Skip navigation |


Category Technical Articles
AuthorTechnical Department
In the concluding part of his article,David Goldberg continues his analysis of the current attitude to anti-avoidance in the courts. David Goldberg QC is a barrister specialising in tax at Gray’s Inn Chambers, London, WC1R 5JA.
Published in the July 2002 issue of Tax adviser The New Approach

Despite this intemperate outburst from the Inland Revenue (see Tax Adviser, June 2002, p. 17), in the courts the law was regarded as settled by Duke of Westminster [1936] AC 1 until, in 1981, there was the explosion – the nuclear explosion – of IR Commrs v Ramsay[1982] AC 300 setting off a chain reaction, leading through IR Commrs v Burmah Oil Ltd [1982] BTC 56, Furniss v. Dawson [1984] BTC 71 and Ensign Tankers (Leasing) Ltd v Stokes [1992] BTC 110 to IR Commrs v McGuckian [1997] BTC 346.

These cases seemed to proclaim a new approach – sometimes called an emerging principle – and they rightly caused a good deal of disturbance in the dovecotes inhabited by tax advisers. The problem was not so much what the cases were actually deciding – that is the actual outcome of the case looking only to its own facts – but, rather, with the way in which the decisions were expressed.

I think exception can be taken to three particular features of the judgments before MacNiven (HMIT) v Westmoreland Investments Ltd [2001] BTC 44.

First there was an appeal to jurisprudence derived from the United States (US). Now this was dangerous and uncertain. It was dangerous because, although there are many features common to all tax systems, there are very considerable differences between the US and the United Kingdom (UK) tax systems; and principles developed in the context of US statutes cannot easily be applied to our domestic law.


It was uncertain, because the American courts seem to have adopted at least four different approaches to tax avoidance cases. They are:

  • a sham doctrine, which is unobjectionable and broadly the same as we have here and which nobody will get very excited about;
  • a no business purpose doctrine, that a transaction may be disregarded if it has no commercial purpose;
  • a step doctrine which is that, where a transaction is carried out in several steps, the steps may be disregarded; and
  • a substance over form doctrine, which is that one looked to the substance and not to the form.

These were all different doctrines, applied from time to time as distinct principles and without much coherence as to when one rather than the other would be applied.

Before MacNiven, the English – and I should add, the Scottish – courts appeared to be creating a single doctrine which was an amalgamation of the step doctrine with the no business purpose doctrine, though the McGuckian case suggested that we might be adopting a substance over form approach which is, of course, a different thing altogether.

It did, however, appear that we were developing a distinct UK doctrine rather than just importing a number of US doctrines; MacNiven shows that, while there remains some universality in the common law (so that we do not altogether disregard what common law judges in other jurisdictions are doing), nonetheless, the doctrines applied here are UK doctrines and not doctrines imported from the US.

The second feature of the judgments before MacNiven to which exception can be taken is that many of the judges were putting a moral principle as if it were a legal principle. There were, undoubtedly, some judges who were in effect, saying that the principle was 'if we do not like it, it is not going to work'. Lord Templeman was, of course, chief among these judges, but he was not the only one.

This is not the way to run a legal system. This is to substitute discretion for law and we should not do that. It is right to say that the judges who took this moral view were very much in the minority, but they were there and were not without influence. The worry was that their view might spread. We can see from MacNiven that it has not done so.

The third feature is that some of the remarks made by some of the most respected judges in these cases were excessively extravagant. Lord Diplock’s well known, if slightly nauseating, remark in Burmah that ‘it would be disingenuous to suggest and dangerous on the part of those who advise on elaborate tax avoidance schemes to assume that Ramsay’s case did not mark a significant change’ is one of these rather extravagant remarks. A friend of mine was once going to write an essay called ‘Lord Doplick’s Ligoc’; and remarks of this sort explain why.

Another extravagant remark was Lord Scarman’s map-making analogy in Furniss v Dawson, where he says that it will be in an ‘area of judge made law that our elusive journey’s end will be found’.

Pausing here, then, and looking at the dicta in these cases up to McGuckian, it would or might appear that we had developed a doctrine that tax avoidance cases were to be decided by some principle of judge made law, which might depend to some extent upon decisions of the US courts, and which could also turn heavily on the degree of indignation and repugnance felt by the judge who was deciding the case in question.

If that is what had happened it would, of course, have been very unsatisfactory. Judges are supposed to decide cases in such a way as to provide a guide to the result of future cases. They ought to be providing a degree of certainty and a doctrine of this sort certainly did not do that.

Happily, however, what is important in our system of jurisprudence is the point actually decided by a case, rather than the peripheral commentary surrounding the decision.

When each of these individual cases up to McGuckian is analysed, it will be found that what it actually decided was something very much narrower than the commentary suggested. If we concentrate on the actual decision in each case rather than on the extravagant remarks, and even though some of the remarks were, at the least, worrying, it can be seen that the cases were not actually deciding any very wide issues. The actual decision in each case was quite narrow.

And then came McGuckian.


McGuckian was and is an interesting and important case for three or four reasons.

First, two of the judges who decided the case were trained in non-domestic jurisdictions – Lord Steyn in South Africa and Lord Cooke in New Zealand – thus bringing to the case a different non-UK and non-US perspective.

Secondly, the case suggests that there is a rule of substance over form which is to be applied in tax cases while, thirdly, it emphasises that the principle being dealt with is one of statutory construction and not something else.

The second and third aspects of McGuckian are somewhat contradictory: if the question is one of statutory construction what is the scope of a rule about substance over form, which is a fact related rule?

It appears from MacNiven that the substance over form aspect of McGuckian does not exist so, while it is undoubtedly a feature of the McGuckian decision, I think we can put it into the extravagant remarks compartment: it is not necessary to the actual decision in McGuckian.

The fourth aspect of the McGuckian case is that it is not entirely clear what it actually decided. Did it decide that capital could be treated as income or did it decide that the taxpayer’s vehicle got income? I want to come back to this in a moment when I have considered the MacNiven decision in a bit more detail.


Although recent, it is not the latest word on the story because we have, since it was decided, had the decision in DTE Financial Services Ltd v Wilson [2001] BTC 159, but it is, I think, the most important case we have had on tax at least since Ramsay and Furniss.

Why is it so important?

It is for two reasons:

  1. Many of the remarks in the earlier decisions looked or sounded like the judicial creation of a principle of law which overrode the statute Lord Hoffmann has emphatically rejected the ability of the court ‘to impose an overlay upon the tax legislation’. The courts should not now attempt to legislate in tax cases, an interesting contrast with their willingness to do so in the common law field as exemplified by the restitution case of IR Commrs v Kleinwort Benson (1969) 45 TC 369.
  2. Partly as a consequence of the refusal to legislate, it has swept away all the clutter and has left us with a rule; and, although neither the content of the rule nor the result of the rule may yet be absolutely clear, the rule itself is clear.

There is no moral content to a tax case: as Rowlatt J said all those years ago ‘there is no equity about a tax’; and a taxpayer is entitled to arrange his affairs so that he pays the least possible tax.

Where the statute does not mention tax avoidance, the question of whether there has been tax avoidance is irrelevant to the analysis; and it does not matter whether the factual background can be described as a device or a stratagem or just as a transaction. None of this has any impact upon the analysis.

The only thing which has any impact is the wording of the relevant statute. This is the rule. It is all a matter of looking at the statute and seeing whether the taxpayer falls within it or without it.

So the case is important because it has swept away a lot of the clutter and all of the extravagance and taken us back to the right place, which is the wording of the statute itself.

It is, I think, even more important because of the instruction which it gives us as to how we should approach the issue of statutory construction.

Purposive approach

In every case it is necessary to ascertain with precision the question being posed by the statute; and, in determining what the statutory question is, a purposive approach is adopted to the construction of the statute.

This is, I think, the absolutely key and fundamental point. It is always necessary to identify the statutory question.

In Ramsay and Burmah the statutory question was: ‘has the taxpayer realised a loss?’

The analytical error made at the time the transactions were being considered in the 1970s was to think that the statutory question was ‘did the taxpayer make a loss on this particular asset?’ and to fail to realise that, implicit in that question, was the issue of whether the taxpayer made a loss at all.

A taxpayer can only make a loss on a particular asset if he suffers a loss: if he does not have a loss at all, he cannot realise a loss on a particular asset. The statute posed the question ‘does the taxpayer as a matter of fact have a loss?’; and the courts were able, by analysing the facts, to see that he did not.

In Ensign Tankers the question was: ‘Did this taxpayer incur expenditure?’; and a factual analysis enabled the court to say that it had not.

In Furniss the question was ‘to whom was this disposal made?’ and the court was again able to analyse the facts and find that there was a disposal to the ultimate purchaser, rather than to the intermediate company.

And it can be seen from an analysis like this that nothing in the cases up to McGuckian at least, infringes the rule laid down in the Duke of Westminster which is that there can be no re-characterisation by reference to substance.

What we learn above everything else from MacNiven is that identification of the statutory question is absolutely key.
I do not think MacNiven is a very useful guide when it comes to identifying what the statutory question is in each case: indeed, in one respect at least, I think it may be rather unhelpful on that point. What is quite clear is that we do need in every case to identify the correct statutory question.

All that MacNiven actually decides is that, as Westmoreland discharged its debt for interest, it had a charge on income for the purposes of Income and Corporation Taxes Act 1988 (ICTA 1988), s. 338, because it paid the interest.

That is all that it actually decides: it decides something, now no longer relevant, in relation to ICTA 1988, s.338. It does not decide anything more and it does not decide anything less. It does not tell us very much about how that decision was arrived at. I think everybody recognises that, using the same analytical theory as expounded in MacNiven, the conclusion that Westmoreland had not discharged its debt could have been arrived at just as easily as the conclusion that it had.

Therefore, I do not think that MacNiven is a very good guide to results in future cases. It tells us what approach to take and it is highly important for that reason, but it does not actually point to the outcome of any future case. I should make six points here:

  1. I suspect that the most important sentence in the whole judgment is Lord Hoffmann’s remark that the Crown made it unnecessarily difficult for itself in McGuckian ‘by failing to notice that the question was different from that in Furniss … and therefore did not necessarily respond to precisely the same analysis’.

    Different cases will raise different questions. Different questions require different analyses.

  2. Although I have emphasised the absence of moral content in the making of any analysis it is, perhaps, worth noting that it is not wholly absent from this decision itself. The House of Lords has said that the feature objectionable to the Revenue was the ability of the pension scheme to recover tax; but I rather doubt if that is what was objectionable to the Revenue.

    What was objectionable to the Revenue was that Westmoreland was put in a position to claim tax relief while being, economically, in exactly the same position as it was in when it could not obtain tax relief.

    However, a point to note here is that Westmoreland really did have a loss. It really had lost money so that MacNiven is not a case where relief was being obtained for something unreal.

  3. In some ways, MacNiven has created rather than eliminated uncertainty. Before MacNiven it appeared that a particular set of circumstances had to exist as a matter of fact before a Ramsay approach could be adopted. There had, for example, to be pre-ordination – and this is no longer the case.

    A rule like that – that certain facts had to exist before Ramsay applied – was obviously inconsistent with an approach which gives the guiding role to the meaning of the statute and looked uncomfortably like judicial law making.

    However, the removal of any such rule means that the facts alone cannot now determine the applicability or non-applicability of the Ramsay approach.

    The removal of this apparent rule coupled with the possibility of different questions and different analyses creates scope for future development.

    I have referred to the removal of the rule: I should, perhaps, have referred to its partial removal because in some cases the statute may ask whether there has been preordination and, in cases like that, preordination and the learning on it will remain relevant.

  4. There is a good deal in MacNiven about the distinction between legal and commercial concepts.

    But I do not believe that this helps to remove the uncertainty created by the decision and I doubt if this distinction is going to be important in other cases.

    The court says the point was important in MacNiven, but I have to say I am not quite clear why: even if commercial concepts were relevant wouldn’t a businessman – if told that Westmoreland had discharged its debt for interest – say that Westmoreland had paid its interest? Some commercial concepts are dependent on or interlinked with legal ones.

    The point which I have made and that I want to emphasise is that the actual decision is only that Westmoreland had paid its interest given the terms of ICTA 1988, s. 338. The reasoning as to why Westmoreland is treated as having paid its interest belongs in the commentary category not in the class of the truly important. While there is much commentary on it, it would be a mistake to focus on the distinction between legal and commercial concepts when the real issue is ‘What question is being posed by the statute?’

  5. Because it is a mistake to pay too much attention to the distinction between legal and commercial concepts, I do not myself think that the DTE Financial Services case in the Court of Appeal is incompatible with MacNiven.

    I do think that there are a number of objections to that decision, partly because I think the interpretation which it puts on the word ‘payment’ is inconsistent with the provisions about ‘trading arrangements’ and partly because I think that the court’s approach to the application of Ramsay principles is somewhat out of line with the approach indicated by MacNiven itself.

    This in a way goes to illustrate how the MacNiven concepts can create uncertainty. Indeed, any purposive approach inevitably brings uncertainty, because the perceived purpose of a statute will vary with the degree of knowledge and familiarity of the person doing the construing.

    But I fully accept that the statutory question asked in the DTE case is: ‘was there a payment by the employer to the employee?’ and for my own part, on the facts of the case, I do not think it too difficult to say that the answer was yes in the context of the particular statutory provision in issue.

  6. As I have been at some pains to expound, I believe that its importance lies in emphasising the paramount role of the statutory question and, on the whole, I do not have too much difficulty with the way in which the statutory question has been identified by Lord Hoffman in relation to MacNiven itself or in relation to the earlier cases, save for McGuckian.
    I do, however, have a little difficulty with the way in which Lord Hoffman has identified the statutory question in McGuckian.

Lord Hoffman has said that the statutory question in McGuckian was: ‘was this receipt income or capital?’ and the answer to the question was that it was income, because the sale of the dividend did not work the alchemy of turning income into capital.

I have difficulty with this if the statutory question really is ‘was this receipt of sale proceeds income or capital?’ because, if that is the right question, it is necessary to look at the receipt itself as sale proceeds and to characterise it as income or capital.

On its face the receipt being sale proceeds of a right to a dividend is capital; and it follows that a conclusion that the receipt is income involves treating the sale proceeds as a dividend. This is to re-characterise the receipt in a way which is prohibited by the Duke of Westminster’s case, the paramountcy of which has been re-affirmed yet again by MacNiven.

Now if that is what McGuckian has decided – that sale proceeds can be taxed as income – it does more than interpret the statute: it would have gone so far as to re-characterise what is essentially capital as income; and this would go further than any case before McGuckian and further than MacNiven suggests it is permissible to go.

The problem I think is caused because Lord Hoffman had slightly mis-stated the statutory question posed in McGuckian. I do not think the question was ‘Is this receipt capital or income?’ I think it was ‘Who received the dividend?’ On the facts of the case, the answer to that question was the apparent vendor because, as a matter of fact, it was the person who really got the dividend.

Somewhere in Four Quartets, T S Eliot has some lines to the effect that in the end is our beginning.

MacNiven has got rid of a lot of extravagant stuff and it has taken us back to the paramount role of the statute, while emphasising the modern approach of purposive construction, which is not new, but which is more greatly emphasised today than it once was.

Is it the end of the journey to which Lord Scarman referred in Furniss v Dawson?
The answer is no for two reasons.

  1. As Lord Hoffman, with a, to me, welcome acidity points out, we are not always on the same journey. Indeed, I suspect each case takes us on a different journey so that, just as you cannot step into the same river once, you cannot end the journey more than once.
  2. More crucially, I think we still have a great deal to learn about how to identify, in the context of modern and elaborate transactions, what the statutory question is.

But I am quite sure that correct identification of the statutory question is critical to the analysis of any tax case and clarification of that point is a significant analytical advance.

Some years ago I had the privilege – it was, I think, then a privilege – of advising the Conservative Party and, afterwards, Mrs Thatcher wrote to me to thank me for my help in what she called ‘the battle against the Inland Revenue’.

Some of us think of tax practice as a war, if not of arms at least of wits, and so it may be appropriate for me to finish with Mr Churchill’s famous remark after the Battle of El Alamein had been won by General Montgomery:

‘Now, this is not the end. It is not even the beginning of the end. But it is, perhaps, the end of the beginning.’

Technical Department
020 7235 9381

div class="taslug">July 2002 by


We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we'll assume that you are happy to receive all cookies on the The Chartered Institute of Taxation website. To find out more about the cookies, see our privacy policy.