Discovery assessments are an important part of HMRC’s enforcement arsenal and a common area for dispute in tax litigation. This means that the Supreme Court’s recent decision in Tooth v Revenue and Customs Comrs [2021] UKSC 17 is particularly significant in defining the nature of the arguments available to taxpayers when challenging such an assessment.


In January 2009, Mr Raymond Tooth, the taxpayer, entered into the Romangate tax avoidance scheme, the purpose of which was to create employment-related losses which could be carried back to the 2007–08 tax year. His accountants subsequently attempted to fill out his tax return for the relevant year but were not able to input the loss in the right box on the form as a result of a technical glitch in the HMRC-approved IRIS software. Accordingly, they entered the loss in the unrelated partnership income section of the return together with an explanation of the loss and why they had done this. The tax computation automatically generated showed that Mr Tooth was entitled to a repayment of income tax.

HMRC took the view that the Romangate scheme had always been ineffective. However, they did not open an enquiry into Mr Tooth’s 2007–08 tax return under section 9A of the Taxes Management Act 1970 (“TMA 1970”) prior to the expiry of the one-year window for doing so in 2010. It was only following the introduction of retrospective measures to counteract the scheme that HMRC finally issued a “discovery assessment” under section 29 TMA 1970 in October 2014.

By way of background, HMRC have two procedural methods open to them to assess unpaid income tax or CGT (or indeed corporation tax). They must either open an enquiry into the relevant tax return within the enquiry window (broadly one year from the date the return is filed). If they do not open an enquiry, HMRC can issue a discovery assessment “if an officer of the Board or the Board discover […] that an assessment to tax is or has become insufficient”, but only where certain further conditions are satisfied and subject to certain further time limits (four, six and twenty years depending very broadly on the level of culpability). 

In the present case, HMRC needed to show that the insufficiency “was brought about … deliberately by the taxpayer or a person acting on his behalf”, as the assessment was made more than 6 but fewer than 20 years before the end of the relevant tax year. Section 118(7) TMA 1970 further defines deliberate conduct in this context as “a situation that arises as a result of a deliberate inaccuracy in a document”. Mr Tooth challenged the assessment on the grounds that (i) HMRC had not made the requisite “discovery” / the relevant “discovery” had become “stale” and (ii) that his tax return did not contain a “deliberate inaccuracy”.

Mr Tooth was successful in the First-Tier Tribunal on the deliberate inaccuracy point but not on the discovery one and HMRC appealed. The Upper Tribunal sided with Mr Tooth on both issues. On the second appeal, in a further reversal, the Court of Appeal found in his favour on the discovery point but not on the deliberate inaccuracy one. The case was finally heard in the Supreme Court in March 2021.

A Fresh Take on Staleness

Most of the interest in the Tooth decision comes from its consideration of what it means to make a discovery, and whether the concept of “staleness” has any part to play in the discovery assessment regime. Ultimately, the Supreme Court sided with HMRC on this issue and provided important guidance on how the regime should be interpreted in future.

First, it emphasised that the regime referred to the discovery being made by “an officer” rather than by HMRC as a whole. This meant that a discovery was not made “once and for all” and so allows for the possibility that a future officer is not prevented from “discovering” a deficiency simply because it has already previously come to the notice of one of his colleagues.

Second, it rejected the idea that a discovery could become “stale” if it was not acted upon within a reasonable time period. Although strictly this part of the decision is obiter, it addresses an idea that has been the subject of much argument in recent case law but which, in truth, has always rested on shaky linguistic foundations. The “staleness” doctrine depends on the implication that “to discover” connotes the finding of something new, and the (supposed) further implication that the requirements for a discovery assessment may not be satisfied if, due to the passage of time, the relevant finding is no longer new at the time of the assessment. It is this further (supposed) implication that the Supreme Court dispelled in its judgment. 

The court clarified that HMRC simply needed to make an assessment within the relevant limitation period, irrespective of when the “discovery” was first made. In the court’s view, any delay on the part of HMRC in actually raising an assessment could only be challenged on public law grounds in judicial review. This would depend on the facts of a particular case but such a challenge is likely to be very difficult indeed in practice. 

In practical terms, the dispelling of the concept of staleness leaves taxpayers even more susceptible to HMRC re-opening their tax affairs, with the consequent loss of certainty, for many years after the events in question. This is especially acute in the context of the ever-increasing powers given to HMRC and the increased time limits introduced in recent years – in particular the 2019 extension of the discovery assessment time limit to 12 years in the case of “offshore” matters or transfers (which is very broadly defined).

Deliberate Inaccuracy: An End to Tunnel Vision

However, Mr Tooth won on the issue of whether his return contained a “deliberate inaccuracy” and it was this which ultimately determined the appeal in his favour. The Supreme Court affirmed the common-sense conclusion that “deliberate inaccuracy in a document” must mean an inaccuracy that is deliberate, rather than a deliberate statement which is inaccurate. In other words, the taxpayer (or his agent) must know and intend that the statement/entry is inaccurate (or perhaps be reckless as to its accuracy). The court accused HMRC of “tunnel vision” in focussing only on the boxes in the return containing the inaccuracy. Rather, the document must be considered as a whole. In this case, that meant that the accompanying explanation provided by Mr Tooth’s advisers needed to be taken into account. Consequently, if there was any inaccuracy in the tax return, it was not a deliberate one.