First-tier Tribunal (Tax)

You are here: BAILII >> Databases >> First-tier Tribunal (Tax) >> Cooke v Revenue & Customs (INCOME TAX/CORPORATION TAX : Assessment/self-assessment) [2017] UKFTT 844 (TC) (24 November 2017)
URL: http://www.bailii.org/uk/cases/UKFTT/TC/2017/TC06239.html
Cite as: [2018] SFTD 633, [2018] STI 454, [2017] UKFTT 844 (TC)

[New search] [Printable PDF version] [Help]

[2017] UKFTT 844 (TC)

TC06239

Appeal number: TC/2016/02608

INCOME TAX – discovery assessment – excessive claim to double tax relief – error could be determined from information in return – whether HMRC officer could not have reasonably be expected to be aware so that s 29(5) satisfied – whether appellant’s representative careless so that s 29(4) satisfied – appeal allowed

FIRST-TIER TRIBUNAL

TAX CHAMBER

	ADAM COOKE	Appellant

	- and -

	THE COMMISSIONERS FOR HER MAJESTY’S	Respondents
	REVENUE & CUSTOMS

TRIBUNAL:	JUDGE SARAH FALK

Sitting in public at Taylor House, 88 Rosebery Avenue, London EC1R 4QU on 25 October 2017

Robert Maas, instructed by Kreston Reeves LLP, for the Appellant

Kate Murphy, Officer of HMRC, for the Respondents

DECISION

1. This is an appeal against a discovery assessment in respect of the year 2012-13, made under s 29 Taxes Management Act 1970 (“TMA”) in the amount of £16,560. The appellant does not challenge the quantum of the tax payable. The only dispute is over the validity of the assessment.

Background

2. The appellant’s 2012-13 return was filed electronically on 22 January 2014. Under s 9A(2)(a) TMA HMRC had until 22 January 2015 to open an enquiry. No enquiry was opened within the time limit. The appellant’s 2013-14 return was filed electronically on 17 November 2014. An enquiry into that return was opened by a Mr Neaves of HMRC on 24 September 2015.

3. Both returns included entries in the “Foreign” pages in respect of foreign dividends from three jurisdictions, two of which were France and Canada. The following information was included in respect of those dividends on page F2, under the headings shown below (the country code is that provided by HMRC’s systems):

A. Country or territory code	B. Amount of income arising or received before any tax taken off	C. Foreign tax taken off or paid
2012-13
FRA	103,083	30,523
CAN	15,636	3,845
2013-14
FRA	59,250	17,603
CAN	21,107	5,077

4. The reason why HMRC opened an enquiry into the 2013-14 return was that by that time HMRC had decided to look at all self assessment returns where foreign tax credit relief (double tax relief) had been claimed. When the 2013-14 return was looked at HMRC found that the relief claimed in respect of both the French and Canadian dividends was in excess of the permitted amount under the terms of the relevant double tax treaty, being 15%. (On the figures the credit actually claimed was around 30% for the French dividends and 25% for the Canadian dividends.)

5. HMRC’s enquiry into the 2013-14 return also led them to check the 2012-13 return, where the same error was detected.

6. A discovery assessment for 2012-13 was issued on 19 January 2016 (within the four year time limit set by s 34 TMA). HMRC issued a closure notice in respect of the 2013-14 return on the same date, which has not been appealed. An appeal was made against the discovery assessment on 10 February 2016. Following a review HMRC’s conclusion was confirmed by a letter dated 11 April 2016. The appeal to the Tribunal was made on 10 May 2016.

The legislative framework

Section 29 TMA

7. Section 29 TMA provides so far as relevant:

“(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, or chargeable gains which ought to have been assessed to capital gains tax, have not been assessed, or

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

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.

…

(3) Where the taxpayer has made and delivered a return under section 8 or 8A of this Act in respect of the relevant year of assessment, he shall not be assessed under subsection (1) above—

(a) in respect of the year of assessment mentioned in that subsection; and

(b) in the same capacity as that in which he made and delivered the return,

unless one of the two conditions mentioned below is fulfilled.

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

(5) The second condition is that at the time when an officer of the Board—

(a) ceased to be entitled to give notice of his intention to enquire into the taxpayer's return under section 8 or 8A of this Act in respect of the relevant year of assessment; or

(b) informed the taxpayer that he had completed his enquiries into that return,

the officer could not have been reasonably expected, on the basis of the information made available to him before that time, to be aware of the situation mentioned in subsection (1) above.

(6) For the purposes of subsection (5) above, information is made available to an officer of the Board if—

(a) it is contained in the taxpayer's return under section 8 or 8A of this Act in respect of the relevant year of assessment (the return), or in any accounts, statements or documents accompanying the return;

…

(8) An objection to the making of an assessment under this section on the ground that neither of the two conditions mentioned above is fulfilled shall not be made otherwise than on an appeal against the assessment.

…”

8. It was not disputed that the burden of proof lies on HMRC to demonstrate that the conditions required to be satisfied to allow a discovery assessment to be made are met (see for example HMRC v Household Agents [2008] STC 2045 (Ch), approved in Hankinson v HMRC [2012] STC 485 (CA)).

9. There was also no dispute that the requirements of section 29(1) were met in this case, such that subject to one of the conditions in s 29(4) or (5) being satisfied a discovery assessment could be made. HMRC clearly “discovered” that there was an error in the double tax relief claim. As has been made clear on a number of occasions, and in particular by the Court of Appeal in Hankinson and by the Upper Tribunal Charlton & others v HMRC [2013] STC 866, the question of discovery is concerned with the inspector’s subjective view and does not require any new facts to emerge. It is enough that the inspector satisfies himself or comes to a conclusion.

10. The dispute between the parties relates to whether either or both of the conditions in s 29(4) or (5) were met. HMRC’s position was that the condition in s 29(4) was met because the appellant’s accountant had been careless, and that s 29(5) also applied because there was insufficient information on the return for an officer of HMRC to know from looking at it that the foreign tax credit figures were excessive.

Double tax relief

11. The double tax treaties with France and Canada each place limits on the tax that can be imposed on dividends paid to residents of the other contracting state. In the appellant’s case the tax is limited to 15%, under Article 11 of the UK/France treaty and Article 10 of the UK/Canada treaty. This is a fairly standard rate for dividends paid to “portfolio” investors (typically those with a shareholding of below 10%). The rate of withholding tax under domestic law is frequently higher. Where tax is deducted at a rate in excess of the rate permitted under the treaty, the taxpayer can seek to reclaim the tax from the tax authority in the relevant foreign jurisdiction. However, time limits will typically apply.

12. Under UK law, double tax relief may only be claimed at the treaty rate of 15%. This is the effect of the incorporation of the treaties into UK law pursuant to what is now Chapter 1 of Part 2 of the Taxation (International and Other Provisions) Act 2010 (“TIOPA”), and the rules governing the amount of credit available in Chapter 2 of Part 2 of TIOPA.

13. With that background I now turn to the evidence and facts.

Evidence

14. Oral evidence, and a short witness statement, was provided by Nathan Steinberg. At the relevant time Mr Steinberg was a partner in Munslows LLP, a firm of chartered accountants based in London. The firm has since merged with a much larger firm, Kreston Reeves LLP, and Mr Steinberg works as a consultant in that firm. The appellant was a client of Mr Steinberg’s at Munslows LLP and remains a client of the merged firm. It was Munslows LLP that prepared the relevant tax returns. Mr Steinberg’s evidence was straightforward and helpful, and I accept it as regards matters of fact.

15. Documentary evidence included copies of the 2012-13 and 2013-14 returns and correspondence between the parties.

Findings of fact

16. Munslows was a small firm of chartered accountants with a general practice. Unlike Kreston Reeves it had no separate tax department. In addition to accounting work Munslows prepared tax returns for clients who instructed them to do so. Partners such as Mr Steinberg would not process tax returns themselves. Instead Mr Steinberg relied on members of his team, comprised of clerical staff and employed accountants, to do this work, but he would check the draft returns before they were sent to clients for approval. The staff who produced the returns were not tax specialists and, whilst they would work on tax returns at relevant times of the year, they would otherwise be employed on other work.

17. The firm used standard commercially available software to prepare returns. It had access to two different products but the one used for the appellant’s returns was a product known as “PerTAX”. I understood from Mr Steinberg that this particular product is widely used. The software enabled the returns, including the foreign pages, to be completed, but there was nothing in the software either to prevent excessive double tax relief being claimed or to flag up to the user that there was a potential issue, or that they needed to check what relief was actually available.

18. Mr Steinberg described the sort of checks he would normally carry out before sending returns to clients for approval. His focus tended to be on the tax calculation pages, checking that nothing had been omitted and that the figures made sense. I understood this to be largely a “sanity check” for obvious errors and omissions, rather than a detailed analysis. At the time neither Mr Steinberg nor other staff were aware that the PerTAX software did not flag up any limits on double tax relief claims or prevent excessive claims being made, and so the errors were not picked up. Clearly, he was now aware and new systems have been put in place. His understanding was that Munslows was not alone in making this kind of error. The merged firm, which had the benefit of a specialist tax department, had also discovered that similar errors had been made with some other clients. Since discovering the problem with the appellant’s return the firm had attempted to recover the excess tax from the French and Canadian tax authorities, and the appeal was only being pursued because the attempt had been unsuccessful.

19. In cross examination Mr Steinberg confirmed that he was aware of double tax relief in general terms but not of the particular rates of relief available, that he had reviewed the appellant’s 2012-13 return, and that he would have focused on the client schedules with the tax calculation in more detail than other parts of the return such as page F2.

20. Following the enquiry being opened into the appellant’s 2013-14 return, Mr Steinberg spoke to HMRC by phone and accepted that the claims should have been restricted to 15%. He also explained to HMRC that this was his firm’s error and not that of the appellant, and on that basis there should be no penalty. In a further call he asked how the firm should go about reclaiming the foreign tax, and was referred to HMRC’s Double Taxation Manual. He wrote to Mr Neaves on 26 October 2015 confirming his firm’s agreement that the claims should have been restricted and apologising for the error which was “due to the fact that our computer system did not pick up the fact that the amount of tax deducted was above the amount that could be reclaimed”. The letter stated that claims would be made to recover the over deducted tax under the treaties and thanked Mr Neaves for forwarding website links to the Manuals to assist Munslows in making the claims.

21. Mr Neaves wrote to Munslows on 12 November 2015. This letter confirmed the revised tax calculation for 2013-14 and explained that given the error in that return Mr Neaves had now reviewed the appellant’s 2012-13 return, and had found that the Canadian and French dividend relief claims in that return were also in excess of 15%. The letter referred to s 29 TMA and stated:

“I do not require you to provide any information as it is all contained within the foreign pages of the return.”

The letter enclosed a recalculation of the tax liabilities for both years and stated that formal assessments would follow.

22. Mr Steinberg replied on 26 November contending that HMRC was not permitted to issue a discovery assessment because the condition in s 29(5) was not met. Mr Neaves responded on 17 December disagreeing on the basis that there was no evidence in the return to show that the foreign tax credit relief claim had or had not been made in line with the relevant double taxation agreements, and that HMRC could not know that the claim was excessive due to Munslows’ accounting system failing to pick the point up. Mr Neaves referred to the guidance notes to the Foreign pages, stating that they made it clear that taxpayers should check to see if double taxation agreements are in place and adhere to them. HMRC had only discovered the problem once the 2013-14 return was reviewed. Mr Neaves added:

“It was from finding the error within that claim it became apparent there was a similar error contained within the 2012-13 tax return.”

23. Mr Steinberg’s reply to this on 21 January 2016 (which must have crossed in the post with the discovery assessment), and the formal appeal filed on the appellant’s behalf, both repeated that HMRC was not entitled to make a discovery assessment because the error could have been picked up by HMRC just as easily as it had been for 2013-14. HMRC had simply chosen not to review the earlier return.

24. HMRC’s review decision concluded that, although Mr Neaves had not considered s 29(4), that condition as well as the condition in s 29(5) was met.

Submissions

25. Ms Murphy, for HMRC, submitted that s 29(4) was satisfied because the appellant’s accountant had been careless. The accountants had acknowledged that the return was incorrect because the computer programme did not alert them to the error and they did not check that the amounts claimed were in line with the relevant treaties. There was information on what was then the HMRC website (now transferred to gov.uk), including HMRC’s Manuals, that would have explained the level of relief that could be claimed for each country. It was reasonable to expect a professional firm, tasked to produce tax returns, to check this. Failure to do so amounted to an absence of reasonable care. Computers do make mistakes and for that reason a checking system should be in place.

26. In addition, in HMRC’s view, s 29(5) applied. There was insufficient information in the 2012-13 return for an officer of HMRC to know from looking at it that the claim to foreign tax credit was incorrect. HMRC work on the principle of process first, check later. The assumption is that tax returns are completed accurately and correctly. There was nothing in the 2012-13 return to indicate that the percentages used to arrive at the foreign tax credit claims were incorrect or to draw the attention of an officer to the fact that the tax credit claims were incorrect. The returns for both 2012-12 and 2013-14 were submitted electronically and were not manually processed by an HMRC officer, meaning that the matter did not come to light until the return for the later year was selected for enquiry. No information was entered in the “white spaces” to alert HMRC to the fact that the amount of relief claimed might have been excessive. The appellant’s agent would have been aware that those white spaces were available as a means to alert HMRC to concerns or queries, or the agent could have contacted HMRC by letter or email. The hypothetical officer postulated by s 29(5) was not a specialist, and specialist knowledge would be needed to be aware of the rates provided under particular treaties.

27. Mr Maas, for the appellant, submitted that s 29 TMA provided a narrow exception to the principle of finality in self assessment. It limited discovery to two situations, essentially brought about by the taxpayer. What HMRC actually did or did not do in processing returns was not relevant. Page F2 of the return clearly had sufficient information for HMRC to be able to ascertain that excessive claims had been made, and it was clear from the correspondence that this was the case since Mr Neaves stated in his letter of 12 November 2015 that he did not require further information. As regards the question of carelessness, this was a small firm of general accountants. Most routine work of this nature is computerised and everyone relies on computer programmes, assuming that they work correctly unless a problem is flagged up. HMRC’s Compliance Handbook states at paragraph 81130 (in the context of penalties) that a person should be treated as taking reasonable care if arrangements or systems exist that, if followed, could reasonably be expected to produce an accurate basis for the calculation of tax.

Discussion

28. This case appears to raise something of a conundrum. Boiled down to its essentials, and leaving to one side some of HMRC’s arguments about their internal processes, HMRC are arguing that the appellant’s accountant was careless in not identifying that the claims were excessive, but at the same time an HMRC officer could not have been expected to pick the point up. The appellant argues the opposite: the accountant was not careless but an HMRC officer should have been able to spot the problem.

29. I will address s 29(5) first, and then s 29(4).

Section 29(5)

30. In my view HMRC’s internal processes are irrelevant. The fact that HMRC work on a principle of process first, check later, initially processing returns electronically and only selecting certain returns for enquiry, is something that they are clearly entitled to do but makes no difference to the interpretation of s 29(5). On the facts of this case s 29(5) requires a determination of whether the information available in the return was such that an HMRC officer “could not have been reasonably expected” to be aware of the insufficiency of the assessment. It assumes that a person, in the form of an HMRC officer, does look at the return, and the question it requires to be addressed is what the officer could reasonably have been expected to identify from the information in it. It may well be that s 29(5) does not fit as well as HMRC might like with their current working practices, and in particular the electronic processing of returns, but that is no basis for interpreting the provision in a way that does not accord with what it actually says. It is of course true that the issue would not have arisen if HMRC’s own software had been programmed to pick up that the claims were excessive, but that does not affect how the legislation applies: taxpayers are entitled to be taxed according to the law, rather than in a way that is affected by the sophistication or otherwise of HMRC’s systems or processes.

31. I therefore reject HMRC’s submissions insofar as they relate to their own processes, and turn to how s 29(5) should be applied in this case.

32. The interpretation of s 29(5) has been the subject of much case law discussion, including in the Court of Appeal. The most recent Court of Appeal decision that considers s 29(5) in any detail is Sanderson v HMRC [2016] STC 638. Patten LJ, who gave the only judgment, summarised points that he considered to be clear as a matter of authority at [17]. These included that the officer referred to is a hypothetical officer (rather than the officer who made the assessment), that the officer “has the characteristics of an officer of general competence, knowledge or skill which include a reasonable knowledge and understanding of the law” (referring to HMRC v Lansdowne Partners LLP [2011] EWCA Civ 1578, [2012] STC 544), and that where the law is complex even adequate disclosure may not make it reasonable for the officer to have discovered the insufficiency (referring to Lansdowne at [69]). He went on to emphasise at [22] that the hypothetical officer “must, on an objective analysis, be made aware of an actual insufficiency in the assessment”, and said at [23] that the officer was not required to resolve or assess every question of law, but where the points were not complex or difficult “he was required to apply his knowledge of the law to the facts disclosed and to form a view as to whether an insufficiency existed”.

33. The Court of Appeal decision in Lansdowne considered whether the hypothetical officer would have been aware of an insufficiency in circumstances where he could have seen from the documents available that deductions had been made from partnership profits for amounts paid to partners, contrary to the House of Law decision in MacKinlay v Arthur Young [1989] STC 898. The Chancellor stated at paragraph [50] of Lansdowne that the hypothetical officer would have known from the Arthur Young case and the relevant statute that payments to partners are not usually deductible, and concluded at [56] that the condition in s 29(5) was not met. Moses LJ agreed, stating at [69] that “legal points were not complex or difficult”. He did however leave open the possibility that, even where there was a sufficient disclosure of factual information, there might be circumstances in which an officer could not reasonably be expected to be aware of an insufficiency by reason of the complexity of the relevant law.

34. The Upper Tribunal decision in Charlton (referred to at paragraph [9] above) was made after, and considered, the Court of Appeal decision in Lansdowne, but it predated the Court of Appeal decision in Sanderson. Charlton included an extensive discussion of the concept of hypothetical officer in s 29(5) that is more nuanced, and potentially differs to some extent from, the summary referred to above in Sanderson (the point was not central to the decision in Sanderson). The discussion in Charlton was in turn commented on in another Upper Tribunal decision, Pattullo v HMRC [2016] STC 2043.

35. In Charlton the Upper Tribunal stated at [58] that there was no single benchmark for the knowledge and experience of the hypothetical officer and that the test of reasonable awareness had to be applied in the circumstances of each case, in the context of the particular information provided by the taxpayer. There was no assumed average to be applied in all cases. The focus is on the quality and extent of the information made available rather than the qualities of the hypothetical officer (paragraph [60]), and on the level of knowledge that it is reasonable for the hypothetical officer to have in the particular circumstances of the case (paragraph [62]). The Upper Tribunal summarised the position at [65] as follows:

“[65] Our conclusion on this point, therefore, is that s 29(5) does not require the hypothetical officer to be given the characteristics of an officer of general competence, knowledge or skill only. The officer must be assumed to have such level of knowledge and understanding that would reasonably be expected in an officer considering the particular information provided by the taxpayer. Whilst leaving open the exceptional case where the complexity of the law itself might lead to a conclusion that an officer could not reasonably be expected to be aware of an insufficiency, the test should not be constrained by reference to any perceived lack of specialist knowledge in any section of HMRC officers. What is reasonable for an officer to be aware of will depend on a range of factors affecting the adequacy of the information made available, including complexity. But reasonableness falls to be tested, not by reference to a living embodiment of the hypothetical officer, with assumed characteristics at a typical or average level, but by reference to the circumstances of the particular case.”

36. This approach was to some extent criticised by the Upper Tribunal in Pattullo, where Lord Glennie stated at [78] that he did not consider that Parliament could have intended to prevent assessments being made simply because the scheme in question was beyond the immediate comprehension and analysis of a reasonably competent HMRC officer. However, he made it clear at [82] that he proceeded to decide the case on the basis set out in Charlton, namely that the test to be applied was that of a hypothetical officer with the “characteristics of knowledge and expertise to be expected of someone receiving and considering the type of information contained in the tax return”.

37. In my view, the hypothetical officer postulated by s 29(5) could have been reasonably expected to be aware that the double tax relief claims in the appellant’s return were excessive. I would expect any HMRC officer of general competence, knowledge or skill to have some understanding of double tax relief, including that there are limitations on the relief that can be claimed. 15% rate is a standard rate, and in fact generally the maximum treaty rate, for portfolio dividends from companies in jurisdictions with which the UK has double tax arrangements in place. The UK has a significant number of treaties, including with all or virtually all Western economies, and certainly with major trading partners such as France and Canada. The point is not a complex one, and in my view it did not require any “white space” entry or other flagging up as Ms Murphy suggested. A hypothetical officer with a “reasonable knowledge and understanding of the law” (Sanderson at [17]) should have been able to ascertain that the claims were excessive. Although percentages are not included, it is pretty obvious from looking at the figures on page F2 that the amounts claimed were materially in excess of 15%. The percentages claimed could readily be calculated from the figures in the return. This is well illustrated by the fact that Mr Neaves did not require any additional information in order to raise the discovery assessment, or to conclude the enquiry for the subsequent year.

38. I do not think it makes a difference to the result in this case whether the focus is on an average HMRC officer of general competence or (as Charlton indicates) on a hypothetical officer with characteristics of knowledge and expertise that would reasonably be expected for someone considering the particular double tax relief claims in question. I do not think that a claim to double tax relief in respect of withholding tax on dividends is a matter of any real complexity or that it requires particular specialist knowledge within HMRC. As Charlton makes clear the focus is on the information made available, and in my view it was clear from that information that the assessment was insufficient.

39. Accordingly, I conclude that HMRC have not established that the condition in s 29(5) is satisfied.

Section 29(4)

40. HMRC’s alternative argument was that the appellant’s accountant was a person acting on the appellant’s behalf, that the accountant was careless and therefore that the condition in s 29(4) was met. The appellant did not seek to argue that the accountant was not acting on his behalf, so the only question is whether there was carelessness.

41. The original version of s 29(4) referred to fraudulent or negligent conduct rather than to the under assessment being “brought about carelessly or deliberately”. The wording was changed by Schedule 39 to the Finance Act 2008, which also made similar changes to language elsewhere in the legislation. The new language is slightly elaborated on by s 118(5) TMA, which states:

“For the purposes of this Act a loss of tax or a situation is brought about carelessly by a person if the person fails to take reasonable care to avoid bringing about that loss or situation.”

42. The original language of s 29(4) was considered by the Upper Tribunal in Colin Moore v HMRC [2011] STC 1784. It was noted without disapproval that the First-tier Tribunal in that case had applied the following formulation of the test of negligence set out by Judge Berner in Anderson v HMRC [2009] UKFTT 206 at [22]:

“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.”

43. This is an objective test. Some other more recent First-tier Tribunal cases have concluded that, in the context of penalties charged under Schedule 24 to the Finance Act 2007, “carelessness” requires that the attributes and experience of the particular taxpayer should be taken into account, rather than simply considering a hypothetical reasonable taxpayer. This reflects the test applied in determining whether a reasonable excuse exists, and has been justified by the absence of any defence of reasonable excuse from Schedule 24, in contrast to the predecessor provisions: see Martin v HMRC [2014] UKFTT 1021 (TC) at [124] to [131]. However, there is no similar justification to apply this approach to s 29(4).

44. There appears to be relatively little case law discussion of the meaning of the amended version of s 29(4). It was referred to in Richard Atherton v HMRC [2017] UKFTT 831 (TC) as an objective test, citing Moore and Anderson, but without discussing whether the change in language has made any difference (paragraphs [130] to [132]). There is a slightly more detailed discussion in a case I decided, Thomas Bubb v HMRC [2016] UKFTT 216 at [34] to [38]. The most detailed discussion I have been able to identify is in Alan Anderson v HMRC [2016] UKFTT 335 (TC), from paragraph [114] onwards.

45. In that discussion Judge Morgan noted at [118] that the explanatory notes to the changes made by Finance Act 2008 stated that the changes were being made to align the position with the terms used in the revised penalty regime in Schedule 24 to the Finance Act 2007, and at [119] that there was some indication in the contemporaneous materials that the new language in Schedule 24 was not intended to be materially different from the old test of negligence. However, Judge Morgan went on to say that this was not enough to conclude that the two terms were interchangeable, and referred to cases relating to Schedule 24 where the test of failure to take reasonable care had been interpreted as requiring the Tribunal to consider all the circumstances, including the position and experience of the taxpayer in question (citing David Collis v HMRC [2011] UKFTT 588 (TC) at [29] and Hanson v HMRC [2012] UKFTT 314 (TC) at [21]; see also Martin referred to at [43] above).

46. In Alan Anderson it was the behaviour of the taxpayer that was in question. Judge Morgan concluded at [123] that the correct approach was to assess what a reasonable hypothetical taxpayer would do in all the applicable circumstances of the actual taxpayer. This is the same approach as I took in Bubb. I have considered the position again and I have reached the same conclusion. I also consider that the test can be no different when one is considering the position of a person acting on behalf of the taxpayer, rather than the taxpayer himself. In my view, whilst the test is an objective one its objective nature is qualified to some extent, because it is necessary to take account of all the circumstances. The question of what is “reasonable” care cannot be decided in a vacuum. It must mean reasonable care in the circumstances.

47. On the facts of this case I have reached the conclusion that HMRC have not demonstrated that the appellant’s accountant, Munslows, failed to take reasonable care. My reasons are as follows:

(1) Munslows was a small firm of general practice accountants. I do not think that this is a factor that should be ignored. A small firm of that nature cannot reasonably be expected to have the same level of expertise and knowledge as a substantial firm with specialist practitioners. It was quite clear from the evidence that Munslows had limited tax expertise: for example, Mr Steinberg relied on Mr Neaves to explain how the firm should seek to reclaim the foreign tax, and to forward relevant links to HMRC’s Manuals (see [20] above).

(2) Like many other small, and indeed larger, firms Munslows relied on commercially available software to complete tax returns. Again, this is a relevant factor. It is not only reasonable to rely on third-party software, but in reality it is essential to the practical operation of the system, both from the perspective of advisers and that of HMRC. It enables large numbers of returns to be completed and filed electronically at a cost which is not prohibitive for taxpayers. The electronic submission of returns, which is of course facilitated by commercially available software, is HMRC’s preferred method of submission from a resourcing perspective and is effectively encouraged by the legislation, in particular via the later filing dates that apply to electronic submissions.

(3) It is important to bear in mind that the test in s 29(4) (read with s 118(5)) is whether the taxpayer or a person acting on his behalf failed to take reasonable care. Munslows was acting on behalf of the appellant but the producers of the software were not: the question is whether Munslows failed to take reasonable care, not anyone else.

(4) It is not realistic to expect that tax returns completed by small firms of accountants will only be completed by tax professionals, or that they will be checked either by someone with sufficient tax expertise to pick up errors of the kind that arose in this case during the checking process, or by someone who has the time to find and read all the relevant information on HMRC’s website. Mr Steinberg is clearly not a tax specialist and Munslows had no tax specialists among its partners and staff.

(5) I agree with Ms Murphy that computers (or, more frequently, those who input information into them) do make mistakes and for that reason a checking system should be in place. However, it was clear from Mr Steinberg’s evidence that a checking system was in place, and that a check was made in this case. The check focused mainly on the tax calculation pages and amounted to a high-level check for errors and omissions. In all the circumstances I do not think that this was an unreasonable approach.

(6) Whilst I think that an HMRC officer of average competence would have some knowledge of the double tax relief system, as discussed above, that is because HMRC officers are tax specialists. Accountants in general practice are not, and whilst there are some matters that I would expect a general practice accountant to be able to identify correctly (for example, basic rules about the kinds of business expenditure that are and are not deductible) I consider that knowledge about limits on double tax relief does not fall into this category.

(7) The correspondence from HMRC included in the bundle includes reference to the guidance notes to the Foreign pages to the return, and indicates that these notes state that taxpayers should check whether double taxation arrangements are in place. However, the guidance notes were not included in the documentary evidence and the point was not relied on by Ms Murphy. In the absence of those notes I do not think that I can place significant weight on this point, particularly since the reference in the correspondence does not suggest that there was any clear indication in the notes that the foreign tax actually suffered by the taxpayer could exceed the relief that can be claimed.

(8) The way in which page F2 is set out is relevant. The column in which the foreign tax needs to be entered (column C, see [3] above) is headed “Foreign tax taken off or paid”, giving no indication on the face of the return that the amount included might need to be restricted.

(9) It is also highly relevant that the errors were of a kind that might have been expected to be picked up by the software (or indeed by HMRC’s own computer programme when it processed the return). The country code included in column A on page F2 is the code provided by HMRC. The foreign tax included in column C exceeds the treaty rate for the relevant countries, a fact that can readily be determined by applying the correct percentage to the income figures included in column B. A key purpose of software of this nature is to calculate the tax due from the entries made. Munslows included what it considered to be accurate numbers based on the description of the boxes in the return, and the software accepted them but calculated the tax incorrectly by not recognising that credit could not be given for the full amount of the foreign tax. Given that a principal function of software of this nature is to calculate the tax due, it was not unreasonable to assume that the foreign tax credit amount would be calculated accurately, or at least that there would be some warning that it might not be.

48. I therefore conclude that HMRC has not established that the condition in s 29(4) is satisfied.

Conclusion

49. Accordingly, the appeal is allowed.

50. 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.