Brought to you by
Share this content
Part of the
Brought to you by Tolley

At what point should somebody’s tax affairs for a particular year become final?

26th Jul 2016
Brought to you by
Share this content

We all want certainty in our own tax affairs and those of our clients. But how should that certainty be achieved? This is an extremely complex issue and I don't think that anybody really thinks that what we have at the moment works properly.

Back to basics

Let's start with some basic points. Everybody would, I think, accept that HMRC must have some powers to enquire into a person's tax return. In a parallel universe there might be a utopian country where every citizen's tax return is correct and complete in all particulars, but until we get there HMRC must have some enquiry powers. At the moment normally HMRC has a year from return submission in which to open an enquiry. Some might think that this is a bit on the long side, but it is hard to say that it is completely unreasonable.  

Almost everybody would I think also agree that in cases where people have deliberately fiddled their taxes HMRC ought to have the power to go back to collect the arrears. At the moment HMRC can go back 20 years in cases of deliberate understatement or concealment. Some might argue that this there should be no time limit, but in practice going back more than 20 years becomes almost impossible because of a lack of information and also because of the effect of inflation. So broadly speaking the 20 year limit is appropriate.

The third thing which I believe everybody would agree with is that if a person a has made a mistake and paid too much tax there should be a mechanism for him/her to correct the mistake and obtain a tax repayment.

Paying too much tax

But from that point consensus breaks down. Let's take the case of somebody who finds that they made an error in their return and paid too much tax because, for example, they failed to claim for an expense which was allowable. At the moment the time limit for a claim is (broadly) 4 years.  Is that long enough? If the state has "my" money shouldn't I be able to get it back whenever I want, provided that I can prove that I am entitled to it? While doing a clear out of papers recently I came across my 1989-90 return. Should I be able to look back at this and claim for expenses I didn't know about at the time? The purist in me says yes, but I think in practice there has to be some limit. Should there be a simple cut-off date or should you be able to go back longer if the amounts are greater? Is four years long enough? Should it be longer if you didn't employ an agent to do your return? All of these are questions which I believe should be open for consideration. 

There is then the related issue of whether the time limits should be the same for overpayments as they are for underpayments. At the moment, as we shall see, in most cases not involving deliberate behaviour HMRC can go back 4 years, so there is symmetry as between the citizen and the state. That seems logical but it is fair? After all an extra £500 refunded to me is much more valuable (at least in my eyes) than another £500 for the state. Might there be an argument that the time limits should be weighted in the taxpayer's favour? 

Practical difficulties

We've started to see a few cases on the way that the time limits for claims for tax repayments work, following changes in the legislation a few years ago. Cases like Raftopoulou and Fessal show that the rules are far from straightforward. The latter in particular is worthy of comment. The taxpayer was a barrister who was subject to the transitional rules for recognising profits. The details don't matter, but the end result was that his returned  profits were understated for 2005/6 and 2007/8 years but overstated in 2006/7. As far as one can tell all of his profits were declared for tax purposes so this was purely a timing difference.  Unfortunately for the taxpayer the system worked against him.

HMRC were able to assess the additional profits for 2005/6 and 2007/8 under the discovery rules (which I discuss in more detail below) but he was out of time to make a claim for repayment of tax for 2006/7. The tribunal agreed that this was indeed the effect of the time limits in the legislation but found in favour of the taxpayer on human rights grounds. While that was undoubtedly a good result for the taxpayer is it concerning that the tax legislation creates the problem in the first place. Shouldn't it be able to recognise the overall position rather than take each year separately?


This brings us to the most difficult area of all.  When should HMRC be able to go back to review a return for a year which is out of time for an enquiry? In the absence of fraud should they be able to do it all? After all there is a generous 12 month limit for HMRC to raise enquiries: if they don't take that opportunity why should they get a second bite of the cherry? But what happens if the taxpayer has only given limited disclosure on a particular point. Say for example he has not analysed legal expenses in the tax computation and it turns out that one of the items should have been disallowed? What about the case where the taxpayer has used an estimated valuation in computing a capital gain without disclosing the fact that she had not obtained an independent professional valuation? And what happens when the taxpayer has entered into an avoidance scheme but has not fully disclosed this in the return.

These are all difficult issues.  In principle the rules are straightforward. If you have given HMRC the information on your return and HMRC fails to enquire you are protected from a discovery assessment. If you haven't HMRC  can go back four years or six in the case of failure to take reasonable care. But simple as the principle should be, dealing with it is practice is anything but. There are now a whole host of cases on the discovery rules and each one seems to bring out more areas of difficult. Indeed to understand some of the recent decisions tax training is not enough: one needs a degree in philosophy and probably one in linguistics as well.  Take the following paragraphs taken verbatim from a recent judgement.

"This condition is that at the time when "an officer" of HMRC ceased to be entitled to give notice of its intention to enquire into the taxpayer's return (in the present case 31 January 2006), "the officer" could not have been reasonably expected to be aware of the situation mentioned in subsection (1), i.e. the insufficiency. To the untutored mind, that reference to "the officer" would appear to be a reference back to the officer ("an officer") mentioned at the beginning of subsection (5) who, at the cut-off date, ceased to be entitled to give notice of his intention to enquire into the taxpayer's return. In other words, it is a particular individual, not an abstract hypothetical officer with the knowledge and expertise mentioned in Charlton nor an average or reasonable officer riding the fiscal equivalent of the Clapham omnibus. As has been noted elsewhere, the section appears to contemplate that, as may well have happened in the past, each tax return would be considered by an individual officer who would consider its contents, form a view as to its accuracy and take a decision whether or not to give notice of his intention to enquire into it. Whether or not it is that same officer who subsequently makes the discovery in terms of subsection (1) is perhaps unimportant. The important point is that the right under subsection (1) to make an assessment upon discovering that the assessment to tax in the return is insufficient is qualified by the proviso in subsection (5) that an assessment cannot be made if the insufficiency could reasonably have been discovered by the officer handling the case at the cut-off date from the information made available to him by the taxpayer. The discovery in subsection (1) finds its counterpart in the state of awareness in subsection (5). They involve the same level of knowledge or awareness and effectively mean the same thing. It seems to me that the question of reasonableness comes in, not in the need to construct a fictional hypothetical officer but rather in the test of whether the actual officer ought reasonably to have been aware of the insufficiency at the cut-off date from information provided by the taxpayer."

And to give you a further flavour of the weird judicial world in which the discovery provisions now are tested, how about the following further quotes from the same case.

"The word "if", like many words in the English language, has a variety of shades of meaning. It may be purely conditional. But it may equally have a temporal aspect, as in the expression "if and when" (e.g. if the sun comes out we shall go to the beach)."

However, a state of awareness, understanding or perception seems to be the result of the application of the test rather than the test itself (cf paragraph 58). Moreover, it is difficult to see why a conclusion is excluded from a state of awareness. One plus one equals two. Two is the result of adding one to one. It is a conclusion. Knowing that one plus one makes two is a state of awareness. We agree, however, that the awareness of a hypothetical officer is not to be measured by probabilities. Probability is normally deployed to determine facts. The deemed awareness of the hypothetical officer is determined by reference.

How on earth did we end up with a system where it is necessary to consider the relative knowledge and belief of

  • the actual officer making the assessment,
  • of a hypothetical experienced officer, and
  • of an reasonable officer riding the Clapham omnibus to his nearest tax office. It has all become truly bizarre.

Of course all of this makes really good ammunition for tax commentators (see for example Adam Cragg's article in Tax Journal of 4 December 2015 or Derek Francis piece in Taxation of 10 December 2015.

Indeed a quick search of discovery within Tolley Library brings up a whole host of commentary and technical analysis. All grist to the mill for the adviser of course, but not surprisingly a source of great frustration to taxpayers who want a simple yes/no answer to the question "Can HMRC go back and assess me or not?"  I don't think a philosophical discussion about the nature of the statement "one plus one equals two" is likely to go down very well with a client who has thousands of pounds at stake on the matter.

We desperately need a fresh look at this whole area because I can't believe that anybody believes that what we have at the moment is fit for purpose. The issue of finality will become even more acute in a digital environment where the plan is to move as much of the tax system into real time.  If I am reporting in real time is it not reasonable for me to expect in return that I will get finality in real time?  So far I have not seen anything to suggest that this will be the case, but I really hope that the digital strategy will be used as an opportunity to revisit this whole area.

One final thought. Many years ago a colleague suggested that to me that there should be a box on the tax return which you could tick to ask HMRC to enquire into the return. The thinking was that if you had nothing to lose you could get certainty either because HMRC would enquire into your return and find nothing, or, more likely, decide not to enquire but would be shut out from a looking again at a later date because they had passed up the opportunity to do so when you asked them. At first though it seems bizarre that anybody would actually invite an enquiry, but the more you think about it the more the idea has its attractions. In my experience the vast majority of clients do want early certainty in their tax affairs.  Such a mechanism, re-engineered for the digital age, must just be one way of allowing them to sleep at night.