The honest tooth


Image credit: © iStock/Victor_85


Keith Gordon considers a recent Upper Tribunal case which revisits the question as to when a discovery assessment is valid

Key Points 

What is the issue?

HMRC’s discovery assessment powers should reflect a remedy of last resort, limited to those cases where broadly it was not appropriate to expect HMRC to have been able to identify the possible tax risk during the enquiry window.

What does it mean to me?

As the Tooth case demonstrates, there are additional protections given to taxpayers which mean that HMRC have to overcome certain hurdles before a discovery assessment will be valid.

What can I take away?

If you or a client faces a discovery assessment, it is worth checking whether or not the statutory conditions are met. If in doubt, ask, because there is no point paying tax unnecessarily.

The arrival of Self Assessment more than 20 years ago meant that most challenges by (what is now) HMRC into a taxpayer’s income tax, capital gains tax or corporation tax position should be undertaken through the statutory enquiry mechanism. Consequently, HMRC’s discovery assessment powers should reflect a remedy of last resort, limited to those cases where broadly it was not appropriate to expect HMRC to have been able to identify the possible tax risk during the enquiry window (or, in cases where an enquiry was opened, during the enquiry itself).

There are no special rules applying in cases involving so-called tax avoidance, although the advent of the rules requiring early notification of tax avoidance arrangements (‘DOTAS’) back in 2004 (and its expansion in 2006) should have meant that it would be even easier for HMRC to identify potential enquiry cases in time. Nevertheless, for reasons that appear to be a combination of inappropriately drastic staffing cuts and institutional disorganisation, it was precisely in many of those types of avoidance case where timely enquiries were not opened and HMRC have been seeking to remedy the situation by playing catch-up through the use of their discovery assessment powers.

Hot on the heels of a defeat in one such case in the First-tier Tribunal (Hicks v HMRC [2018] UKFTT 22 (TC)), HMRC were hoping for better luck in their appeal in the case of HMRC v Tooth [2018] UKUT 38 (TCC).

Facts of the case

In the 2008/09 tax year, Mr Tooth participated in a tax avoidance arrangement which was designed to give rise to employment income losses, such losses to be set off against other taxable income enjoyed by the taxpayer. As did many other participants of that scheme (and similar arrangements), Mr Tooth sought to use the carry-back rules so as to set off the losses against the income of the previous tax year (i.e. 2007/08).

The pro forma 2007/08 tax return included a space for later years’ losses to be claimed and, as the loss had by been suffered by the time that the return was submitted, Mr Tooth duly made the claim.

Many readers will recall the case of HMRC v Cotter [2013] UKSC 69 where the Supreme Court had to decide the precise procedural consequences of such carry-back claims. The Supreme Court concluded that the claims do not constitute part of the earlier year’s tax return, even if they are legitimately referred to and included on the tax return form for that earlier year. In other words, there is a difference between the ‘tax return’ as that term is understood for some purposes within the Taxes Management Act 1970 and the tax return form, being the set of papers that might colloquially be considered to constitute the tax return.

An associated question that arose in Cotter was how to give effect to the carry-back claim, when so included on the tax return form. It was generally accepted that relief could be obtained pretty quickly (and that was the purpose of including it on the earlier year’s form), but different views existed as to whether it reduced the self assessment calculation for that earlier year or merely provided a credit against the liability arising under SA. This debate was also resolved by Cotter which concluded that the claim gave rise to a ‘free-standing credit’. (Technically, the argument was conceded by Mr Cotter and therefore the Supreme Court did not formally decide the matter, although it agreed that the concession had been rightly made.)

Mr Tooth’s 2007/08 tax return, however, was submitted long before the Cotter case came to Court and Mr Tooth had been advised that the relief should be effected through the earlier year’s self assessment calculation. The difficulty was that HMRC’s forms did not permit such an approach (as they had been prepared in accordance with the view which eventually prevailed in Cotter). Mr Tooth was therefore required to improvise, but included a ‘white space’ note in the tax return explaining that he had treated the loss as an in-year partnership loss specifically to ensure that the self assessment figure reflected the reduction to which he believed to he was entitled.

In October 2014, HMRC issued a discovery assessment to recover the tax that Mr Tooth had underpaid in his 2007/08 self assessment. As this assessment was made more than six years after the tax year to which it related, HMRC were required to demonstrate that Mr Tooth’s under-assessment was attributable to deliberate conduct by him or a person acting on his behalf.

The First-tier had rejected HMRC’s contentions quite comprehensively. HMRC duly appealed to the Upper Tribunal, where the case came before Mr Justice Marcus Smith and Judge Charles Hellier.

The Tribunal’s decision

The Tribunal considered the argument put forward by HMRC. They had argued that Mr Tooth’s tax return had been submitted deliberately (an assertion which I do not think anyone would challenge). They also pointed out that Mr Tooth’s under assessment arose because his tax return contained an error (a point which I shall revisit further below). Consequently, according to HMRC, the under assessment as a result of a deliberate act by Mr Tooth (i.e. the deliberate submission of his tax return).

Although the argument follows the literal wording of the legislation, it clearly makes a mockery of the fact that the deliberate conduct test was meant to limit HMRC’s assessment opportunities after six years to the most serious of cases, a point which the Tribunal readily acknowledged. As the Tribunal further noted, using the phrase ‘self-evidently’, the mere completion of a tax return cannot of itself be the deliberate conduct intended by Parliament: ‘the deliberation must relate to the inaccuracy, not merely the completion and submission of the document’.

Of course, the inclusion of an inaccurate figure could be deliberate. But, as the Tribunal continued, it is not necessarily so. In particular, an allegation of deliberately bringing about a tax loss is a serious one, tantamount to an allegation of fraud. Given that Mr Tooth had taken steps to draw the point to HMRC’s attention, the Tribunal concluded that Mr Tooth’s conduct did not entitle HMRC to assess out of time.


The Upper Tribunal’s decision on the meaning of deliberate conduct represents a good deal of common sense. Indeed, other than perhaps a sense that for HMRC tax avoiders are so evil and should be pursued at all costs, it is really hard to understand why anyone would actually advance the argument that the mere submission of a tax return was a deliberate act within the meaning of the legislation.

There are other cases in the First-tier which similarly turn on the meaning of deliberate conduct and perhaps one should be grateful that HMRC sought to overturn the First-tier’s decision because there is now a binding precedent which should be used as a guide in those other cases. I do fear that HMRC’s motives were not so altruistic and that they were hoping for a more sympathetic hearing. If my fears are correct, then HMRC must be regretting their decision (although they might try their luck at the Court of Appeal – although in my view such a course of action would represent a complete waste of public funds).

In the meantime, another cause for HMRC’s regret was the fact that, given that HMRC were appealing against the decision of the First-tier, Mr Tooth took the opportunity to challenge part of the First-tier’s decision that went against him. One condition that needs to be met before a discovery assessment can be issued is that HMRC must have made a discovery. (It seems like an obvious point but for many years tax advisers and HMRC officers alike tended to focus on the additional conditions that were introduced by the Self Assessment changes from 1996 without remembering that the first hurdle for HMRC to overcome was the making of a discovery.)

Of course, most (but not all) assessments are based upon a reasonable belief that tax is payable and this belief will usually be the basis of a discovery. However, as has come to the fore in recent years, the word discover also implies a relative newness to the belief.

In Tooth, HMRC had first written to the taxpayer back in 2009 pointing out, inter alia, that retrospective legislation meant that his losses would be disallowed. However, reflecting some of the general confusion which was not resolved until Cotter, HMRC initially sought to remedy this by opening an enquiry into the claim (rather than the tax return itself). As Cotter later explained, that was generally the right approach but not in cases where the self assessment in the earlier year had been reduced to give effect to the claim. In such cases, HMRC were obliged to open an enquiry into the earlier year’s return itself.

The Upper Tribunal considered that, contrary to the approach taken by the First-tier, the correspondence in 2009 demonstrated that HMRC had made the relevant discovery (i.e. the belief that Mr Tooth’s self assessment calculation was insufficient). Given the fact that the assessment should ordinarily follow whilst the discovery is still new (or fresh or not stale) a five-year delay before the assessment was issued meant that the assessment was invalid and gave Mr Tooth a second reason to win the case.

The above two conclusions (concerning deliberate conduct and staleness) are ones that I fully endorse and am running in my own discovery cases. However, there was a third aspect of the Upper Tribunal’s decision which caused me some surprise. The Tribunal set out in full the conditions for HMRC to make a discovery assessment on the basis of a taxpayer’s deliberate conduct. In its analysis the first condition was that there must be an inaccuracy in a document given to HMRC. Ordinarily, I might have said that this would be satisfied in all cases because the assessment is meant to correct the inaccuracy. However, referring to Mr Tooth’s white space entry, the Tribunal considered that this represented the taxpayer’s bona fide attempt to explain what he recognised might be a controversial basis for completing his tax return. For as long as the tax return is thought to be correct by the taxpayer then, according to the Upper Tribunal, it will not be an inaccurate tax return and therefore the first condition is not met. I will be interested to see how this argument fares in later cases. However, even if it is not followed, the next step in the Tribunal’s thinking (about the deliberateness of the error) is in my view hard to disagree with.

What to do next

The most important thing to remember is that the discovery assessment rules give HMRC a second bite of the cherry when they have failed to open a statutory enquiry into a tax return. However, as the Tooth case demonstrates, there are additional protections given to taxpayers which mean that HMRC have to overcome certain hurdles before a discovery assessment will be valid. Standard HMRC literature does not usually mention these and therefore taxpayers will often assume that they are liable for the additional tax being charged. In short, if you or a client faces a discovery assessment, it is worth checking whether or not the statutory conditions are met. If in doubt, ask, because there is no point paying tax unnecessarily.

No votes yet