Jaws 2

Background

When I ran my first few discovery cases (including Charlton, Sanderson and Pattullo), it was necessary to take the Tribunals through the history of what is now the Taxes Management Act (TMA) 1970 s 29, comparing the current rules with those that existed pre-Self Assessment. In the subsequent few years, however, there have been so many cases concerning the discovery rules that it is fair to assume the Tribunal judges are now fully acquainted with them (and also the various issues that arise from these rules).

As had already been established by the Court of Appeal in Tower MCashback, the current discovery rules are based on the pre-Self Assessment code but imposed additional restrictions on HMRC’s right to make a discovery assessment so as to ‘underline the finality of the self-assessment’ itself. Accordingly, most of the subsequent case law has concerned the application of the additional conditions required to justify a discovery assessment (looking at the quality of the disclosure made on the tax return and other related documentation and on the existence of what is now described as careless or deliberate conduct).

However, this meant that many advisers started to overlook the need for there to be a discovery and, more importantly, what is actually meant by the word ‘discovers’. Nevertheless, as the Charlton and Pattullo cases make clear, this remains an essential part of the process: an assessment will not be valid unless there is a prior discovery. For example, it is not enough for an officer to make a ‘protective assessment’ (so as to get around an imminent time limit) and then embark upon an enquiry to see whether there is any basis for the officer’s assessment. On the other hand, the Upper Tribunal in Charlton and Pattullo made it clear that ‘discovers’ does not merely mean ‘forms the view’ (in the sense that the officer reasonably and honestly believes that tax has been underpaid). Implicit within the word ‘discover’ is the requirement that the officer acts upon the information within a reasonable period of time – whilst the discovery is still ‘fresh’ or not yet ‘stale’.

HMRC have steadfastly refused to accept that the word ‘discovers’ carries this additional requirement although the Charlton/Pattullo view has become quite firmly established in the First-tier and Upper Tribunals. The Tooth case was its first outing in the Court of Appeal.

The facts of the case

The facts were set out more fully in my previous article on this case, ‘The honest Tooth’, as published in the May 2018 issue of Tax Adviser. In short, Mr Tooth took part in an avoidance scheme which (he thought) entitled him to accelerate tax relief through the carry-back rules. As the tax return prepared by HMRC did not allow for such a claim to be accelerated in this way, Mr Tooth used a wrong box on the tax return to achieve his desired purpose, albeit with a white space disclosure explaining what he had done.

HMRC quickly recognised the fact that Mr Tooth’s return did not comply with their views of the law and promptly opened an enquiry. However, rather than enquiring into the return under the TMA 1970 s 9A, they considered that the carry-back claim had been made outside the tax return (albeit on the tax return form). 

Accordingly, HMRC opened their enquiry under the TMA 1970 Sch 1A provisions instead. The correctness of this course of action was upheld by the Supreme Court in the case of Cotter. However, the facts of Cotter proved to be an exception and it appears that most taxpayers (including Mr Tooth) had put themselves in a situation whereby the correct approach was in fact to use the s 9A provisions. Accordingly, in relation to these taxpayers (including Mr Tooth) HMRC made a series of ‘discoveries’ in or around 2014 that their enquiries had been embarked on the wrong statutory basis. Accordingly, they issued a number of discovery assessments in order to capture the tax that they ought to have pursued via s 9A.

The difficulty for HMRC is that s 29(1) is quite prescriptive as to what may be discovered: it does not include discovering a procedural error by HMRC. What must be discovered is, essentially, that the original tax was wrong. On the facts of the Tooth case, that must have occurred back in about 2009, although there was no clear finding of fact in that respect. Given the apparent five-year gap between the original discovery and the actual assessment, the question of staleness has naturally arisen – in Pattullo, an 18-month delay was described as too long ‘on any view’.

The Upper Tribunal felt that the First-tier Tribunal had reached an unjustified conclusion when it formally decided that the relevant discovery had been made in 2014. However, because of a lack of clarity as to the First-tier Tribunal’s decision, the Upper Tribunal would have remitted the case back to the First-tier Tribunal for further findings of fact (which would probably have led to the formal conclusion that the relevant discovery had been made in 2009). However, this proved unnecessary because the Upper Tribunal endorsed the First-tier Tribunal’s views in concluding that the taxpayer had not been guilty of a deliberate inaccuracy in his tax return.

HMRC appealed against the latter finding and also against the Upper Tribunal’s conclusion that the relevant discovery had not been made in 2014.

The Court’s decision

The case came before Lords Justice Patten, Floyd and Males.

The Judges were unanimous so far as the discovery issue is concerned, but reached different views concerning the deliberate inaccuracy.

In relation to discovery, HMRC were considered to have adopted an argument which they had not run previously and which was factually inconsistent with the documentation that they had put before the First-tier Tribunal. In short, the Court considered that the documentation pointed clearly to the fact that the relevant discovery had taken place in 2009 and that, in 2014, the only thing ‘discovered’ was that HMRC had previously taken the wrong procedural path – such a discovery not in itself coming within the scope of s 29(1). In this regard, the Court’s conclusion was very similar to that of the Upper Tribunal.

However, the Court considered that the Upper Tribunal was wrong to remit the case back to the First-tier Tribunal for further finding of fact. This was because HMRC’s case before the First-tier Tribunal was that the discovery had occurred in 2014 whereas the evidence put forward was clearly inconsistent with that (or at least did not prove the point being made by HMRC). In the circumstances, it was not appropriate for HMRC to have a second bite of the cherry. Accordingly, it was inappropriate for the matter to be remitted for further findings of fact by the First-tier Tribunal. HMRC’s appeal was therefore dismissed.

As a result of the decision on discovery, it was not strictly necessary for the Court to consider the question of deliberate inaccuracy. However, the point was fully argued and therefore the Court expressed its views.

The Upper Tribunal had considered that there was not even an inaccuracy in the return because, even though Mr Tooth had used the wrong box so as to accelerate his loss relief claim, he had made his position clear in the white space. Therefore, taken as a whole, the return reflected Mr Tooth’s honest view and was therefore not inaccurate. As I mentioned in my article last year, this was a conclusion which ‘caused me some surprise’. Although Floyd LJ upheld this part of the Upper Tribunal’s decision, the majority considered that an inaccuracy in a return could not cease to be an inaccuracy merely because it is corrected or explained elsewhere.

Where the three judges were at one, however, was in relation to the second part of the test: being whether the inaccuracy (in Floyd LJ’s case, assuming that there was indeed an inaccuracy) can be said to have been deliberate. All three Judges decided that if the inaccurate return entries were made with the intention of achieving ‘a particular fiscal result’, then that is sufficient to satisfy the statutory test.

Therefore, but for the question on discovery, the Court of Appeal would have allowed HMRC’s appeal and upheld the assessment.

Commentary

In respect of the discovery point, the Court of Appeal has reflected the longstanding view that a party needing to prove something in a case should usually be expected to do so at the first hearing and is not to be given a second chance with the benefit of hindsight.

One of the reasons why the Tooth case was so eagerly awaited was that some clarity was being sought on the need for a discovery assessment to be made whilst the underlying discovery was still new or ‘fresh’, i.e. not stale. Although a series of First-tier and Upper Tribunal decisions have emphasised the importance of HMRC acting reasonably promptly, HMRC have never quite accepted that they need to act with any alacrity and say that they are bound only by the statutory time limits. Consequently, recent Upper Tribunal decisions (whilst following the previous case law) have referred to the need for clarity from the higher courts.

In Tooth, the question of staleness was not directly relevant. However, it was clearly in the background because HMRC would not otherwise have needed to put all their efforts in showing that the discovery took place as late as in 2014. Indeed, the Court of Appeal expressly endorsed the view that a discovery represents something new and that this is implicit from the word itself. Furthermore, the discovery must relate to the conclusion itself (i.e. that too little tax had been paid) rather than the reason for that conclusion. These comments will no doubt prove encouraging to Mr Beagles whose appeal (due to be heard in October 2019) is when the Court of Appeal is next due to consider the requirement for HMRC to issue the assessment whilst the discovery is still fresh.

I must, however, admit to being astonished by the Court’s decision on deliberateness. As the Judges made clear, ‘there is no question of Mr Tooth or his advisers having acted dishonestly or even reprehensibly’. However, they considered that this did not prevent the inaccuracy from being a deliberate one. In my view, the conclusion (whilst, as I noted last year, being supported by a literal interpretation of the statutory words) takes no account of the broader statutory landscape – indeed, it does not immediately explain why HMRC should be given longer to assess a taxpayer who has honestly made an error after proper thought than a taxpayer whose error is attributable to carelessness. And a similar point can be made by looking at the penalty code where deliberate errors are treated as more serious than careless ones.

In my view, HMRC have got lucky on this point and, although they are likely to be disappointed by the fact that they lost on the discovery point, I think that they ought to be reluctant to risk the Supreme Court revisiting the meaning of deliberate inaccuracy, although the grapevine says that an application is pending. On the other hand, I am hopeful that the Tribunals will manage to distinguish most cases from the particular facts of Tooth so as to bring the meaning of ‘deliberate’ much closer to the generally-accepted view that it requires a deliberate attempt to pay less tax than is known to be due, and the First-tier’s decision in Leach appears to be such an example.

What to do next

So far as the staleness point is concerned, it is probably best to await the Court of Appeal’s consideration of HMRC’s appeal in Beagles.

However, the broad definition of deliberate inaccuracy is likely to cause considerable worry. Where an error in a tax return has been identified, there is a large risk that HMRC will now seek the higher penalties that are applicable to deliberate errors and/or will go back up to 20 years because of the extended time limits available for deliberate errors. There is also the risk of naming and shaming.

Clients should be warned of these risks and the likelihood that someone is going to have to take another case to the Court of Appeal (or beyond) to ensure that the previously-accepted meaning of deliberate is reinstated.