages entitled the taxpayer to made a claim for relief in Box 3 under the hearing ‘income tax losses’:

Relief now for 2008-9 trading, or certain capital, losses.

The appellant entered his manufactured employment loss of £6,149,999 which arose in 2008/09 in this box. 

(2)   White space disclosure:  the appellant made a fairly lengthy white space disclosure in his tax return.  We found that the white space disclosure related only to the entry in Box 3.  The disclosure referred to a 2008/09 employment related loss being claimed ‘using box 3’ and explained that box 3 was being utilised, despite its title about trading or capital losses, as there was no other equivalent box for employment losses.  The disclosure went on to mention details about the scheme, including its DOTAS number. It said:

I acknowledge that my interpretation of the tax law applicable to the above transactions and the loss (and the manner in which I have reported them) may be at variance with that of HMRC. 

It repeated at the end that, although the claim was made in box 3, it might not be seen by HMRC as appropriate to that box and concluded, for these reasons,

I assume you will open an enquiry.

(3)   Box 20:  the assessment the subject of this hearing was made because the appellant also entered the figure of £5,048,602 into Box 20 of the Partnership pages of his tax return for 2007/8.  It was accepted that this figure was the employment income loss (capped at the amount of Mr Atherton’s actual income). He had no partnership loss:  the loss claimed had nothing to do with any partnership.

However, box 20 was under the main heading:

Your share of the partnership’s trading or professional losses

Box 19 under the same heading asked the taxpayer to enter the loss for 2007-08 from the working sheet.  Box 20 itself was headed:

Loss from this tax year set-off against other income for 2007-08

6.       The effect of entering the loss in Box 20 of the tax return was that it was carried into Mr Atherton’s self-assessment calculation for 2007/08:  it meant that his tax return self assessed him to nil (actually, to a small repayment of tax).  Without the entry in Box 20, his self-assessment would have been in the amount of the discovery assessment, some £2,010,855.20.

Standalone claim v a claim made in a tax return

7.       It is well understood and not in dispute that taxpayers are entitled to make claims to have certain losses from the subsequent year (‘Year 2’) set against income arising in the previous year (‘Year 1’), leading to a diminution in Year 1 tax liability.  It was also well understood and not in dispute that where such claims were made entirely outside the tax return form, normally by letter, they were ‘standalone’ claims made under Schedule 1A Taxes Management Act (‘TMA’).  HMRC does not have to give immediate effect to a standalone claim (by reducing or repaying Year 1 tax) if HMRC open an enquiry into the claim (paragraphs 4(1), 4(3) and 5 of Sch 1A).  And such a claim can only be challenged by an enquiry opened under the provisions of Sch 1A TMA (which we will refer to as a ‘Sch 1A enquiry’).

8.       On the other hand, where a taxpayer claims a deduction in his tax return, as income tax is a self-assessed tax, such a claim is given immediate effect.  HMRC can raise an enquiry into the tax return under s 9A TMA but they cannot refuse to give effect to the self-assessment pending closure of the enquiry and an amendment to the self-assessment.

9.       HMRC’s view, we find from their manuals, has historically been that any claim to carry a Year 2 loss against Year 1 profits is a standalone claim whether or not made separately by letter or on the face of the tax return.  They considered it could only be challenged by a Sch 1A enquiry.

10.    A Mr Cotter, who, like the appellant entered into the Romangate scheme, but, unlike the appellant, only made entries concerning the manufactured employment loss in Box 3 and in the white space of his return, challenged HMRC’s view.  He considered that his entry in box 3 was making a claim ‘in’ his return, reducing his tax liability to nil, and that it could only be challenged by a s 9A TMA enquiry.  HMRC had only opened a Sch 1A enquiry and pursued him in the County Court for the amount of tax shown as owing in his tax calculation.  HMRC were out of time to open a s 9A enquiry so, if Mr Cotter was right that the Sch 1A enquiry was ineffective, he could not be made to pay the tax that was owing.

11.    He was not successful.  The Supreme Court ruled that an entry ‘on’ the face of the tax return was not necessarily ‘in’ the tax return.  The parties were agreed that the effect of the Supreme Court’s decision in Cotter [2013] STC 2480 was that those parts of the tax return form which invite taxpayers to claim Year 2 losses against Year 1 tax are outside the tax return proper and amount to no more than standalone claims.  They must be challenged, if at all, by a Sch 1A enquiry:

[25]  ...The word "return" may have a wider meaning in other contexts within TMA. But, in my view, in the context of sections 8(1), 9, 9A and 42(11)(a) of the TMA, a "return" refers to the information in the tax return form which is submitted for "the purpose of establishing the amounts in which a person is chargeable to income tax and capital gains tax" for the relevant year of assessment and "the amount payable by him by way of income tax for that year" (section 8(1) TMA).

[26] In this case, the figures in box 14 on page CG1 and in box 3 on page Ai3 were supplemented by the explanations which Mr Cotter gave of his claim in the boxes requesting "any other information" and "additional information" in the tax return. Those explanations alerted the Revenue to the nature of the claim for relief. It concluded, correctly, that the claim under section 128 of ITA in respect of losses incurred in 2008/09 did not alter the tax chargeable or payable in relation to 2007/08. The Revenue was accordingly entitled and indeed obliged to use Schedule 1A of TMA as the vehicle for its enquiry into the claim (section 42(11)(a)).

12.    This point certainly had not been clear before the decision of the Supreme Court.  On the contrary, the Court of Appeal’s decision in Cotter [2012] STC 745 was  that anything reasonably on the face of the tax return was ‘in’ the tax return and could only be challenged by a s 9A enquiry (or a correction of the return).  So, the Court of Appeal’s view in 2012 was that an entry in Box 3 was ‘in’ the return and not a standalone claim and had to be challenged by a s 9A enquiry.  Had HMRC not appealed that decision, Mr Cotter would have succeeded in avoiding the tax.

Forcing a claim into a tax return

13.    As we have already explained, Mr Atherton did not complete his tax return identically to Mr Cotter.  In addition to the Box 3 entry and white space disclosure which Mr Cotter made, Mr Atherton also entered the Year 2 loss into Box 20.  And the effect of that, as we have said, was that the loss was carried into his Year 1 self-assessment calculation resulting in a reduction in tax liability for 2007/08 (year 1) to nil (actually, a small repayment).

14.    HMRC referred to what the appellant did as ‘forcing’ a claim into his tax return. The appellant considered the expression pejorative, but we adopt it as (a) it is a useful way of distinguishing between a standalone claim made ‘on’ but not ‘in’ a tax return and a claim actually made ‘in’ a tax return such that it affects the self-assessment, and (b) in any event, it is clear following the Supreme Court decision in Cotter  at [16], and not in dispute, that it was wrong to make a Year 2 loss claim ‘in’ a tax return.  Such claims should only be made as standalone claims.

15.    Unlike Mr Atherton, Mr Cotter had not forced his Year 2 loss into his Year 1 tax return.  Nevertheless, the Supreme Court in Cotter did go on to consider (technically obiter) the position of a taxpayer who had ‘forced’ the loss claim into his year 1 self-assessment calculation, thus reducing the amount of tax shown as owing for Year 1.  Would a Sch 1A enquiry be effective to challenge such a claim?  This was highly relevant to Mr Atherton, as he had completed Box 20 and thereby put the claim into his self-assessment calculation.

16.    It was HMRC’s position before the Supreme Court, consistent with their view summarised at §9 above, that doing so would be ineffective:  a standalone claim was a standalone claim however made.  The Supreme Court did not agree.  They said:

[27] Matters would have been different if the taxpayer had calculated his liability to income and capital gains tax by requesting and completing the tax calculation summary pages of the tax return. In such circumstances the Revenue would have his assessment that, as a result of the claim, specific sums or no sums were due as the tax chargeable and payable for 2007/08. Such information and self- assessment would in my view fall within a "return" under section 9A of TMA as it would be the taxpayer's assessment of his liability in respect of the relevant tax year. The Revenue could not go behind the taxpayer's self-assessment without either amending the tax return (section 9ZB of TMA) or instituting an enquiry under section 9A of TMA.

17.    In other words, if a taxpayer did as Mr Atherton had done, and put his loss claim into a box which forced the claim into his self-assessment calculation, reducing his declared liability to tax for year 1, HMRC had to open a s 9A enquiry into the tax return:  it was not a standalone claim so opening a Sch 1A enquiry into a standalone claim would be ineffective.  While it was wrong to force such a claim into the tax return, the only effective way to challenge it was a s 9A enquiry into the tax return itself.

18.    Both parties accept the decision of the Supreme Court.  So, whatever was understood by the parties in 2009, it was agreed by the time of the hearing that the appellant had made a claim in his tax return for 2007/8 for an employment loss incurred in 2008/9.  It was agreed that because the claim was made in the tax return, the Sch 1A enquiry which HMRC opened in May 2009 was ineffective to challenge it.  If HMRC had wanted to enquire into the Box 20 entry, they should have opened an in-time s 9A TMA enquiry into the tax return.  They had not done so.

19.    We consider that the most accurate explanation of what the appellant had done was that he had made two claims:  he had made a standalone claim for the 2008/9 loss on his 07/08 return, but not ‘in’ the return when he completed Box 3; at the same time he made a claim for exactly the same loss ‘in’ his 07/08 return when he completed Box 20.  The Sch 1A enquiry was effective to challenge the Box 3 entry.  It was ineffective to challenge the Box 20 entry.

20.    It is also worth mentioning how the taxpayer in the case of Rouse [2013] UKUT 383 (TCC) completed his return, as he completed it in a slightly different way to either Mr Cotter or the appellant in this case.  Unlike the appellant, Mr Rouse did not put his year 2 loss into any box, such as Box 20, that carried the loss into his Year 1 self assessment calculation.  He only made the Box 3 entry.  However, having carried out his self-assessment in accordance with the law, he then deducted the loss claim from his calculation of the sums payable as a result of the self-assessment.  The Upper Tribunal concluded that doing so fell on the Cotter side of the line: the claim for the year 2 loss to be carried back to year 1 was made as a standalone claim and had not been ‘forced’ into the Year 1 tax return.  The Sch 1A enquiry was effective to defeat the claim.

21.    It was the appellant’s position that at the time his 2007/08 return was submitted, he neither knew that the arrangements he entered into were ineffective to generate an employment loss for tax purposes nor that, even if they had been effective, that he was not entitled to claim this loss in his 2007/08 tax return. At the point that he submitted his return, the retrospective legislation had not been promulgated nor could he foresee the decision of the Supreme Court in Cotter, some four years later.

The assessment

22.    Nevertheless, on 31 March 2014, the appellant was issued with a discovery assessment in relation to tax year 2007/8 assessing him to the £2,010,855.20 tax which it is accepted was underpaid.  The appellant’s position is that the unpaid tax is only recoverable from him by assessment and, says the appellant, the 31 March 2014 discovery assessment is ineffective as, he says, HMRC did not meet the statutory conditions to make it.

23.    The dispute between the parties therefore entirely turns upon the validity of the 31 March 2014 assessment (‘the Assessment’).  If the Assessment was validly made, the appellant accepts the scheme was ineffective and his liability to pay the Assessment.  If the Assessment was not validly made, the appellant is not liable to pay the tax which the 2007/08 self-assessment should have shown as owing to HMRC.

The extended hearing

24.    The two days for which the case was listed proved to be inadequate.  We do not seek in this decision to apportion any blame for this:  there may or may not be a subsequent dispute over costs.  It seems to us the main reasons why a two day case took five days was that there were a number of interlocutory applications made which took up time, and while two days may well have been adequate to hear the three witnesses originally intended to give evidence, after the first two days, the appellant was given leave to adduce the evidence of a fourth witness (Mr Jenner) whose oral evidence was extensive.

What HMRC must prove

25.    As we have said, the only matter in dispute between the parties was the procedural validity of the discovery assessment.  The appellant accepted that if the discovery assessment was procedurally valid, he was liable to pay it.  The validity of the assessment depended on whether it met the preconditions of s 29 Taxes Management Act 1970 (‘TMA’) which provided as follows:

29 assessment where loss of tax discovered

(1) If an officer of the Board or the board discover, as regards any person (the taxpayer) and a year of assessment

(a) That any income which ought to have been assessed to income tax...have not been assessed, or

(b) That an assessment to tax is or has become insufficient, or

(c) that any relief which has been given is or has become excessive,

the officer or, as the case may be, the Board may, subject to subsections (2) and (3) below, make an assessment in the amount, or the further amount, which ought in his or their opinion to be charged in order to make good to the Crown the loss of tax.

Subsection (2), which dealt with assessments made in accordance with generally prevailing practice was agreed rightly not to be relevant; subsection (3) provided that one of two conditions had to be met.  The first condition was contained in subsection (4) and the second, relating to awareness of the insufficiency, was contained in (5).  HMRC did not advance a case that the second condition was met, so the Tribunal was only concerned with HMRC’s case that the first condition was met.  That was:

(4) the first condition is that the situation mentioned in subsection (1) was brought about carelessly or deliberately by the taxpayer or a person acting on his behalf.

26.    It was accepted by HMRC, rightly, that they have the burden of proving to the Tribunal that they made a discovery within the terms of s 29 TMA.  They had to prove that there was a ‘discovery’ within the meaning of s 29(1) and carelessness/deliberate behaviour within the meaning of s 29(4).  As, therefore, HMRC had the burden of proof on the only two issues in dispute between the parties (discovery and careless/deliberate behaviour), it was agreed that HMRC would open the hearing.

27.    Before HMRC did so, the appellant made an application in respect of HMRC’s case on the condition contained in s 29(4).  Ms Balmer’s skeleton argument alleged that Mr Atherton, or someone acting on his behalf, completed his tax return incorrectly deliberately or through carelessness.  Mr Gordon’s position at the start of the hearing was that HMRC’s statement of case had done no more than allege carelessness by Mr Atherton himself, and HMRC’s case at the hearing should be restricted to that allegation. 

Were HMRC entitled to make a case that the appellant acted deliberately?

28.    We announced our decision on this application at the time.  We gave oral reasons then, but agreed to record them in the written notice of our overall determination, in order that our decision on this interim matter could be appealed, should either party wish to do so.

29.    We decided that HMRC would not be allowed to allege that Mr Atherton had acted deliberately.  The reasons we gave were:

(1)    an allegation of ‘deliberate’ behaviour was much more serious than an allegation of ‘careless’ behaviour; whether or not it was an allegation of dishonesty, it was certainly an allegation that the appellant had intentionally done something wrong.

(2)   Pleadings have to be clear so that a party knows what is alleged; pleadings of serious misconduct, such as dishonesty or intentionally doing something wrong, have to be specifically pleaded, so a party is in no doubt what is alleged against him.

(3)   Yet in our opinion, HMRC’s statement of case did not make it clear that ‘deliberate’ behaviour was alleged against Mr Atherton.  We accept that the statement of case did set out s 29(4) of the legislation in full and that that subsection refers to both careless and deliberate behaviour (for the text see §25), but that meant nothing without a clear statement of what HMRC alleged the appellant had done.  Paragraph 26 of the statement read to us as if it was saying that, while HMRC considered the appellant to have behaved deliberately, HMRC were choosing not to pursue the allegation because to succeed in their case they only had to prove carelessness.  Elsewhere the statement of case referred to the appellant being ‘at least careless’ which we do not consider to be a clear pleading of deliberate behaviour but, on the contrary, consistent with the implication from paragraph 26 that HMRC only intended to prove carelessness.  That HMRC in their statement of case were only relying on carelessness was also demonstrated because the author attempted to give a definition of ‘carelessly’ but not of ‘deliberately’.

Was carelessness by appellant’s advisers a part of HMRC’s case?

30.    We came to the opposite conclusion on Mr Gordon’s application in so far as it related to HMRC’s case that the appellant’s advisers had acted carelessly.  The reasons we gave were as follows:

(1)   Specific pleading of carelessness was not required in the same way it is for dishonesty.  We found in any event that carelessness of the appellant’s advisers was pleaded: paragraph 28 and 30 of the statement of case in particular refer to carelessness by the appellant or someone on his behalf. 

(2)   Moreover, it is not for HMRC to anticipate in the statement of case a defence to be raised to the statement of case.  HMRC alleged the appellant was careless; his defence was that if there was carelessness in the completion of his tax return, he was not responsible for it because he was acting on the advice of others.  It was perfectly proper for HMRC to then put the case that the (alleged) carelessness of those others was sufficient for s 29(4) purposes.   That was particularly the case where HMRC’s statement of case had in fact contemplated the possibility that the alleged carelessness was due to someone acting on behalf of the appellant.

(3)   And in so far as Mr Gordon suggested HMRC had to identify in the statement of case the person alleged to be careless, that would operate to require HMRC to foresee a defence to their statement of case, which they were not required to do.

31.    We ruled that carelessness by someone acting on the appellant’s behalf was pleaded and HMRC could make that allegation in the hearing; as it was part of the appellant’s case that he relied on the advice of other people, it was for him to say on whom he relied, and, if he chose, to call them as witnesses.

32.    At this point, HMRC offered the appellant an adjournment of the hearing without costs.  The appellant did not accept and elected to continue with the hearing. 

Submission of no case to answer

Burden of proof

33.    Ms Balmer finished opening her case and calling her witnesses in the morning of the second day of the hearing.  Mr Gordon then made an application without warning for the appeal to be allowed on the basis that HMRC had failed to make out a case of carelessness by the appellant. The application and its resolution took up the rest of the day.    In the afternoon of the second day, the Judge read out the decision of the panel dismissing the appellant’s application.  Directions were then issued for the postponed hearing of the rest of the case.

34.    The appellant indicated later, but before the next hearing date, that he wished the Judge to record in writing the reasons for its 5 and 6 May interim decisions as he wanted to appeal them, and, further, he wanted the adjourned hearing of the case postponed until after the Upper Tribunal determination of the appeal against the interim decisions.  The Judge in chambers refused the postponement application:  reasons for that decision were communicated to the parties by letter and it is understood that that decision is not under appeal, so we do not repeat them here.  As postponement was refused, the Judge indicated that the written reasons for the interim decisions given on 5 and 6 May would be contained in the written decision to be issued determining the case.  The written reasons for refusing the submission of no case to answer are therefore set out below.

What is a prima facie case?

35.    There appeared to be no dispute between the parties that, as HMRC had the burden of proving they had made a valid discovery assessment, HMRC had to prove the constituents of such an assessment, in other words that there was a discovery and the appellant or someone on his behalf carelessly completed the tax return containing the insufficiency the subject of the discovery assessment.  Again, both parties appeared to agree, as we do, that the appellant should not be required to reply to HMRC’s opening case unless HMRC had satisfied the Tribunal that it was more likely than not that there was a discovery within the meaning of s 29(1) and carelessness within the meaning of 29(4).  In other words, HMRC had to do enough in their opening, in the absence of any reply from the appellant, to satisfy the Tribunal on the balance of probabilities of their case. This is referred to by the Latin tag of a ‘prima facie’ case which we will use as a convenient short hand.

36.    Mr Gordon did not suggest that HMRC had not raised a prima facie case with respect to discovery.  Mr Gordon’s submission of no case to answer was made solely on the basis that there was no evidence to support a case of carelessness by the appellant or his advisers.  HMRC’s two witnesses, Mr Clarke and Mr Taylor, had given evidence (which we set out below at §§72-77) which only related to the issue of discovery.  Moreover, Mr Gordon said, bearing in mind that the appellant had properly implemented the tax avoidance scheme, and it had only been defeated by retrospective legislation, how could it possibly be said that the tax return was completed carelessly?

No documentary evidence?

37.    We agree with Mr Gordon that we had no oral evidence of any carelessness by the appellant.  But we were surprised by his submission that the Tribunal had no evidence at all to consider, as the tax return was in the bundle of documents.  HMRC had submitted a list of documents including the appellant’s tax return, and although neither of HMRC’s witnesses had specifically stated that the tax return in the list of documents was the appellant’s tax return in issue, the Tribunal had not understood this point to be in dispute.

38.    So was Mr Gordon right to say that we had no evidence of carelessness to consider and must therefore allow the appeal?

39.    Mr Gordon referred us to the decision in the case of Gardiner [2014] UKFTT 421 (TC) where a preliminary issue of whether HMRC had adduced any evidence to support a prima facie case of negligently delivering an incorrect tax return was decided in favour of the appellant, and the appeal succeeded.  HMRC’s case in Gardiner was that the scheme documentation was obviously defective and the taxpayer should have known this:  the judge threw out the case on the basis that the scheme documentation was not in evidence [20-33].  We found this an odd decision as it appeared to us, reading the decision, that the scheme documentation was on HMRC’s list of documents, its authenticity did not appear to be in dispute, and, moreover, comprised documents put forward by the taxpayer to support his tax return.  The decision seemed to suggest that if an HMRC officer had been called by HMRC simply to give the evidence that the scheme documentation in the list of documents was the scheme documentation produced by the taxpayer to support his tax return, then the Judge would have accepted that there was evidence for him to consider.  As it did not appear that there was any dispute about the authenticity of the scheme documentation, it seemed an unnecessary formality to require oral evidence of its authenticity.   

40.    We do not agree with the decision in Gardiner if it requires, in order for a document to be in evidence before the Tribunal, a witness to speak to the authenticity of a document when its authenticity is not in dispute.  That seems to us to run counter to the overriding principle (Rule 2(2)(b)) of avoiding unnecessary formality and would be likely to require much unnecessary oral evidence.  The decision is not binding on us and we would not be inclined to follow it.

41.      In this case, HMRC did not rely on the scheme documentation to support their case of carelessness:  they relied on the manner in which the tax return was completed and in particular that the appellant had claimed a loss in Box 20 of his tax return.

42.    Was the tax return in evidence before us?  We considered that it was.  It was listed on HMRC’s list of documents and the appellant had never put in doubt its authenticity.  Indeed, Mr Gordon confirmed to us that its authenticity was not in doubt and withdraw his suggestion that it was not in evidence before us. 

43.    We proceeded to consider whether the tax return was sufficient evidence to raise a prima facie case of carelessness by the taxpayer or someone acting on his behalf.

Wrong figure in box

44.    HMRC did not suggest that the scheme implemented by Mr Atherton was ineffective at the time the tax return was completed and filed; what they did say is that the loss should not have been claimed in Box 20 of the tax return. 

45.    What we knew at the end of HMRC’s closing is that HMRC alleged, and the appellant (from the skeleton) appeared to accept, that the scheme he implemented created an employment income loss in tax year 2008/09.  Nevertheless, the appellant had claimed relief for the loss in Box 20 of his 2007/08 return.  Box 20 was in the partnership  pages of the tax return, under the heading of ‘partnership trading or professional losses’ and which related to ‘this’ tax year.  The claim in box 20 directly led to the insufficiency in the self assessment.

46.    On its face, we found that HMRC had established a prima facie case that the loss should not have been entered in Box 20.  It was not the type of loss which should have been entered into Box 20, as it was not a partnership loss and did not arise in the tax year of 2007/08. 

Are errors in tax returns prima facie evidence of carelessness?

47.    Mr Gordon did not think the fact that a tax return was incorrectly completed was prima facie evidence of carelessness in any case and certainly not in this case.  He thought this so obvious that it was not a point that had even been considered in the case of Gardiner.  Mr Gordon’s view was that if HMRC thought that a taxpayer may have been careless, they would have to use their information powers in order to ask the taxpayer questions about how he completed his tax return in order to establish whether he had acted carelessly.  Because HMRC had not done so in this case, it was Mr Gordon’s view that they could not establish a prima facie case of carelessness.

48.    We did not accept this.  We do not think that Parliament intended the bar for a prima facie case of carelessness to be set high.  It is the nature of tax that the taxpayers hold the evidence of their liability to tax.  We doubt Parliament intended that HMRC could only prove carelessness by relying on the taxpayer’s answers to information notice requests: moreover, it is contrary to justice to require the defendant to give evidence or answer questions before a case of negligence could be made out. Actions can speak for themselves.

49.    Mr Gordon went on to rely on statutory construction to say that a wrong entry on a tax return was not by itself prima facie proof of carelessness because, if it was, why have s 29(4) at all?  This point, however, was also not a good one.  A prima facie case can be rebutted:  so even if a wrong tax return was prima facie proof of carelessness in every case, it would not render s 29(4) otiose as a taxpayer has the right of rebuttal and might be able to demonstrate that a particular error was not carelessly made.

50.    Mr Gordon also said that a wrong entry on a tax return was not necessarily careless because there might be a good reason for it:  in this case, he suggested that the reason for completion of Box 20 may have been that the taxpayer had been advised he had the right to claim the loss in the tax return for the previous year and there was no appropriate box on the return in which to claim the relief, so he had had to utilise Box 20 for lack of anything else.  We considered Mr Gordon was here confusing the question of rebuttal with whether or not making a wrong entry was prima facie careless.  It is the nature of prima facie cases that it may be possible to rebut them: that there may well be an explanation does not prevent HMRC making out a prima facie case.

51.    Our decision was that an error on the face of the tax return was prima facie proof of carelessness.  Without an explanation, a wrong entry on a tax return seems more likely than not to be caused by carelessness. A taxpayer has a duty to render complete and correct tax returns, and when he fails to do so that is prima facie proof of a breach of that duty and carelessness at the very least.  Where a tax return is incorrectly completed, it seems to us that if the appellant did not defend the allegation of carelessness, HMRC would win the case. Moreover, as it is the taxpayer who knows how and why he completed his tax return as he did, it makes sense that the law requires the taxpayer to explain why he completed his tax return inaccurately, and defend the allegation that the inaccuracy was careless.

Was the error in this case prima facie evidence of carelessness?

52.    Even if we are wrong to consider that as a general rule an error on a tax return is prima facie evidence of carelessness, we find that on the facts of this particular case, the claim to a 2007/08 partnership loss was carelessly made as it was accepted that the appellant didn’t have a 2007/08 partnership loss.  More specifically, it was not in dispute that the entry in box 20 referred to the 2008/09 manufactured employment loss.  So a loss arising from employment in 2008/09 was entered into a box expressly limited to partnership losses arising in 2007/08.  We were satisfied that HMRC had established a prima facie case that return was at the very least carelessly completed.

53.    It was suggested to us that the tax return was not prima facie evidence of carelessness in that any errors in it were disclosed in the ‘white space’, the area of the tax return where the taxpayer can make explanations.  We refer to this in more detail below at §§142-146.  Suffice it to say, that in the absence of any explanation from the appellant, we found that on its face the white space disclosure related only to the Box 3 entry and did not explain that the appellant was making a duplicate claim for the same loss in Box 20, let alone the appellant’s reasons for so doing. 

54.    We were of the opinion in any event that white space disclosure explaining why a wrong entry was made would not necessarily prevent a wrong entry being careless, but in this case the white space disclosure did not even attempt to identify or explain the wrong entry in box 20.  It did not, therefore, detract from the prima facie case of carelessness HMRC had established by pointing out that a 2008/09 employment loss was entered into the 2007/08 return under the heading partnership losses arising in that tax year (ie 2007/08).

55.    We found that HMRC had established against the appellant a prima facie case of carelessness within the meaning of s 29(4).  We therefore refused to allow the appeal.  As it was near the end of Day 2 of the hearing, the hearing was adjourned to the next day of availability for the parties and panel which was unfortunately not until July.

56.    We move on to consider the appeal itself and start with our findings of facts.

The facts

Mr Atherton’s advisers

57.    N T Advisors (‘NTA’):  Mr Jenner was a partner in NTA which promoted the Romangate scheme in which Mr Atherton participated.  He accepted that NTA marketed aggressive tax avoidance schemes, by which he meant that taxpayers did not incur the economic cost of the transaction, and that Romangate was such a scheme.  He considered the firm’s schemes to be within the letter but not the spirit of the law.  NTA would not promote any scheme unless it had received an opinion from Mr R Bretten QC that it was more likely to succeed than not.

58.    In around late 2008/early 2009, Mr Jenner became aware that Mr Atherton was interested in implementing the scheme, although at this point Mr Atherton’s contact with NTA was with Mr Jenner’s partner, Mr Mehigan.  Later on, after implementation, Mr Jenner was Mr Atherton’s representative in interactions with HMRC over the scheme.  Mr Jenner, but not Mr Mehigan, gave evidence.

59.    Fitzgerald and Law (‘F&L’):  F&L were a firm of accountants, who at the time Mr Atherton implemented Romangate in January 2009, had already for a number of years been retained by Mr Atherton to complete his tax returns.  Mr Cockburn, a partner at F&L, was at the time the partner with responsibility for Mr Atherton’s return and it was Mr Cockburn who completed and filed on 30 January 2009 Mr Atherton’s return for 2007/08.  We had no evidence from Mr Cockburn or anyone else at F&L.

Mr Jenner’s evidence

60.    The application to admit the first two of Mr Jenner’s statements was made before the third day of the hearing on 6 July but without a great deal of warning (the statements were dated 22/6/16 and 30/6/16).  HMRC did not object to their admission although Ms Balmer did suggest that she might ask the Tribunal to place less weight on Mr Jenner’s evidence on the basis the short notice had denied HMRC the opportunity of seeking rebuttal evidence.  We indicated our preliminary view that, while the short notice might be grounds on which to oppose the admission of the evidence, we did not see how it could be grounds for an application to put less weight on the evidence.  In the event, Ms Balmer did not pursue this matter, perhaps because the hearing did not finish on the third day but was adjourned again for a couple of months.  HMRC did then seek to admit rebuttal evidence, at extremely short notice, although as explained below at §§106-108, we did not admit it.  In any event, the fact that HMRC had little warning about the application to admit it did not cause us to put any less weight on Mr Jenner’s evidence than we would otherwise have done.

61.    Mr Jenner’s evidence included evidence about advice received from counsel, Mr Rex Bretten QC.  Although, as we have said, HMRC did not oppose the admission of the evidence, having heard Mr Jenner’s oral evidence Ms Balmer applied during the third day for disclosure of the instructions to Mr Bretten and an adjournment of the hearing.  The appellant then disclosed the instructions and a note of the conference made by NTA, and Mr Jenner’s cross examination continued.  But it became clear that the hearing would not finish on day three.  Mr Atherton had still not given his evidence, and a question arose in respect of the instructions to Mr Bretten which could not be resolved during the hearing.  The question arose because the instructions referred to specific boxes on a tax return in 2005/06 and no one could be certain what those boxes were, yet it was potentially very important to know whether Mr Bretten had advised a taxpayer to put a standalone claim in a box that would feed into the self assessment calculation.   This question was ultimately resolved as explained below at §95. 

62.    These factors led to the second adjournment of the hearing.  During that adjournment, the appellant applied to admit a third witness statement from Mr Jenner.  HMRC did not oppose this application either and Mr Jenner was re-called to give further evidence in September.

63.    On the whole, we considered Mr Jenner to be a reliable witness. His evidence was challenged extensively, and on Day 4 it was put to him that what he was saying was inconsistent with what he had said on Day 3.  In particular, on Day 3 he accepted that Box 3 did not, on its face, include Year 2 employment losses yet on Day 4 he indicated that it was proper to put a standalone claim for Year 2 employment losses in Box 3.  He explained that he had researched the point since the Day 3 hearing and had discovered that in more recent years HMRC had accepted that, there being no other more appropriate box for Year 2 standalone claims, it was proper to put a standalone claim for Year 2 employment losses in Box 3.  We accepted that to the extent that there was inconsistency in what he was saying, it was explained by his change of view, and did not indicate any unreliability in his evidence generally. 

64.    We did have a few fairly minor reservations over his evidence, one of which we discuss and resolve at §79 below, relating to the question of whether the white space disclosure covered box 20.  There was also some lack of clarity over the advice NTA gave to F&L and Mr Atherton on exactly how to force a claim into a tax return and we discuss and resolve that evidence at §§174-184. We also note, although it is irrelevant to any issue in this appeal, that Mr Jenner was of the view that the three Court of Appeal judges in Cotter had agreed with him that it was lawful to ‘force’ a Year 2 loss claim into a Year 1 tax return:  but he was not right on this as the Court of Appeal expressly chose not to give a ruling on this point:  [33-34] of their decision. 

Mr Atherton

65.    Mr Atherton’s career had been in financial institutions and now and at the time of the facts in issue he was a partner of a hedge fund.  He agreed that before becoming a partner he had filed his own tax returns, appointing F&L when his income had substantially increased on joining the hedge fund partnership.  We consider that this background meant Mr Atherton had a reasonably good lay understanding of tax returns and tax matters.

66.    He was introduced by F&L (via an intermediary) to NTA because he wished to avoid tax on what was to him unusually large partnership income in 2007/08 of some £5 million.  He knew the scheme was devised by NTA (and not F&L). He was given and read Mr Bretten QC’s advice on the scheme.  He understood the scheme involved fictional employment, used to create losses, and that those losses were nothing to do with his partnership.  He knew the scheme was risky in the sense HMRC were bound to enquire into his use of it and that ultimately it might not survive a challenge.  It was a risk he was happy to take. 

67.    His 2007/08 tax return was prepared by F&L.  They explained to him that he had two options on how to claim the relief generated by the scheme he had entered into.  He could make a standalone claim for relief which would result in his paying tax upfront and later obtaining a repayment if and when HMRC accepted the scheme was effective:  or he could claim the relief in his 2007/08 self assessment to avoid paying any tax on his 2007/08 income when it became due on 31 January 2009.  He opted for the latter course.

68.    He knew that it was, broadly, NTA’s advice that he could put the claim ‘in’ his  tax return and reduce his 07/08 self assessment, albeit F&L were responsible for completion of his tax return.  He had never seen Mr Bretten’s advice on putting the claim ‘in’ the tax return but he knew the advice had been given to NTA and he knew F&L had a copy of it.

69.    He met with Mr Cockburn of F&L to discuss his tax return prior to its submission.  He says that he did notice that on its face Box 20 was not appropriate to the loss which was being claimed, although he also said that he considered that Box 20 could be read as referring to losses arising in 2008/09.  He said that any concerns he had about how the return was completed were assuaged by what he considered to be a very full disclosure in the white space.  His view at the time was that the scheme was risky:  he had not really considered the manner in which he completed his tax return as risky.

70.    It was put to him that there was inconsistency in his evidence and contemporaneous email exchanges between NTA and F&L.  His evidence was that while he preferred not to pay the tax upfront, his main concern was avoiding the tax in principle.  The email exchange, on the other hand, indicated it was important to him to avoid an upfront tax charge.  We don’t think that this inconsistency indicates that his evidence was unreliable:  the explanation could simply be a difference of emphasis by different people.  His advisers saw his preference as more significant to him than it was.  Nothing turns on it in any event.

71.    Broadly we accepted his evidence; one slight reservation we had was his evidence over whether he had considered the appropriateness or otherwise of box 20 at the time of completion of his tax return:  we discuss and resolve this at §139.

Mr Roger Taylor and Mr Nigel Clarke

72.    Although recorded here, we actually heard Mr Taylor’s and Mr Clarke’s evidence before the submission of no case to answer.  Nevertheless, it is relevant only to the substantive issue (was there a discovery within the meaning of s 29(1)?) and so we record it here with the rest of the evidence.

73.    Mr Taylor issued the assessment the subject of the appeal.  Mr Gordon submitted in his skeleton that the Tribunal should be wary of his evidence, and the evidence of the other officer, Mr Clarke, because Mr Taylor’s contemporaneous letters indicated that he had made the discovery and issued the assessment which was inconsistent with the witness statements of both officers which said the discovery was really made by Mr Clarke, and Mr Taylor merely acted as a ‘front’ topping and tailing Mr Clarke’s letters to taxpayers.

74.    In oral evidence, Mr Clarke explained that, at the time in question, HMRC left the investigation of avoidance schemes to dedicated teams; he had been in the team investigating Romangate.  However, these teams were not given assessment powers.  The exercise of formal powers, such as opening enquiries and assessing was left with local officers, such as Mr Taylor.  So Mr Clarke would write the letters to Mr Atherton which would be sent out in Mr Taylor’s name.

75.    We agree with the appellant that the discovery letter in particular is not clear in that, although it did refer to specialist advice being taken, the ‘I’ appeared to be referring to Mr Taylor, who signed the letter, when it was actually referring to Mr Clarke.  Nevertheless, we do not put less weight on the evidence of either officer because of this misleading way of writing letters. It reflected nothing more than HMRC’s method of operating at the time.

76.      We find Mr Taylor took over from a Mrs Dawson as officer in charge of Mr Atherton’s tax affairs around the end of 2009 and early 2010.  His was largely an administrative role, issuing the assessments when and in the form he was instructed to do so by Mr Clarke. We accept his evidence.

77.    Mr Clarke was the officer who made the ‘discovery’ (if it was a discovery in the sense meant by s 29(1).)  We accepted his factual evidence.

The appellant’s 2007/8 tax return

78.    We have already described above the 3 relevant entries on Mr Atherton’s 2007/08 tax return, being the entries in Box 3 of the additional information pages, Box 20 of the partnership pages, and the white space disclosure.

79.    Nothing we heard in evidence or submissions from the appellant caused us to change our view of the return.  In particular, in so far as it was Mr Jenner’s evidence, rather than merely an opinion he expressed, that the white space disclosure did cover Box 20, we do not accept that evidence as reliable.  It is inconsistent with the disclosure on its face and with the other evidence as explained in the next two paragraphs.

80.    On its face, the disclosure dealt with both the Romangate scheme and the way the relief was mentioned on the return.  It dealt with the latter twice:  in both cases stating (as we have already mentioned) that the claim was made in box 3 due to there being no other more appropriate box and even though HMRC might not consider the loss to be a trading or capital loss.  Nowhere in the disclosure was any mention made of box 20, or of inserting an employment loss claim into a box designed for partnership losses in the earlier year.  A reader of the white space disclosure was directed to read the box 3 entry:  there was nothing to put the reader on alert about an entry in box 20 of the partnership pages.

81.    Moreover, we do not accept that the white space disclosure was ever intended to relate to the box 20 entry.  That would be inconsistent with Mr Jenner’s other evidence that, firstly, NTA provided the identical white space disclosure to all users of Romangate whether or not they intended to force the claim into the tax return, and secondly, that NTA neither advised nor knew that F&L would use box 20 to force the claim into Mr Atherton’s tax return.

Events after the submission of the 2007/8 tax return

82.    The return was submitted on 30 January 2009 and, we find, considered by HMRC fairly shortly thereafter.  There was an internal email exchange between Mr Clarke and another officer.  While it seems on 3 March 2009, Mr Clarke was doubtful about the effect of Mr Atherton’s tax return for 07/08, by 20 March 2009 he had concluded:

...We cannot repair the return – whilst confusingly worded what we have here is an actual claim to carry 09 losses back to 08.  You will need to set the no repayment signal.  Process the return and arrange for an enquiry to be opened under the provisions of Sch 1A TMA 1970....

83.    On 7 May 2009, as we have said, HMRC opened an enquiry into the loss relief claim using their powers in Schedule 1A TMA.  Those powers allow HMRC to enquire into standalone claims, in other words, to enquire into claims not made in a tax return.

84.    HMRC wrote again to the appellant on 30 September 2009, a few months after the Sch 1A enquiry was opened.  Mrs Dawson said:

Further to my letter of 7 May 2009, I have removed the losses of [Box 3 figure] from the 2007-08 Tax Return and enclose a revised computation showing tax now due of £2,055,777.28.  The changes are to ensure that no effect is given to the loss claim before the enquiry is completed.  Schedule 1A 4(3)(a) TMA 1970 gives us authority to do this.

Enclosed with this letter was the computation which showed no loss claim and the tax liability of just over £2million, compared to Mr Atherton’s tax calculation submitted with his tax return which showed a nil tax liability.

85.    Fitzgerald and Law must have queried this in a telephone call, as on 14 October 2009 HMRC wrote again to the taxpayer in response.  The letter was quite long and we do not reproduce it but we find it was the writer’s clear and consistent view that a claim to carry back a 2008/9 loss to the previous year was a ‘standalone claim’ and even if actually mentioned in the 2007-8 return, it was not (within the meaning of the legislation) ‘included’ in the return, and in reality the taxpayer had done nothing more than use the return to communicate to HMRC his standalone claim. The letter went on to say that, as HMRC were enquiring into the standalone claim under Sch 1A, and had chosen not to give effect to it pending closure of the enquiry, what HMRC saw as Mr Atherton’s self-assessment liability of just over £2million was due and payable.

86.    A reply, it seems from Fitzgerald and Law, dated 6 November 2009 stated:

“We are informed by NT Advisors, that they continue to dispute (based on their own QC’s opinion) the fact that a claim, physically made within a return, is in fact determined by HMRC to be a stand-alone claim....”

87.    HMRC were clearly not persuaded on this point as on 20 November 2009 they issued a claim form in the County Court against the appellant seeking to collect the £2million plus tax.  The appellant defended this on the basis he made a loss relief claim ‘within’ his 07/08 tax return.  It appears a stay of action was granted in the County Court.

88.    HMRC did open a s 9A TMA enquiry into Mr Atherton’s 07/08 tax return on 21 April 2010.  It was accepted by HMRC that this enquiry was ineffective to challenge the loss claim because it was opened too late, although no one took issue with this at the time. Mr Clarke explained that there was a computer error which had led HMRC at the time to consider the enquiry timeous. More relevantly, while Mr Clarke knew the enquiry was being opened, he had not requested that it be opened, and we find it was not opened because of the Romangate scheme or the Box 20 entry.   As was clear from the letter opening the enquiry, it was an enquiry raised into Mr Atherton’s savings and investment income.  We find the enquiry was triggered because HMRC had received (erroneous) information that Mr Atherton had under-declared his interest. The enquiry showed Mr Atherton had correctly returned his savings and investment income and it was closed without any amendment being made to Mr Atherton’s return.

89.    In the same year, Mr Clarke opened a general s 9A TMA enquiry into the appellant’s 2008/09 return, and sought various pieces of information from the appellant.  HMRC considered, we find, that because Mr Atherton had made, in their view, a standalone claim for a 08/09 loss to be carried back to 07/08, the loss should have been stated in his 08/09 return, yet no mention was made of it.  (Of course, its omission may have been because by the time of filing of the 08/09 return retrospective legislation blocking the scheme had been enacted).  This enquiry is not relevant to issues in this appeal and we don’t mention it again.

90.    On 19 December 2013, Mr Jenner of NTA wrote to HMRC accepting that the Romangate scheme failed due to the retrospective legislation, the challenges to which had by then failed, although Mr Jenner did not accept that the users of the scheme were necessarily liable to pay the tax.  Indeed, in February 2014 the appellant made it clear to HMRC that he considered the Sch 1A and the s 9A enquiries were invalid.

91.    Mr Clarke then consulted with technical colleagues and was informed at this time that the Sch 1A enquiry was invalid.  His evidence is that he understood that the Sch 1A enquiry was invalid because of the late 2013 Supreme Court decision in Cotter mentioned above.  He considered he had therefore newly made a discovery of an insufficiency in an assessment under s 29 TMA and arranged for Officer Roger Taylor to issue, in early 2014, the discovery assessment which is the subject of this appeal.

Counsels’ opinions on forcing a Year 2 claim into a Year 1 self-assessment

92.    Mr Bretten QC’s advice:  Mr Bretten QC advised NTA that Romangate was more likely to succeed than not. As we have said, Mr Atherton saw a copy of that advice (§66). That advice is not relevant to any issue in this appeal and we were not shown a copy of it.

93.    What we were shown were instructions to, and an opinion by, Mr Bretten in 2007.  We have referred to this above at §61. This opinion was obtained long before Mr Atherton was a client of NTA’s and was therefore not given in respect of him.  It was also accepted that the instructions related to a different scheme to Romangate, albeit one where similarly there was a manufactured loss in Year 2 being claimed in the Year 1 tax return. 

94.    There was some dispute as to what Mr Bretten’s opinion actually said.  The appellant’s view was that Mr Bretten advised that it was lawful to ‘force’ a carry-back loss claim into the earlier year’s return by utilising an unrelated box on the return form.

95.    Indeed, the instructions to Mr Bretten asked him to assume that NTA’s hypothetical, or at least unnamed, client had implemented the scheme in Year 2 and completed his tax return for Year 1, including the claim for the Year 2 loss in both the white space disclosure and in box 15.8/18.8 to take the effect of the loss into the calculation of tax due for Year 1.  It was accepted, after we reconvened in September, that box 15.8/18/8 of the tax return for 05/06 (the year to which the instructions related) was a box, similarly to Box 20 in this appeal, which was unrelated to the claim being made and a box which fed directly into the tax calculation.  It was a box which was, however, quite unrelated to box 20 and partnership losses.

96.    Nevertheless, reading the instructions, it is clear that Mr Bretten was neither asked to advise on, nor did advise on, whether it was right to co-opt an unrelated box in this manner to claim an unrelated Year 2 relief.  He was asked to advise on the effect of doing so on the taxpayer’s liability to pay tax for the tax return year, and in particular whether HMRC would be obliged to give immediate effect to the loss claim or whether Schedule 1A or 1B would mean that HMRC could delay giving effect to the claim.  In short, he was asked to advise whether such a claim was made ‘in’ the tax return.

97.    The conference between Mr G R Bretten QC and Mr Anthony Mehigan of NTA took place on 16 August 2007.  Mr Bretten’s advice (apparent from the settled note dated 2 October 2007) was that the effect of doing as described was to put the claim into the tax return, and Sch 1A was not relevant.  It was also his opinion that taxpayers ‘were within their rights’ to offset the specific loss arising in Year 2 which was generated by that scheme (under s 574 ICTA) against their Year 1 income and thus reduce what they owed on 31 January in Year 2.

98.    We find that this shows it was Mr Bretten’s opinion that taxpayers were entitled to put a particular kind of Year 2 loss ‘in’ the Year 1 tax return; and that utilising an unrelated box was effective to force the Year 2 loss into the Year 1 tax return.  What he was not asked to express a view on and did not express a view on was whether a taxpayer was fulfilling his duty of care in completing his tax return in the manner described and in particular if the taxpayer entered an unrelated Year 2 loss into a box restricted to a different loss arising in Year 1.  It is also worth noting that he was told to assume there was white space disclosure although the extent of it was not specified.

99.    In retrospect, the Supreme Court decision in Cotter shows that Mr Bretten was wrong to consider that taxpayers were entitled to put a Year 2 loss in the Year 1 tax return; however, while his view was inconsistent with HMRC’s view at the time, we do not understand HMRC to suggest, and we don’t consider, that his view at the time he gave it was unreasonable. (We don’t agree with the appellant, however, that the Court of Appeal shared it:  it was the question on which they declined jurisdiction:  §64).

100.Cotter also shows that the second of Mr Bretten’s opinions was correct:  entering a Year 2 loss in a box on a Year 1 tax return was effective to force the claim into the tax return.  So this view was not only reasonably held but shown to be correct in law.

101.Mr Rory Mullan, counsel:  we were shown a copy of an opinion by Mr Mullan dated February 2009.  However, this advice was not advice given to NTA nor Mr Jenner nor on the Romangate scheme.  It was merely a copy of advice which had come into Mr Jenner’s possession sometime after Mr Atherton’s tax return had been submitted and was not relied on when that return was submitted.  He produced it in order to show that Mr Bretten QC was not the only counsel holding the view that it was lawful to force a Year 2 claim into a Year 1 self-assessment.  We agree, as we have already said, that Mr Bretten’s view was not unreasonable at the time it was held:  §99.

102.We note in passing that Mr Mullan did not advise taxpayers to put one kind of loss in a box on their tax return which clearly related to a different kind of loss:  it was Mr Mullan’s opinion that the loss merely had to be claimed somewhere on the four corners of the return even though the form of the return did not have a particular place for the claim to be made.

103.Mr Ewart QC’s opinion: it was Mr Jenner’s evidence that he had seen a copy of an opinion by Mr Ewart, dated 31 March 2009, very similar to that of Mr Bretten.  As the opinion was not given to NTA, but was owned by third parties, he was not at liberty to show it to the Tribunal.  He did not suggest that it was relied upon by, or even known to, NTA or Mr Atherton at the relevant time:  he mentioned it to support his view that Mr Bretten’s advice on forcing a claim, while it had been shown to be wrong in the Supreme Court’s decision in Cotter, was nevertheless a reasonable opinion at the time it was given.  We have concluded that it was:  §99.  We reach that conclusion without putting any weight on this evidence of an opinion that was not produced to the Tribunal.

Who drafted the white space disclosure?

104.We find that NTA was responsible for the terms of the white space disclosure, which was adopted by F&L and placed on Mr Atherton’s return. 

What was Fitzgerald & Law’s advice to Mr Atherton?

105.We find, following Mr Jenner’s and Mr Atherton’s evidence, and in the absence of any evidence from F&L, that F&L advised Mr Atherton that it was lawful to claim his Year 2 manufactured employment loss in Box 20 of his 2007/08 (Year 1) tax return, although that box related to partnership losses for 2007/08, and that they did not advise him to make a specific disclosure relating to the box 20 entry, but implicitly advised him that the white space disclosure which was made was adequate.

Application to admit evidence of meeting notes and emails

106.On the fourth day of the hearing, HMRC applied to admit certain meeting notes and emails passing between HMRC officers and Mr Mehigan of NTA back in 2007.  Their reason, as we understood it, was to show what NTA knew in 2008/9 about HMRC’s views of forced claims in tax returns.  Ms Balmer accepted that it was unfortunate that HMRC had only discovered them, and passed them to the appellant, a few days before the reconvened hearing was due to start.

107.Mr Gordon objected to their admission on the grounds of relevance and procedural prejudice.  He considered them irrelevant because (on his view of the law) the Bessie Taube case (discussed below at §§192-194) made it clear NTA’s knowledge was irrelevant; he considered it procedurally prejudicial because HMRC had only produced the information a few days before and the appellant had not had proper time to consider it.

108.We were satisfied that the new material was potentially relevant:  HMRC did not accept what Mr Gordon said about Bessie Taube and s 29(4) was correct.  If we were to agree with HMRC on this, NTA’s knowledge might be relevant.  However, we refused the application because we considered that the procedural prejudice of admitting such late evidence could not be cured and was significant.  Mr Jenner was not alleged to be a party to the meeting or the emails and so the information was new to him and yet the appellant was given no time to investigate the possibility of rebuttal evidence.  The evidence was not admitted.

The law

The law on carelessness

109.While s 29(4) follows on from s 29(1), we deal with the issue of carelessness first as it follows logically from our decision that HMRC had made out a prima facie case of carelessness.

110.As we understand it, the appellant’s case in rebuttal was that it was not careless to make the carryback loss relief claim in Box 20 and therefore reduce his self assessment because:

(1)   it was reasonable at the time to think he was entitled to do this, albeit in retrospect the Cotter decision has made it clear it was wrong to do so; and

(2)   in any event, if it was careless, it was not carelessness by anyone whose carelessness mattered for the purposes of s 29 TMA.

111.So we go on to consider, firstly, whether there was carelessness, and if there was, whether it was the responsibility of the appellant’s or someone whose carelessness can be attributed to the appellant.  We note here that the below discussion is predicated on the basis that carelessness by Mr Atherton or F&L is enough for s 29(4).  The appellant accepted this.  S 29(4) catches carelessness

‘...by the taxpayer or a person acting on his behalf”

and Mr Gordon accepted that F&L was a person acting on behalf of Mr Atherton.

112.The carelessness must be relevant carelessness so we first consider what is relevant carelessness.

Carelessness and causation

113.S 29(4) requires the carelessness to cause (‘brought about ...by’) the ‘situation mentioned in subsection (1)’. 

...the situation mentioned in subsection (1) above was brought about carelessly or deliberately by the taxpayer or a person acting on his behalf

What is the situation mentioned in subsection (1)?  What must the carelessness cause?

114.It seems clear to us that the ‘situation’ are the three situations mentioned in (a), (b) and (c) of s 29(1): Those three situations are (a) an underassessment (b) an insufficiency in an assessment and (c) an excessive claim to relief.  Any of those three subsections could describe the entry in Box 20.  Box 20 claimed a relief to which the appellant (in retrospect) was not entitled leading to an underassessment/insufficiency in his self-assessment/excessive claim to relief.  For simplicity we will refer to the ‘situation’ as being the insufficiency in the self-assessment.

115. S29(1) does also talk of discovery of the position mentioned in (a), (b) and (c) and a right to assess consequent on that discovery.  The ‘situation’ referred to in s 29(4), however, is clearly not a reference to the discovery of the insufficiency, let alone the assessment consequent on that discovery.  S 29(4) is clearly referring to the situation discovered by HMRC, and is not treating the discovery as the situation.   It would not be logical to interpret the section otherwise because Parliament cannot have thought the carelessness would ‘cause’ the discovery and/or assessment: it is obvious that the situation referred to is the insufficiency in assessment.

116.In other words, the carelessness must be the cause of the insufficiency in the assessment.  The carelessness is not required to be the cause of any lack of knowledge of the insufficiency by HMRC.  In other words, if a self-assessment is carelessly completed and as a result insufficient, the fact it does not fool HMRC would not prevent s 29(4) from applying.

117.The point is significant in this appeal.  This is because it is clear that, even if it was careless for Box 20 to be completed, HMRC were not fooled by the completion of Box 20.  It is clear from the start that they considered the Box 20 figure related to the Year 2 carry back loss relief claim, being a claim which they considered the appellant was not entitled to make (§82).

118.The fact HMRC were not misled by the Box 20 entry does not prevent s 29(4) being applicable.  The question is not whether the carelessness (if there was any) fooled HMRC into accepting a claim which should not have been made:  the question is whether the carelessness (if there was any) resulted in the self-assessment being insufficient.

119.The alleged carelessness is the entry of a figure in Box 20.  Did the completion of box 20 cause the insufficiency in the assessment?  We find it did.  The completion of box 20 led to the insufficiency because it directly fed into Mr Atherton’s self assessment, as we have said, leading to a nil tax liability rather than his true tax liability of a little over £2million (§6).  So if entering the figure in box 20 was careless, that carelessness is relevant carelessness for s 29(4).

120.We note that Mr Gordon at one point put the case that it was not the completion of Box 20 so much as the submission of the tax return (with the incorrectly completed box 20) which caused the insufficiency.  We simply don’t see any point in this:  the insufficiency was the result of the submission of Mr Atherton’s tax return with the incorrectly completed Box 20.  Incorrectly completing Box 20 caused the tax return which was submitted to be insufficient.  So the question is whether it was careless to make the entry in box 20 which is the same question as whether it was careless to submit the tax return with the entry in box 20.

121.Box 3 and carelessness:  HMRC suggest that putting the relief claim into box 3 was careless; they also suggested that completing both box 3 and box 20 was careless as it amounted (they said) to a duplicate claim for the same relief.

122.We find that even if Box 3 had been completed carelessly, it was not relevant carelessness for s 29(4).  The carelessness relevant to s 29(4) had to bring about the situation mentioned in s 29(1).  That situation was the insufficiency in the tax assessment.  As we have already explained, a standalone claim, even one made ‘on’ a tax return, is not made ‘in’ the tax return and does not affect the self assessment.  The self assessment was insufficient but not because of the Box 3 entry, which did not feed into the self assessment calculation.

123.Had the appellant done no more than make the Box 3 entry, there would have been no insufficiency in the 2007/08 self-assessment and the Sch 1A enquiry would have been effective to defeat the claim.  We will not consider whether the box 3 entry was careless, as such carelessness is irrelevant to the discovery assessment.

124.Nor did making duplicate claims cause the insufficiency:  it was the box 20 entry which caused the insufficiency.  The completion of box 3 was irrelevant to the question of causation of the insufficiency.  However, the completion of box 3, in our view, is relevant to the question of whether it was careless to complete box 20 as we explain below at §§158-161.

125.White space disclosure and causation: it was a part of the appellant’s case that the entry in box 20 was not careless because of the disclosure made in the white space of the tax return.  We deal with our factual findings relevant to this submission below.  But we consider the submission from a legal point of view here because of a possible interpretation of what was said by the Upper Tribunal in Moore [2011] UKUT 239 (TCC).  In that case, the taxpayer put the wrong amounts in boxes on the return, although he made a disclosure with his return to explain how he had arrived at those figures.  Had an officer read the disclosure, he ought to have realised that the taxpayer’s entries in the boxes were wrong. The question was whether the taxpayer had been careless within the meaning of s 29(4) so that HMRC could raise a discovery assessment. 

126.The Upper Tribunal said:

[16] It seems to me that ...it is necessary to look closely at what [29(4)] provides.  It allows an officer to assess where ‘the situation mentioned in [s 29(1)]...is attributable to ....negligent conduct on the part of the taxpayer’.  The ‘situation mentioned in [s 29(1)]’ includes... ‘an assessment to tax is ...insufficient.’  The assessments in this case....were based not upon what [the taxpayer] wrote on the additional sheets, but on what he entered in the boxes.  ... His setting out the information on an additional sheet would have given Mr Moore the protection of [s 29(5)], but not of [s 29(4)].....

127.This might be read as meaning that where there was a course of conduct (in that case, making a wrong entry in a box on a return plus giving disclosure), the Tribunal should only assess carelessness as if that course of conduct comprised only the act which actually caused the insufficiency.  This appears to have led the Upper Tribunal to exclude consideration of the disclosure when considering carelessness.

128.We do not think it should be read that way.  That would be unfair to taxpayers.  Whether it was careless to do the act that caused the insufficiency must be judged as a whole:  the taxpayer should be judged on his course of conduct.  Was his course of conduct careless? What the Upper Tribunal must have meant was merely that disclosure of a careless entry won’t prevent that entry being careless.  White space disclosures do not (necessarily) negate an entry being careless, but the Tribunal is entitled to look at the taxpayer’s overall actions to determine whether a wrong entry was made carelessly.

129.In other words, we consider as a matter of law that Moore does not prevent the Tribunal considering Mr Atherton’s actions in both (a) giving white space disclosure and (b) entering the loss in box 3, when considering whether it was careless for him to make the entry in box 20.

What is carelessness?

130.The test is objective.  In Anderson [2009] UKFTT 206 at [22], which was approved by the Upper Tribunal in Moore [2011] UKUT 239 (TCC), the FTT said:

The test to be applied, in my view, is to consider what a reasonable taxpayer exercising reasonable diligence in the completion and submission of the return, would have done.

131.In Gedir [2016] UKFTT 188 (TC)  the FTT said:

[19] in my view carelessness can be equated with ‘negligent conduct’ in the context of discovery assessments....

132.In so far as there is any difference between these two statements, we prefer the Anderson test of carelessness as it is more descriptive, except that we would add to it that the reasonable taxpayer would know that it was his duty to submit a complete and correct tax return and would intend to do so, although that is really implicit in what the Judge said in Anderson.

133.We understood Mr Gordon to say that mistakes were not necessarily careless, and innocent mistakes were not careless.  We agree that mistakes are not necessarily careless.  This is consistent with what we said in the earlier interim application that we considered that mistakes in tax returns, and in particular the mistake in this tax return, were prima facie careless, as we expressly allowed for the possibility that the taxpayer could rebut that prima facie case.  It is possible, therefore, that a mistake was not carelessly made, but it is for the appellant to discharge the evidential burden on that.

134.We do not agree that innocent mistakes cannot be careless, unless innocent is taken to mean a mistake without negligence.  However, if ‘innocent’ is taken to mean, as it usually is, a non-intentional error, then clearly some innocent mistakes could be carelessly made.

135.So the question is whether the insufficiency in Mr Atherton’s assessment was an insufficiency which arose from behaviour which would not have been the behaviour of a reasonably diligent taxpayer, mindful of the need to make a complete and accurate tax return.  In other words, would such a taxpayer have made the entry in box 20 in the circumstances in which Mr Atherton did?

136.As we have said, the entry in box 20 was prima facie careless for two reasons: firstly, the loss related to Year 2 but box 20 related to Year 1; secondly, the loss was an employment loss but Box 20 related to partnership losses.  The appellant put forward a number of explanations of why in his view the entry was nevertheless not careless and they were:

(a)    It was reasonable to think, and the appellant did think, that box 20 related to year 2 losses;

(b)   There was no other box in which to enter the loss;

(c)    There was white space disclosure to explain the entry.

(d)   Mr Atherton and/or F&L relied on advice it was lawful to enter the loss in the wrong box.

We will deal with each in turn.

Reasonable to think box 20 related to year 2 losses?

137.We were not referred to HMRC’s guidance notes on completion of the tax return at the point that Mr Gordon made his submission of no case to answer, so we did not mention them above.  However, we find that the notes to the partnership pages said, in respect of box 20:

Loss from this tax year set-off against other income for 2007-8

You can claim relief for the 2007-08 loss by reducing your other taxable income for 2007-08 by entering the amount you are claiming to offset in box 20.

138.We find, looking at the clear wording of box 20 and the notes to it, it is clear ‘this tax year’ is the tax year being returned (2007/8).  This is doubly clear when the partnership pages and tax return as a whole are considered as throughout it is clear the return is seeking information about the tax year being returned. We find that no reasonable person completing the 2007/08 tax return could reasonably think Box 20 related to anything other than a partnership loss incurred in the year 2007/08.

139.Mr Atherton, however, said in evidence that he would read ‘this tax year’ as referring to the year in which he was completing the tax return (ie 2008/09) and not the tax year being returned (2007/8).  Mr Gordon said that because this was Mr Atherton’s evidence, we could not go behind it.  We do not accept that.  Firstly, Mr Atherton appeared really to be expressing his current view and only assuming that this was what he would have thought at the relevant time.  It was not, in our view, reliable evidence of what he thought at the time.  Secondly, and more importantly, the question is whether the tax return was completed carelessly and that is an objective and not subjective question.  It does not matter if Mr Atherton actually misunderstood the tax return but whether it was reasonable for him to misunderstand the tax return in this way.  And as we have said in the previous paragraph, it would not have been a reasonable mistake to make.

140.And in any event, there was no suggestion that Mr Atherton made any mistake about the fact that Box 20 was on its face limited to partnership losses, and Mr Atherton admitted he knew the loss Romangate generated was an employment loss. 

141.So we find that this defence does not rebut the prima facie case that the insertion of the Romangate loss into Box 20 was careless.

Does white space disclosure negate careless mistake?

142.Factual findings:  The appellant’s position was that even if it was careless to make the entry in box 20, that carelessness was counteracted by the explanation given in the white space disclosure.  As a matter of fact, we do not accept that.  As we have stated before, the white space disclosure did not refer to the entry in box 20.  We find it related solely to the box 3 entry.  It is not surprising that it did not refer to or relate to box 20 in that the tenor of Mr Jenner’s evidence is that the disclosure in Mr Atherton’s return was identical to the generic disclosure recommended by NTA to all their clients, only some of whom may have gone on to ‘force’ the claim into the tax return rather than merely make a box 3 entry, and in any event NTA did not even know that F&L decided to utilise box 20.

143.We had no explanation from F&L why no further disclosure was given, and in particular why HMRC’s attention was not drawn to the entry in box 20 being a 2008/09 employment loss rather than a 2007/08 partnership loss to which the box actually referred, and moreover that it duplicated the entry in box 3.

144.Mr Atherton’s evidence seemed to be that he considered the white space disclosure adequate to cover box 20 as well as box 3.  We do not consider that a reasonable view.  The disclosure made no mention of box 20.   There was no mention that the loss entered into box 20 was neither a partnership loss nor a loss arising in 2007/08 nor that it duplicated the box 3 entry.

145.Therefore, as a matter of fact, we find that no explanation of any kind was given to HMRC for the wrong entry in box 20.  This was not the action of a reasonable taxpayer mindful of his obligation to make a correct return.  The white space disclosure is no defence to HMRC’s prima facie case that the entry in box 20 was carelessly made.

146.We have already referred to the point on causation and do not repeat it fully here:  we accept that the wrong entry in box 20 did not appear to mislead HMRC or at least not for more than a few weeks.  By the end of March 2009 (§82), it appears Mr Taylor was well aware that the box 20 entry related to the Romangate scheme and not Year 1 partnership losses, although we do not know when and why he formed that view.  That the box 20 entry did not mislead HMRC is, for the reasons we have given, irrelevant to the question whether it was careless to enter the Romangate losses into box 20 without giving HMRC any kind of explanation.

There was no other box in which to insert the loss

147.We think that it would have been obvious to F&L, and to Mr Atherton, who had some basic understanding of tax matters, that the Romangate loss was not a 2007/08 partnership loss and therefore obvious that it did not meet the criteria for being entered into Box 20.

148.The tenor of Mr Jenner’s evidence was that Mr Atherton was understood to want to ‘force’ his loss claim into his 2007/08 return in order to get immediate tax relief, and that F&L knew that NTA’s and Mr Bretten’s view was that it was effective to do so.  The obvious inference is that F&L decided it was appropriate to force the claim into the return by entering the loss into Box 20, as that would feed into the tax calculation.  Mr Atherton’s evidence is that he understood that was why the entry was made in box 20.

149.Mr Gordon’s case, as we understood it, was that it was not careless to do so, despite Box 20 being for quite different losses, because it was reasonably believed to be lawful to make the claim in the Year 1 tax return, and there was no appropriate box on the tax return to do so.  So either box 20, or another box unrelated to Year 2 employment losses, would have had to be utilised for that purpose.

150.Put another way, HMRC had failed to provide a box ‘in’ the return which permitted Year 2 losses to be claimed in the Year 1 self-assessment.  While in retrospect they were right to omit such a box from the return form, the appellant’s case was that it was reasonable in 2009 to consider that it was wrong for HMRC to omit such a box, and therefore reasonable to co-opt an unrelated box.

151.There is some force in this view. We have said that, even though it was not HMRC’s view at the time, and was ultimately shown to be wrong, it was not unreasonable to hold the view in early 2009 that Year 2 losses could be claimed in the Year 1 self-assessment, obtaining an immediate benefit.  Ms Balmer’s case was that NTA was well aware that their view was inconsistent with HMRC’s:  we think, even if true, it is irrelevant.  Taxpayers are entitled to hold reasonable views which are at variance with HMRC’s.  So, as it was reasonable to hold the view that the Year 2 loss could go into the Year 1 return and as there was no suitable box on the return in which to insert the loss in question, how else could the appellant force his loss into his return other than by co-opting an unrelated box, such as box 20?

152. In Rouse (§20 above) the taxpayer attempted to ‘force’ his Year 2 claim into his Year 1 self-assessment by merely deducting the loss relief claim from the self-assessment calculation:  this was held by the Upper Tribunal to be ineffective to ‘force’ the claim into the self-assessment.  So how else could a Year 2 claim be forced into the Year 1 tax return other than by co-opting an unrelated box?

153.It would not be possible to adapt an electronic return form in any event; and while not canvassed as a possibility at the hearing, we don’t think the taxpayer could amend a paper return to include a new box for his Year 2 claim as we think under s 8(1) TMA the return is the information which HMRC requires.  In other words, it seems to us that the only way the appellant could force his Year 2 claim into his Year 1 tax return was to use an unrelated box such as box 20.  And, as we have said, his view at the time that it was lawful to force his claim into his return was not unreasonably held although ultimately shown to be wrong.

154.So, as we have said, there is force in the appellant’s submission that it was reasonable behaviour to put his Year 2 employment loss into box 20.  Utilising box 20 or another unrelated box was the only way Mr Atherton could force his claim into his Year 1 return, and he reasonably believed he was entitled to do this. 

155.But the question is whether the entry was carelessly made and that means we have to consider objectively whether a reasonably diligent taxpayer, conscious of his obligation to make a complete and correct return, and similarly of the reasonable view he was entitled to force his claim into the Year 1 tax return, could have acted as Mr Atherton did.  We do not think that such a taxpayer would have declared his year 2 employment loss in a box related to year 1 partnership losses, without expressly making it very clear to HMRC what he was doing.   He would have explained in a covering letter or white space disclosure what he was doing and why, so it would have been clear to anyone reading the return that he was not claiming that he had a Year 1 partnership loss. 

156.We have already mentioned that it is not relevant that HMRC were not misled by the appellant’s failure to explain his entry in box 20.  The question is whether it was careless of him to complete box 20.  We think it was careless because a conscientious taxpayer mindful of his obligation to file a correct and complete tax return would not have put the unrelated loss into box 20 without any explanation to HMRC. 

157.We have already stated that no explanation of the box 20 entry was given.  The white space disclosure only related to the box 3 entry.  So we reject this defence.

158.We note that the entry in Box 20 was actually careless for a second reason and that was because it was a duplicate entry.  Mr Gordon said this was a red herring because HMRC had not been misled by the duplication into thinking there were two different losses.  We think that is irrelevant:  the question was whether the entry in box 20 was carelessly made because that entry caused the insufficiency.

159.So was it careless to make an entry to claim a loss in Box 20 when the appellant had already made a standalone claim for the same loss in Box 3?  Had neither claim been challenged by an enquiry or assessment, it seems to us that the appellant would have been entitled to the benefit of double relief.  While clearly wrong to make a double claim, it would have been effective to do so.  We think a conscientious taxpayer mindful of his obligation to make a complete and correct tax return, even one who believed he was entitled to force the Year 2 loss into his Year 1 tax return, would know he was not entitled to the same relief twice, once as a deduction from his self-assessment and once as a standalone claim.  It was careless to make the same claim twice:  it was certainly careless to do so without explaining to HMRC that the two entries were duplicates of each other.  There was no such explanation as the taxpayer gave no explanation of his entry in box 20.

160.The problem for Mr Atherton is that he did not behave as we think a diligent taxpayer conscious of his obligation to make a correct return could have acted:  he co-opted an unrelated box without explaining it to HMRC; he made a duplicate claim for the same loss without explaining to HMRC that he was doing so.  We think that that means his completion of box 20 was careless.

161.We would have accepted, in the circumstances that it was reasonable to believe, and he did believe, that he was entitled to claim the Year 2 loss in Year 1, and that there was no appropriate Box in which to make such an entry, that it would not have been careless if he had explained in disclosure that the box 20 entry was for the same loss as the box 3 entry and that it related to Year 2 employment losses and not Year 1 partnership losses, and if he had disclosed why he thought he was entitled to co-opt an unrelated box in this way.  But, as he did not draw any of this to HMRC’s attention, the manner in which he completed his tax return was other than how a reasonably diligent taxpayer would have done and he was careless.

Reasonably relied on advice it was lawful to utilise box 20 for other losses?

162.It was also the appellant’s case that Mr Atherton and F&L had reasonably relied on advice from NTA and/or Mr Bretten that it was lawful to utilise unrelated boxes to claim the relief generated by the scheme.

163.This tied in with their case, which we consider below, that even if that advice was careless, the carelessness was not carelessness which could be attributed to the appellant.  In other words, the appellant’s case was that even if the advice to Mr Atherton was careless, it was reasonable for Mr Atherton and F&L to rely on it.

164.This part of their case raises a number of issues.  We consider the question of whose carelessness matters for s 29(4) below.  Here we consider whether in fact Mr Atherton and/or F&L were advised to enter the loss into box 20 and whether it was reasonable for them to rely on such advice.

165.No advice to enter loss in box 20:  the instructions to Mr Bretten concerned a different box to box 20; we accepted Mr Jenner’s evidence that he did not advise F&L or Mr Atherton to put the loss in box 20.

166.But we don’t think the question is whether there was express advice to utilise box 20 but whether there was express advice to co-opt any unrelated box.  HMRC state the case too narrowly when they imply Mr Atherton would have to show there was advice to use box 20.

167.Did Mr Bretten say it was lawful to utilise an unrelated box? Putting aside questions of to whom Mr Bretten’s advice was addressed and whether it was reasonable to rely on it, as we have said, the tenor of Mr Bretten’s advice was that it was effective to reduce the Year 1 self-assessment to utilise an unrelated box for a Year 2 loss.  He was not asked, and gave no specific advice, whether doing so breached a taxpayer’s duty of care in completing a tax return; and in any event was told to assume that there was white space disclosure (§98).

168.We find Mr Bretten did not give advice that it was lawful to declare a loss in a box that related to an entirely different kind of loss, particularly in circumstances when no explanation was to be given to HMRC that that was what had been done.

169.We find it was not reasonable to rely on Mr Bretten’s opinion as justifying the co-opting of box 20.  Even putting aside issues that Mr Bretten’s opinion was given in relation to an entirely different scheme and box on the tax return, and was not addressed to Mr Atherton, it was unreasonable to rely on it because:

(a)    It was clear from the instructions (disclosed to F&L) that Mr Bretten was told to assume that there would be white space disclosure of the entry;

(b)   The advice was not given on the basis of the relief being claimed in 2 boxes and therefore did not, for this reason as well, apply to the appellant’s circumstances;

(c)    Mr Bretten never actually advised that it was lawful to co-opt an unrelated box:  he only ever said it was effective.

170.Did NTA advise it was lawful to utilise an unrelated box?  It was clear to us that while Mr Jenner objected to the term ‘forcing’ a claim into a tax return he understood it to mean placing the loss generated by an NTA scheme into an unrelated box on the Year 1 tax return which fed into the self-assessment calculation, and that NTA advised all their clients entering schemes to generate Year 2 losses that they were entitled to do this if they wished rather than make a standalone claim.  Mr Jenner said he did not know how many of his clients had chosen to do it each way. He accepted that NTA provided to clients interested in forcing the claim into the Year 1 return, and had provided to F&L, Mr Bretten’s advice that it was effective to do so.

171.It was clear that NTA advised Mr Atherton on the implementation of the scheme and also that NTA did not itself complete Mr Atherton’s 2007/08 tax return.  Nevertheless, the letter of engagement between NTA and Mr Atherton was that NTA would assist with the making of the claim and it was clear NTA and F&L liaised to some extent over Mr Atherton’s 2007/08 tax return.  We accept Mr Jenner’s evidence that NTA had not advised F&L or Mr Atherton to utilise box 20 to force the loss into the Year 1 return, and had not known at the time that that is what F&L actually did.

172.Mr Jenner gave a great deal of evidence surrounding this issue of forced claims.  It was clearly something which he had considered lawful and a ‘justified’ means to an end.  Nevertheless, conscious in 2007 that HMRC did not share NTA’s view on this,  NTA had sought Mr Bretten’s opinion.  NTA relied on that opinion as indicating it was lawful to co-opt an unrelated box.   Mr Jenner pointed out to us that another NTA client who had utilised an unrelated box to claim a loss had not been criticised for this by HMRC or the Tribunal:  Chappell [2013] UKFTT 98 (albeit we note that in that case there was specific disclosure around the use of the unrelated box).

173.At the same time, however, his witness statement and oral evidence seemed to wish to distance NTA from F&L’s decision to utilise box 20, as he indicated that F&L had made the decision to utilise an unrelated box relying on Mr Bretten’s opinion and pointed out that NTA had never specifically advised the use of any particular box on the return and did not know F&L  had chosen to use box 20.  It was also the case that he did not know exactly what advice F&L was given by NTA because at that point the client relationship was with his partner Mr Mehigan.  Later, he stated he just handed over Rex Bretten opinion to F&L and let them decide what to do.

174.The appellant’s case, as articulated by Mr Gordon, was that F&L relied on both NTA and Mr Bretten’s advice when reaching the decision to claim the loss in box 20.  We consider that the evidence of Mr Jenner, taken as a whole, irresistibly leads to the conclusion that NTA did advise F&L, on behalf of Mr Atherton, that it was lawful to claim the year 2 loss in an unrelated box on the year 1 self-assessment return, albeit that they did not specify which box. 

175. What did NTA advise F&L about disclosure?  It was clear that NTA provided the text of the disclosure used, that Mr Jenner considered that it was adequate to cover Box 20 as well as Box 3, and that NTA advised F&L to utilise an unrelated box. On balance, we consider that NTA more likely than not were careless.  This is because it is more likely than not that having advised F&L and Mr Atherton to utilise an unspecified unrelated box for the Year 2 manufactured employment loss, it seems they did not also advise F&L and Mr Atherton, as they should have done, to make specific disclosure of the nature of the entry in that unrelated box.

176.The entry in box 20 was a double claim for the same relief claimed by box 3:  there was no suggestion that Mr Atherton or F&L had been advised by Mr Bretten to make such a double claim, and certainly not to make a double claim without explaining to HMRC that that was what was being done. 

177.But we have a different view with respect to NTA for the same reasons as given above:  NTA advised F&L and Mr Atherton to make the entry in Box 3 and provided the text of the disclosure and  that the claim could be forced into the Year 1 tax return by co-opting an unrelated box which fed into the self assessment:  it seems more likely than not that NTA failed, as they should have done, to advise that a double claim should not be made or, at the very least, if a double claim was made, that this was drawn to HMRC’s attention.

178.HMRC’s case was that F&L and Mr Atherton should have been wary of relying on NTA’s advice as they knew NTA were invested in the Romangate scheme and other aggressive tax avoidance schemes and could not be expected to give an independent view.

179.We certainly agree that F&L and Mr Atherton should not have simply followed the advice from NTA without applying some independent thought to it.  While it was at the time reasonable to accept the advice it was lawful to make the loss claim in the earlier year’s self-assessment, we think it was quite unreasonable to accept advice that it was lawful, without giving any explanation to  HMRC, to insert a year 2 employment loss into a box which was clearly related to an entirely different year 1 loss.  Such advice was obviously wrong.

180.We reiterate the point that in Chappell the co-option of the unrelated box was disclosed to HMRC.

181.Mr Atherton’s explanation for doing so seemed to be that he considered that the white space disclosure covered it.  But as we have said, it could not reasonably be read that way.  It did not explain the box 20 entry.  We had no explanation from F&L.

182.We find it was not reasonable to rely on NTA’s advice as justifying the co-opting of box 20 without specific disclosure as it would be so obviously wrong to put a Year 2 employment loss into a Year 1 partnership loss box without any kind of explanation to HMRC that F&L and Mr Atherton should not have relied on such advice; and in the alternative,  it was also  so obviously wrong to claim the same relief twice without any kind of explanation to HMRC that such advice should not have been relied upon.

Have HMRC accepted implicitly that there was no carelessness?

183.Mr Gordon also submitted that because HMRC chose not to rely on s 29(5), they had implicitly accepted there was no carelessness in completing box 20. As we understand it, his point was that by expressly not relying on the second of the two conditions for a discovery assessment (s 29(5)), HMRC accepted that the white space disclosure on the tax return was adequate.   In other words, it was the appellant’s case that HMRC could reasonably have been expected to be aware of the insufficiency on the basis of the information contained in the tax return.

184.We do not accept the premise in law.  The two conditions were clearly intended as alternatives.  That is how s 29(3) is expressed.  It would be wrong to assume that because HMRC could reasonably have been aware of the insufficiency, that the insufficiency was not careless.  As we have already said, the carelessness must cause the insufficiency; it is irrelevant whether or not HMRC are actually misled by the carelessness.

185.We note in any event that the Upper Tribunal in Moore came to the conclusion at [7] and as cited above that white space disclosure might prevent HMRC relying on s 29(5) but would not necessarily prevent HMRC relying on s 29(4).

186. Nor do we accept the premise in fact in any event.  We do not know why HMRC did not rely on s 29(5) and as our conclusion is that the white space disclosure did not relate to the box 20 entry, it is not immediately obvious to us why s 29(5) did not apply.

187.We reject this defence.

Conclusions on carelessness

188.The entry in Box 20 caused the insufficiency; s 29(4) applies if that entry in Box 20 was made carelessly.

189.We found that there was a prima facie case of carelessness when box 20 was completed; we have rejected the appellant’s case in rebuttal.  The entry in box 20 was careless.  In the same circumstances as the appellant’s, with a reasonable belief he was entitled to force his year 2 claim into his year 1 tax return, a reasonably diligent taxpayer, mindful of his obligation to make a full and correct tax return, would not have made the entry in box 20 without giving an explanation to HMRC that (a) it was a Year 2 employment loss and not a year 1 partnership loss and (b) it was the same loss as claimed in box 3.  Because the appellant did not do this, his entry in box 20 was careless; it was misleading.  It is irrelevant it did not actually mislead.

190.We find that there was carelessness within the meaning of s 29(4); we find both Mr Atherton and F&L were careless in making the entry in Box 20 and submitting the tax return with Box 20 completed as it was.  It was accepted by the appellant that F&L’s carelessness (if proved, as it has been) fell within s 29(4) as F&L were persons acting on behalf of Mr Atherton in the matter of the completion of his tax return.

191.While we find NTA did act carelessly, because, having advised it was lawful to force a Year 2 loss into the Year 1 tax return by co-opting an unrelated box, they did not go on to advise that the taxpayer should make explicit to HMRC that that is what he had done.  Nevertheless, even if Mr Atherton and F&L relied on that advice, they were careless to do so because the advice was so obviously wrong.  It was obviously wrong to put a Year 2 manufactured employment loss into a box which on its face was limited to a Year 1 partnership loss without explaining that that is what was being done and the reasons why it was being done.

192.There is, therefore, little point in going on to consider whether carelessness by NTA would have been carelessness within s 29(4).  Nevertheless, we mention our views in case this appeal goes higher.

Whose carelessness matters?

193.The appellant relied on the FTT decision in Trustees of Bessie Taube Discretionary Settlement Trust [2010] UKFTT 473 (TC) and HMRC did not suggest it was wrongly decided.  The Tribunal stated:

 

[93] ...In our view, the expression ‘person acting on ...behalf’ is not apt to describe a mere adviser who only provides advice to the taxpayer or to someone who is acting on the taxpayer’s behalf.  In our judgement the expression connotes a person who takes steps that the taxpayer himself could take, or would otherwise be responsible for taking.  Such steps will commonly include steps involving third parties, but will not necessarily do so.  Examples would in our view include completing a return, filing a return, entering into correspondence with HMRC, providing documents and information to HMRC and seeking external advice as to the legal and tax position of the taxpayer.  The person must represent, and not merely provide advice to, the taxpayer.

194.The appellant’s view was that while NTA was retained to provide advice on the Romangate scheme, only F&L was retained to prepare and submit Mr Atherton’s return and only F&L were therefore ‘acting on his behalf’ within the meaning of s 29(4).

195.We struggle with the views expressed in Bessie Taube.  Our view is that, unless expressly stated otherwise by Parliament, a person cannot pass on to someone else an obligation which Parliament has imposed on that person.  It is contrary to good governance and sense for a person, with a statutory obligation, to be able to avoid liability for its improper performance simply by having passed it on to someone else, who owes no obligation to the government to carry out that duty.

196.This view is reflected in the VAT penalties legislation at s 71(1)(b) of the Value added Tax Act 1994 where it is expressly stated that the fact of reliance on a third party, or any carelessness by the third party relied on, is not a reasonable excuse for failure to comply with an obligation by the taxpayer.  Our view would be that that would be the natural way to interpret ‘reasonable excuse’ and it did not really need an express statement to that effect.  Any other view would allow a person with the obligation to file a complete and correct tax return to escape the obligation to do so by passing it on to someone else:  that not only appears to subvert Parliament’s intention but favour those who can afford advisers over those who can’t.  Why should a taxpayer who makes a careless error have liability whereas a taxpayer who employed an agent who made the same careless error avoid liability?

197.Nevertheless, we are aware that recently Parliament have chosen to provide exceptions in other circumstances, most notably in Schedule 24 of Finance Act 2007.  That provides for penalties for certain inaccurate direct tax returns, and at paragraph 18(3) expressly provides that a taxpayer is not liable for mistakes by his agent if the taxpayer ‘took reasonable care to avoid’ the mistake.

198.This Tribunal has also interpreted other legislation in such a way that an agent’s unanticipated carelessness is not attributed to the taxpayer. Aside from Bessie Taube, in the Special Commissioners’ decision in AB (a firm) (2006) SpC 572 the panel said:

We are of the view that the question of whether a taxpayer has engaged in negligent conduct is a question of fact in each case.  We should take the words of the statute as we find them and not try to articulate principles which could restrict the application of the statutory words.  However, we accept that negligent conduct amounts to more than just being wrong or taking a different view from the Revenue.  We also accept that a taxpayer who takes proper and appropriate professional advice with a view to ensuring that his tax return is correct, and acts in accordance with that advice (if it is not obviously wrong), would not have engaged in negligent conduct.

We note in passing that neither party suggested that there was any relevant distinction between ‘negligence’ and ‘carelessness’ and for the purposes of this appeal we proceed as if they are interchangeable terms.

199.We were also referred to the case of Hanson [2012] UKFTT 314 (TC) where the judge at [21] said there was no liability if the taxpayer reasonably relied on a reputable accountant for advice on completion of his tax return, but as the case concerned Sch 24 it cannot be read across to s 29(4).  This was applied in Litman & Newall [2014] UKFTT 89 (TC) which was also a Sch 24 penalty case, where the conclusion was that, despite professional advice, the taxpayers were careless as they failed to consider the commercial reality of what they had done (or not done) to implement the scheme.  It was applied in Gedir [2016] UKFTT 188 (TC) where reasonable reliance on professional advice avoided a Sch 24 penalty.

200.However, these were all cases on Sch 24:  there is nothing in s 29(4) which expressly limits the normal interpretation of the law that where Parliament has imposed an obligation on a person, that person cannot avoid liability by passing it on to someone else.  In other words, while obligations can be out-sourced, the person with the obligation remains responsible for its execution.  His agent’s carelessness is his carelessness.  Indeed, the wording of s 29(4) seems intended to put beyond doubt that liability cannot be passed on by stating expressly that it applies to the carelessness of the taxpayer or anyone acting on his behalf.

201.We also struggle with Bessie Taube as taken to its logical conclusion, it suggests that the taxpayer is liable for the carelessness of an agent employed to complete his tax return, but not for the carelessness of an agent employed to advise him on how to complete his return.  The logic of such a distinction escapes us.

202.If it mattered in order to resolve this case, which it does not, we would consider s 29(4) should be read broadly to encompass all advisers to Mr Atherton, including those who, like NTA, gave general advice on completion of the tax return.

203.However, it would not extend to the opinion given by Mr Bretten QC.  It was clear Mr Bretten was not acting on behalf of Mr Atherton.  He gave his opinion to NTA on the basis of a hypothetical/unnamed taxpayer who had implemented a different scheme and at a time when Mr Atherton was not a client of NTA.

204.That concludes the half of the appeal which concerned s 29(4):  we have found, for the reasons given above, that the insufficiency in Mr Atherton’s 2007/08 tax return, namely the loss stated in Box 20, was brought about carelessly by Mr Atherton or by a person acting on his behalf.  We move on to consider the second half of the appeal, which is whether HMRC have proved that they discovered the insufficiency within the meaning of s 29(1).

 The law on discovery

205.As we have said, the parties were agreed that the appellant’s liability to pay the assessment depended entirely on whether HMRC had made an assessment within the parameters of s 29 TMA. We have dealt with condition s 29(4) on carelessness, but HMRC also had to prove that there had been a discovery.  S29 provided as follows:

(1) If an officer of the Board or the board discover, as regards any person (the taxpayer) and a year of assessment

(a) That any income which ought to have been assessed to income tax...have not been assessed, or

(b) That an assessment to tax is or has become insufficient....

The officer or, as the case may be, the board may, subject to subsections (2) and (3) below, make an assessment in the amount, or the further amount, which ought in his or their opinion to be charged in order to make good to the Crown the loss of tax.

What is a discovery?

206.As case law has shown, it is not entirely straightforward to define what a discovery within 29(1) actually is.  What we think case-law has shown is required is as follows:

(a)    An HMRC officer crosses a threshold from non-awareness to awareness of an insufficiency;

(b)   He acted reasonably when so doing;

(c)    HMRC as a body did not previously have the awareness of the insufficiency;

(d)   The assessment must be proximate to the discovery.

207.We deal with the authority for each of these propositions in turn:

208.(a) crossing a threshold:  The Upper Tribunal in Charlton [2012] UKUT 770 (TC) said:

[28] ...the word ‘discovers’ does connote a change, in the sense of a threshold being crossed.  At one point an officer is not of the view that there is an insufficiency such that an assessment ought to be raised, and at another he is of that view.  That is the only threshold that has to be crossed.  We do not agree that the lawyer, in Lord Denning’s example, would be regarded as having made a discovery any the less by waking up one morning with a different conclusion from the one had had earlier reached, than if he had changed his mind with the benefit of further research.  It is, we think, evident that the relevant threshold for there to be a discovery may be crossed as a result of a ‘eureka’ moment just as much as by painstaking research.

209.From this it is clear that there is a fairly low threshold for an officer to make a discovery.  It can be a change of mind.  It could include, for instance, a change of mind following on from a release of a judgment.  A discovery does not require new factual information to have newly come to HMRC’s knowledge.

210.(b) the officer must be acting reasonably:  s 29(1) permits an assessment where otherwise none would be permitted.  It must be implicit in s 29(1) therefore, that when crossing the threshold to awareness, the officer is acting reasonably in the Wednesbury sense.  Otherwise, it would be open to HMRC to make discovery assessments on an unreasonable awareness.

211.In Charlton,  the Upper Tribunal said:

[24] ... ‘discovers’ cannot mean to ascertain by legal evidence.  But is nevertheless the case that an officer’s discovery must be a reasonable conclusion from the evidence available to him.  To that extent, although the test in s 29(1) is a subjective test, an element of objectivity is introduced in examining the reasonableness of the officer’s conclusion....

212. (c) the discovery must be new to HMRC.  Again it must be inherent in the word ‘discovery’ that the discovery is new to HMRC.  Otherwise, by passing the file on to a new HMRC officer, that officer could always ‘discover’ an insufficiency that he was previously unaware of.  There was some discussion of this in Charlton  at [41-42] and in particular whether involving a new officer in an old case would enable the new officer to make a ‘discovery’ of something already known to the original officers dealing with the file.  The Upper Tribunal did not appear to conclusively determine the point, and indeed it is not easy to determine exactly what they meant, but the Upper Tribunal did seem to indicate that, as a discovery must be new, the newness is more than just newness to the officer making the assessment.  We think that must be right.

213.The point is not particularly relevant here where Mr Clarke was involved with the case from the start; there is no suggestion that he was not aware from early 2009 that Mr Atherton had implemented Romangate. 

214.Mr Clarke described his discovery in a letter dated 5 June 2014 as follows:

The conclusion I reached in this case is that the assessment to tax on your 07/08 tax return is insufficient.  I reached the conclusion when I became aware that the previous amendment to your 07/08 tax return was potentially not valid.

We find, taking all the evidence into account, that what was new to Mr Clarke in early 2014 was the implication of the Cotter decision that because Mr Atherton had made his claim in Box 20, he must be treated as having made the claim ‘in’ his self-assessment, and therefore the Sch 1A enquiry opened to defeat the claim and/or the restatement of the self-assessment at a figure of just over £2 million (§84) was ineffective to do so. 

215.It was Ms Balmer’s case that Mr Clarke’s team in 2009 had so many returns from taxpayers who had implemented Romangate to look at, they did not look at any closely and did not look at them to ascertain whether the claim was ‘in’ the self-assessment until after Cotter.  We do not accept that that case is made out on the facts.   It is clear that Mr Clarke did look closely at Mr Atherton’s return in 2009 and was aware of the entry in box 20.  He was aware (see §§82-85) in 2009 that the claim was made in a box which fed through to the self-assessment:  what he thought that meant as a matter of law is something we move on to discuss below.

(d) the discovery must be proximate to the assessment:  The Upper Tribunal in Charlton also said, obiter or in passing, that the assessment must follow on the heels of the discovery with some alacrity:

[37]...all that is required is that it has newly appeared to an officer, acting honestly and reasonably, that there is an insufficiency in an assessment....The requirement for newness does not relate to the reason for the conclusion reached by the officer, but to the conclusion itself.  If an officer has concluded that a discovery assessment should be issued, but for some reason the assessment is not made within a reasonable period after that conclusion is reached, it might, depending on the circumstances, be the case that the conclusion would lose its essential newness by the time of the actual assessment.  But that would  not, in our view, include a case, such as this, where the delay was merely to accommodate the final determination of another appeal which was material to the liability question.  Such a delay did not deprive [the discovery] of their essential newness for s 29(1) purposes.

216.While it is inherent in the word ‘discovery’ that the discovery must be of something new, there is nothing overt in s 29 which requires the assessment to be proximate to the discovery:  this obiter comment in Charlton was therefore criticised in three FTT decisions:  Pepper [2015] UKFTT 615 (TC), Gakhal [2016] UKFTT 356 (TC) and Miesegaes [2016] UKFTT 375 (TC).

217.Nevertheless, it was followed by the Upper Tribunal in Pattullo [2016] UKUT 270 (TC) at [52], released on 14 June 2016 and what was said in Pattullo is binding on this Tribunal as it formed a part of the operative decision.  So while in the May 2016 hearings of this appeal, Ms Balmer sought to persuade us Charlton was wrong on this point, by the July and September hearings she accepted we were bound by Pattullo.  We understand that HMRC reserve the right to challenge this interpretation of s 29 if this decision is appealed.

218.For this hearing, it was HMRC’s case that the discovery occurred no earlier than the Supreme Court decision in Cotter which was released late in 2013.  Mr Gordon accepted that if the date of the discovery was no earlier than the Supreme Court decision in Cotter on 6 November 2013, then the assessment on 31 March 2014 (§23) was sufficiently proximate to the discovery so that the discovery had not lost its essential newness.  He did not accept, of course, that there was a discovery after the Cotter decision.

What must be discovered?

219.None of the above propositions were really in issue in this appeal. What was really in issue was what must be discovered.  The appellant considered the point was simply that HMRC knew in 2009 that the appellant had implemented Romangate, and from the date of the retrospective legislation in 2009, knew that the appellant was not entitled to the loss generated by Romangate claimed on his tax return.  Mr Clarke could not, therefore, newly discover this in early 2014.

220.HMRC say that they discovered the assessment was insufficient when they received the Supreme Court decision in Cotter. They accept that they were aware of the ‘tax loss’ since 2009 but Ms Balmer said her case was that they newly discovered in 2014 an insufficiency of the collection of that tax. At root, it seemed to be Ms Balmer’s position that it was enough for Mr Clarke to discover that the steps HMRC had taken to correct the insufficiency were themselves insufficient.

221.Mr Gordon appeared to agree that that was what HMRC had done but it was his view that did not meet the test in s 29(1).  His view was that what HMRC discovered in early 2014 was not that the assessment was insufficient, which, he says, they had known since 2009, but that the steps they had taken to correct the insufficiency were ineffective.  They had opened a Sch 1A enquiry whereas the Cotter decision in late 2013 showed that to challenge the claim to the relief they should have opened a s 9A enquiry.  It was his view, as he said in submissions, that the discovery was ‘whoops, we opened the wrong kind of enquiry’.

222.But we do not think that the point is that simple.  It is clear from s 29(1) (set out at §25 above), in so far as relevant here, that the discovery must be that an assessment is (or has become) insufficient (29(1)(b)) or that income which ought to have been assessed has not been assessed (29(1)(a)) or that relief was claimed which should not have been  (29(1)(c)).  This can be summarised, as we have said, as a discovery that an assessment is insufficient. 

Did HMRC newly discover in 2013 that Mr Atherton’s assessment was insufficient?

223.The flaw in the appellant’s case is that it equates the discovery that the loss relief claim was invalid with a discovery that the self-assessment was insufficient.  But there is a distinction, as the Supreme Court made clear, between a tax return form and the self-assessment.  The tax return form can include standalone claims that are not a part of the tax return itself and which do not form a part of the self-assessment. 

224.So the question is not when HMRC discovered that the appellant had made an invalid loss relief claim on his tax return form, but when they discovered that the tax return itself, the self-assessment part of the tax return form, included an invalid loss relief claim.

225.We find the loss relief claim was quite clearly made in the self-assessment part of the tax return form.  This was because Box 20 was completed and the claim fed through to the self-assessment, reducing it to nil.  Nevertheless, we find that it is also quite clear that HMRC proceeded on the assumption from the start in 2009 that the tax return form should be treated as if the Box 20 entry had not been made, and must be treated as if a standalone claim had been made (which it had been as Box 3 was also completed).  We considered the evidence set out at §§82-91 and concluded that it shows that HMRC were at the time of the consistent view that Mr Atherton had only made a standalone claim even though box 20 was completed.

226.This explains why HMRC opened a Sch 1A enquiry but not a s 9A enquiry; it explains why they sought to re-calculate the self-assessed tax to show the £2 million liability without using the normal TMA procedures (s9ZB) because they were convinced that in law the taxpayer had to be treated as making a standalone claim.  It explains why they proceeded to attempt to collect the £2 million from Mr Atherton.  All these actions were consistent with the belief that Mr Atherton had made a standalone claim to the loss and inconsistent with the belief that Mr Atherton had claimed the loss in his self-assessment.

227.Mr Gordon suggested that this analysis of the facts is wrong because HMRC’s attempt to re-calculate the self-assessment (see §84) shows that HMRC were  aware the claim was made ‘in’ Mr Atherton’s self-assessment. We don’t agree.  Section 9ZB allows HMRC to correct obvious errors in self-assessments; if HMRC had thought that the loss relief claim was ‘in’ the tax return but should not have been, they would have made a correction under s 9ZB.  We find it is clear that they did not seek to use this power because they did not consider the claim was made ‘in’ the tax return proper.

228.In other words, from the first, we find Mr Clarke was of the view that as a matter of law, Mr Atherton’s self assessment did not include the loss relief claim.  It is clear from his letters and actions that he held this view despite the completion of box 20 and the fact that the self-assessment calculation incorporated the claim.

229.The evidence clearly shows that HMRC only realised that the self-assessment was insufficient following the decision in Cotter when it was made plain that a claim for carried back losses which was included in a self-assessment was not a standalone claim, but simply an invalid claim leading to an  insufficient self-assessment.

230.This is quite an unusual case:  we are not aware of any other case raising a similar issue.  We have therefore considered the point at length.  It is clear that HMRC were from 2009 cognisant of all the facts necessary to conclude that the 07/08 self- assessment was insufficient and they discovered no new relevant facts after 2009:  but their understanding of the law up until Cotter led them to conclude that the self-assessment in law was, or had to be treated as being,  for £2 million  and therefore was not insufficient despite in practice the self-assessment on its face including the invalid loss relief claim, and being expressed to be nil.  Mr Clarke discovered in early 2014 that his view of the law was wrong and that the self-assessment in law was for nil and was therefore insufficient.

231.In a typical discovery case, HMRC would know what comprised the tax return, by which we mean those parts of the tax return form which are the tax return proper and feed into the self-assessment.  They would discover that an entry on that return was insufficient.

232.Here the position is different.  In 2009, HMRC thought the tax return did not include certain entries on the form, in particular the entries in box 3 and 20.  They thought then, and continue to be of the opinion, that those entries claim a relief to which the appellant is not entitled.  But for the period 2009-2013, they did not think that the return was insufficient, or that it included an excessive relief claim, because they did not think that the tax return proper included those entries in box 3 and box 20.

233.It turns out that HMRC in general and Mr Clarke in particular did not know what comprised the tax return proper.  Mr Clarke discovered, after Cotter, that the tax return proper included the entry in box 20 (but not the entry in box 3). It was only when in early 2014 that Mr Clarke was advised that the entry in box 20 had to be treated as part of the tax return proper, that he became aware that the tax return was insufficient as it actually contained the relief claim which he had long ago concluded was excessive. So it is the case that HMRC discovered in early 2014 an insufficiency in Mr Atherton’s tax return.  They newly discovered, in the words of s 29(1), that income which ought to have been assessed was not assessed, that the assessment was insufficient, and that a relief claim which had been given in the self-assessment was excessive.

234.In the hearing, rather loosely, HMRC were referred to as having known of the insufficiency since 2009, but that only goes to show that loose language is misleading.  HMRC did know from 2009 that the claim to the employment loss relief was invalid; but because Mr Clarke only discovered that that claim comprised a part of the self-assessment in 2014, the insufficiency in the self-assessment was only discovered in 2014.

235.We go on to consider whether that discovery meets the test as set out above.  But we note that we entirely reject Ms Balmer’s case that it is enough to discover that HMRC’s actions have been insufficient to correct an insufficiency in a self-assessment.  That is inconsistent with the plain words of s 29(1).

236. (a) did Mr Clarke cross a threshold in early 2014?

237.We find that he did.  Prior to the advice from policy referred to at §91, which was based on the Supreme Court decision in Cotter  a few months earlier, Mr Clarke had consistently been of the view that the box 20 entry was not a part of Mr Atherton’s self-assessment and that the self-assessment was for just over £2million and not insufficient. After receiving that advice, he completely changed his mind and discovered that the box 20 entry was a part of the self-assessment and that therefore the self assessment was for nil and was insufficient.

238.Mr Gordon relied on Mr Clarke’s description of his discovery as being based on the finding out that HMRC had failed to correct the insufficiency by amending the return. As we have already said, his evidence as a whole was that his discovery was that the self-assessment was insufficient because of Cotter and that therefore the Sch 1A enquiry was ineffective to challenge it as was the amendment to the tax return.  This seems accurate to us:  it was because he discovered that the box 20 entry comprised a part of the self-assessment that he first realised the self-assessment was for nil, and that HMRC had failed to amend it to £2 million, it remained at nil and was insufficient.

239.(b) did Mr Clarke act reasonably?

240.It was clearly reasonable for him to form a view consistent with that of the Supreme Court in Cotter.  But was it reasonable for him to have consistently held the opposite view up to that point that the self-assessment did not include the box 20 entry when (a) box 20 did in practice feed into the self-assessment; (b) he was aware that it was the taxpayer’s and his advisers’ view that box 20 did feed into the self-assessment; (c) the tax return showed a nil and not £2 million liability?

241.In retrospect, the Supreme Court’s view, that a taxpayer who makes an invalid claim in a return is nevertheless to be treated as making that claim, is entirely predictable:  it is a self-assessed tax so it follows that taxpayers can complete their tax returns incorrectly.

242.But we do not think that Mr Clarke’s view to the contrary in so far as carry back loss relief claims was concerned was an unreasonable view.  It was not shown to be inconsistent with any court decisions of the time, and was consistent with HMRC’s view at the time.

243.He ought perhaps to have been aware that some tax specialists considered that HMRC was wrong on this point:  that ought to have been apparent from the Cotter litigation which was ongoing for years before it reached the Supreme Court.  But we do not think it is unreasonable for HMRC to have considered that those views were wrong and indeed HMRC were prepared to litigate to uphold their view.  Mr Clarke’s and HMRC’s view on this, although ultimately shown to be wrong in law, was not an unreasonable view to hold up until the point of the Supreme Court decision.

244.(c) was the discovery new to HMRC?

245.We find it was.  It follows from what we said in the previous paragraph that Mr Clarke and HMRC considered, before November 2013, that Year 2 loss relief claims had to be made as standalone claims and had to be treated as such even if entered into the Year 1 self-assessment.  Cotter in November 2013 newly showed that that view was wrong and Mr Clarke reversed his opinion in early 2014.

246.At one point, Mr Gordon suggested that the S 9A enquiry into Mr Atherton’s tax return (§88) was opened because HMRC realised that they might be in difficulties with the Sch 1A enquiry.  But as a matter of fact, we do not accept that.  We found, as we have said, that Mr Clarke did not instigate the opening of that enquiry and that it was opened for reasons unconnected with the Romangate enquiry.  The opening of that enquiry did not indicate that Mr Clarke or anyone else in HMRC had formed the view that the box 20 entry was a part of Mr Atherton’s self-assessment.

247.Although nothing was made of it in the hearing, we note that it appears HMRC and the appellant agreed to a stay of the County Court proceedings.  It seems an obvious conclusion that HMRC recognised that the appellant thought he had put his claim into his self-assessment.  Indeed, we would say that was obvious from the manner in which the tax return was completed.  Nevertheless, we remain of the view that HMRC were entitled to take, and reasonably took, the position that the claim was not in the self-assessment.  We also find that although HMRC as a whole must have been aware that their view was disputed, they only newly discovered that their view of the law was actually wrong after the Supreme Court in Cotter.  So at that point Mr Clarke newly discovered, on changing his mind on the law, that the self-assessment was insufficient because it did contain the loss relief claim.

248.The fact that HMRC ought to have known before that date that others disagreed with their view, does not mean that Mr Clarke made no discovery following the Cotter  decision:  it is one thing to know your view is challenged and quite another to know it is wrong. The discovery in late 2013/early 2014 was new to HMRC and new to Mr Clarke.

249.(d) Was the discovery proximate to the assessment?

250.The discovery was made in early 2014 by Mr Clarke when he received advice from policy:  Mr Gordon accepted, as we have said at §218 and as we find, that in such a case it was not a stale assessment.

Conclusion

251.We consider that because of the entry in box 20 there was an insufficiency of the amount assessed in Mr Atherton’s 2007/08 tax return; that insufficiency was brought about by the carelessness of Mr Atherton and F&L in inserting, without explanation, into box 20 an employment loss for 2008/09 although box 20 related to partnership losses for 2007/08, and that claim duplicated the claim in box 3.  That insufficiency was discovered by HMRC when, following the Cotter decision in late 2013 they discovered that the box 20 entry comprised a part of the taxpayer’s self-assessment and that it had effectively reduced the self-assessment to nil, and that Mr Atherton’s self-assessment for 2007/08 was therefore insufficient.

252.We consider that the discovery assessment was procedurally properly made and must be upheld.  As the taxpayer does not dispute his underlying liability to the assessment, the appeal is dismissed.

 

253.This document contains full findings of fact and reasons for the decision. Any party dissatisfied with this decision has a right to apply for permission to appeal against it pursuant to Rule 39 of the Tribunal Procedure (First-tier Tribunal) (Tax Chamber) Rules 2009.   The application must be received by this Tribunal not later than 56 days after this decision is sent to that party.  The parties are referred to “Guidance to accompany a Decision from the First-tier Tribunal (Tax Chamber)” which accompanies and forms part of this decision notice.

 

 

BARBARA MOSEDALE

TRIBUNAL JUDGE

 

RELEASE DATE:  16 DECEMBER 2016