Judgments - Jones (Respondent) v.Garnett (Her Majesty's Inspector of Taxes) (Appellant)

(back to preceding text)

    36.  In Young v Pearce (1996) 70 TC 331 the share capital was divided into two classes: 50 preference shares of £1 each, of which 25 were allotted at par to the wives of each of the two taxpayers, and 50 ordinary shares of £1 each which divided equally between the taxpayers themselves. The preference shareholders were entitled to 30% of the net profits for any year in which the profits of the company were to be distributed, the balance of the distributed profits to be paid to the ordinary shareholders. The articles provided that in the event of liquidation the preference shareholders were to be entitled to repayment of the sums subscribed for their shares, but no more. They were entitled to attend and to speak but not to vote at general meetings of the company.

    37.  In the light of those facts Sir John Vinelott said, at p 346, that the property given in the form of the preference shares was wholly or substantially a right to income. This was because the preference shares entitled the holders to a preferential dividend if the directors decided that the whole or part of the profits arising in any given year were to be distributed and because, apart from that right to income, the only rights were to repayment of the nominal sum paid on the allotment of the shares and the right to attend but not to vote at general meetings of the company. It would not have been accurate to say that the rights attached to the preference shares were wholly a right to income, because of the right to repayment of capital. But there is no doubt that, on the facts of that case, the rights were "substantially" a right to income. That is because there was no right under the articles, after repayment of capital, to participate in any other assets of the company.

    38.  The position would have been different if the shares in question in Young v Pearce had been ordinary shares. The rights which an ordinary shareholder enjoys are not confined wholly, or even substantially, to a right to income. The residue of the assets of the company belongs to the ordinary shareholders, after the rights of creditors and of any preference shareholders have been satisfied. So property given which consists of ordinary shares in a company will always attract the exception in section 660A(6).

    39.  It is true that Arctic Systems Ltd had no assets when the formation agents sold the two ordinary shares of £1 each at par to Mr and Mrs Jones. But the question whether the property given to Mrs Jones was wholly or substantially a right to income cannot depend on the state of affairs of the company at time of the gift. Parliament cannot have intended that the exception in section 660A(6) would apply only if - to take shares as an example - the settlement was made when the company had built up sufficient surplus assets to be distributed among the ordinary shareholders as capital. The critical words are "a right to income". It is the rights attached to the asset comprised in the settlement, not the product of their exercise from time to time, now or in the future, that determine whether the exception applies to it.


My Lords,

    40.  I have had the advantage of reading in draft the opinions of my noble and learned friends Lord Hoffmann and Lord Neuberger of Abbotsbury. I agree with them, and for the reasons which they give I consider that the taxpayer succeeds on the section 660A(6) issue, but not on the wider issue of "arrangement". Because these points are of general interest I add some observations of my own. I cannot usefully add to what my noble and learned friends say about "bounty" but I wish to address some other problems arising on the statutory definition of "settlement" as including an "arrangement".

    41.  In IRC v Plummer [1980] AC 896, 911-912, Lord Wilberforce referred to the statutory origins of Part XVI of the Income and Corporation Taxes Act 1970:

    "[Part XVI] includes a number of provisions which have been enacted at different times, the general effect of which is to cause income of which a person has disposed in various ways to be treated, in spite of the disposition, as the income of the disposer. These had been successively enacted in the Finance Acts 1922, 1936, 1938, 1946, inter alia, with increasing severity."

    The same is true of sections 660A and 660B of the Income and Corporation Taxes Act 1988 as amended by the Finance Act 1995.

    42.  The very first of these enactments, section 20 of the Finance Act 1922, was aimed at revocable and short-term dispositions of income, and primarily (though not exclusively) at gratuitous covenants to make annual payments. Before 1922 sums payable under such covenants were treated as charges on (and so deductible from) the payer's taxable income, even though the covenant did not effect any disposition of property, and created no more than a personal obligation. Section 20(5) provided,

    "the expression 'disposition' includes any trust, covenant, agreement or arrangement."

    A comparable definition of "settlement" ("includes any disposition, trust, covenant, agreement, arrangement or transfer of assets") appeared for the first time in subsection (9)(b) of section 21 of the Finance Act 1936 (provisions as to income settled on children), which was capable of applying to covenanted payments as well as to the income of settled property. Covenants continued to have some efficacy for tax purposes (though subject to increasingly stringent conditions) until a general abolition of the relief (for surtax purposes) by section 12 of the Finance Act 1965. The general rule now (for all income tax purposes, though subject to some limited exceptions) is that in order to escape a tax charge the income disposed of must be income from property of which the disponer has entirely divested himself.

    43.  So the very wide definition of "settlement" in section 660G(1) of the Income and Corporation Taxes Act 1988 (as amended) has a long and fairly complicated pedigree. It is in striking contrast to the definition of "settlement" which applied (through its 80-year life) under the statutory code charging estate duty. Section 22(1)(i) of the Finance Act 1894 defined the term referentially, initially by reference to section 2 of the Settled Land Act 1882: a "classic" settlement, in the convenient expression used in Chinn v Hochstrasser [1981] AC 533. A definition of that sort was appropriate for the purposes of estate duty, which was a mutation tax charged on the capital value of property.

    44.  The progressive elimination of covenanted payments as a means of tax avoidance has in this respect (only) brought the income tax definition of "settlement" a little closer, in its practical application, to the classic meaning. But trying to apply the statutory definition to an arrangement which includes (but is not limited to) a classic settlement can produce a sort of blurred double vision. In Chamberlain v IRC (1943) 25 TC 3 , for instance, the court (in applying section 38(2) of the Finance Act 1938) had to identify "the property comprised in the settlement" and to decide whether that settlement ("or any provision thereof") was revocable, and unsurprisingly encountered difficulties in doing so: see in this House the speeches of Lord Thankerton (with whom Viscount Simon LC and Lord Atkin agreed) at 329, Lord Macmillan at 331 and Lord Romer at 333. That case may be contrasted with IRC v Payne (1940) 23 TC 610, a decision on section 38(1) of the Finance Act 1938, where there was no classic settlement of property, only an income covenant in favour of a controlled private company.

    45.  The most striking example of what I have referred to as double vision is Chinn v Hochstrasser [1981] AC 533, in which this House had to consider section 42 of the Finance Act 1965. That section, long since repealed, was the earliest anti-avoidance provision aimed at offshore trusts as a means of avoiding capital gains tax. By the section Parliament lifted the wide income tax definition of "settlement" and applied it to one category of classic settlements (those with settlors based in the United Kingdom, but trustees resident outside the United Kingdom), possibly with insufficient consideration of how this hybrid provision would work.

    46.  In this case there is no classic settlement, but the exception for an "outright gift" in section 660A(6) raises a similar problem as to identifying the constituents or components of the arrangement to which (as I would hold) Mr Jones was a party.

    47.  The inclusion in the statutory definition of the very wide word "arrangement" shows that Parliament recognised, as long ago as 1922, that a wealthy taxpayer might be advised to dispose of what would otherwise be his taxable income by relatively complicated or artificial means. These might include a classic settlement, especially when the intended beneficiaries were minor children, but even in that case a classic settlement was not essential (Copeman v Coleman [1939] 2 KB 484 is an early example). Other components of the arrangement might be the formation or acquisition of a private company with an unusual share structure, the declaration of abnormal dividends and the granting and exercise of options (as in Vandervell v IRC [1967] 2 AC 291) or entering service agreements on unusual terms (as in Crossland v Hawkins [1961] Ch 537).

    48.  An intention to avoid tax is not, I think, absolutely essential. It is possible to imagine that an arrangement planned for some other purpose (such as pre-empting the consequences of insolvency or divorce) could unexpectedly prove efficacious for tax avoidance and amount to an arrangement (and so to a settlement). But usually an intention to avoid or minimise tax can readily be inferred (in this case it was candidly admitted) and that intention is part of the factual material that has to be looked at in the round. Sir Wilfred Greene MR put it trenchantly in IRC v Payne at 626:

    "It appears to me that the whole of what was done must be looked at; and when that is done, the true view, in my judgment, is that Mr Walter Payne deliberately placed himself into a certain relationship to the company as part of one definite scheme, the essential heads of which could have been put down in numbered paragraphs on half a sheet of notepaper. Those were the things which it was essential that Mr Payne should do if he wished to bring about the result desired. He did it by a combination of obtaining the control of the company, entering into the covenant, and then dealing with the company in such a way as to achieve his object. Now, if a deliberate scheme, perfectly clear-cut, of that description is not an 'arrangement' within the meaning of the definition clause, I have difficulty myself in seeing what useful purpose was achieved by the Legislature in putting that word into the definition at all."

    49.  Some arrangements are planned in minute detail and carried out "with timetables, in almost military precision" (Lord Wilberforce in IRC v Plummer at 907). Highly artificial arrangements of that sort led to W T Ramsay Ltd v IRC [1982] AC 300 and the other well-known cases which came in its train (which your Lordships need not consider further on this occasion). But a high degree of complexity, artificiality and pre-planning is not essential in order to produce an arrangement. That is well illustrated by Crossland v Hawkins [1961] Ch 537 and Butler v Wildin (1988) 61 TC 666, both of which are covered at some length in Lord Hoffmann's opinion. Like Lord Hoffmann, I would adopt the passage in Donovan LJ's judgment in Crossland v Hawkins at 549, where he refers to "sufficient unity." The taxpayer's intention to minimise his tax liability by a "definite scheme, the essential heads which could have been put down in numbered paragraphs on half a sheet of notepaper" explains the rationale of the sequence of events, and gives it unity.

    50.  The Court has been reluctant to try to lay down any precise test for identifying the components of an arrangement or for assessing the "sufficient unity" to which Donovan LJ referred. Sometimes it has been content to conclude that wherever the boundary line is to be drawn, the taxpayer and his advisers have got themselves into forbidden territory (see for instance the Master of the Rolls in Payne at 626, already quoted; Lord Wilberforce in Chinn v Hochstrasser [1981] AC 533, 549, and Lord Diplock, dissenting, in IRC v Plummer [1980] AC 896, 924). In my opinion the Court's caution has been well-advised. "Arrangement" is a wide, imprecise word. It can ("like settlement" or "partnership", or indeed "marriage") refer either to actions which establish some sort of legal structure (in this case, a corporate structure through which the taxpayer's income could be channelled) or those actions together with the whole sequence of what occurs through, or under, that legal structure, in accordance with a plan which existed when the structure was established. The planned result may be far from certain of attainment. It may be subject to all sorts of commercial contingencies over which the taxpayer has little or no control. But if the plan is successful and income flows through the structure which he has set up, it is "income arising under the settlement."

    51.  That seems to be the approach taken in most of the authorities. In Crossland v Hawkins Donovan LJ said (at 550),

    "Bearing in mind the ultimate object of securing money free from the burden—or the full burden—of surtax, can it matter for present purposes that the precise way of securing this result was not decided upon at the very outset? I think not. An alternative way of looking at the matter would be this: Here the repayment claim is made in the year 1956-57. In that year the arrangement is complete, and that is enough. It would be irrelevant that it came into being by instalments in the year 1954-55. The Revenue looks at the facts of the year being taxed or for which repayment of tax is being sought, and asks in this year 'Is it true to say that there is a settlement of the kind mentioned in the section, and in this year is it true to say that the settlor has provided funds for the purpose of the settlement?'"

    In referring to "instalments" Donovan LJ was referring (see at 549) to the formation of Roehampton Productions Ltd, the service agreement of 10 December 1954, the settlement of 3 March 1955 and the transfer or issue (before 31 March 1955) of all the Roehampton shares to the trustees. He does not seem to have regarded Jack Hawkins' performance in 'Fortune is a Woman' or Roehampton's payment of a dividend (which occurred in 1956) as part of the arrangement, but as the arrangement being put to its intended use.

    52.  That approach is, I think, consistent with what Vinelott J said in Butler v Wildin (1988) 61 TC 666, 678. The point that Vinelott J was making was that the Special Commissioner had misunderstood the facts and misdirected himself by focusing on the date of incorporation of an off-the-shelf company whose shares the taxpayers did not acquire until over two months later. It was during that period that the taxpayers came across their business opportunity. That opportunity might have come to nothing (the judge, at 685, clearly thought it a very risky venture.) But when it did prove profitable the dividend income paid out by the company was income paid "by virtue or in consequence of" the statutory settlement constituted by the arrangement. I would add that in my opinion the wording of section 660A(1) ("income arising under a settlement") does not impose a more demanding test, and may impose a rather less demanding test of causal connection, than the expression "by virtue and in consequence of."

    53.  I have gone into these points in some detail because they do to my mind have a bearing on the "outright gift" issue. It has been said that it is necessary to identify "the arrangement": Vinelott J said that in Butler v Wildin (at 684), and the Master of the Rolls said much the same in IRC v Payne (at 626) nearly fifty years before. Normally (there may be exceptions) the arrangement is to be identified by the constituent parts or components of the legal structure designed for a purpose, and not by what is done (sometimes months or even years later) in using the structure for its intended purpose.

    54.  The Revenue argued that the arrangement entered into by Mr and Mrs Jones included, but was larger than (and so different from) the establishment of the original corporate set-up under which each had half of the issued share capital of Arctic Systems Ltd ("Arctic"). I do not accept that argument. There was no written service agreement between Arctic and Mr Jones comparable to the service agreement between Roehampton and Jack Hawkins. The establishment of the corporate set-up, together with the common intention that Mr and Mrs Jones would use it to minimise tax in accordance with their accountants' advice, was the essential arrangement. What happened afterwards was that the arrangement was put to its intended use.

    55.  Mr Jones did not actually make a transfer by way of gift to his wife of one of the two issued shares in Arctic. She bought it at par from the company formation agents. But it was not the sort of arrangement that would have been made between strangers dealing with each other at arm's length. Arctic was the chosen vehicle through which Mr Jones was to offer his valuable services as an IT consultant, and it was an act of bounty on his part to permit his wife to acquire half its equity for the nominal sum of £1. In my opinion that amounted to an outright gift of the share within the meaning of section 660A(6). I respectfully disagree with Park J's contrary conclusion because I think he took too expansive a view of the scope of the statutory settlement. I prefer the tentative view expressed by Sir John Vinelott in Young v Pearce (1996) 70 TC 331, 346.

    56.  I have found the condition in section 660A(6) (b) "the property given is wholly or substantially a right to income" rather more difficult. The property given was an ordinary share—in fact, half the issued ordinary share capital of Arctic—and so it was certainly not "wholly . . . a right to income." If the plan worked it could be expected to produce a healthy dividend income but not to attain any significant market value (it would hardly be marketable at all on the basis that Arctic was a going concern, since Mr Jones could not be expected to continue to work under an arrangement which channelled nearly half his earnings to a stranger). But at the outset there was at least the possibility that Arctic would build up a reserve of undistributed income, and the agreed statement of facts and issues suggests that this occurred (because of IR35) during 2000-1 and 2001-2. The decision of Sir John Vinelott in Young v Pearce (1990) 70 TC 331 is distinguishable because of the very unusual rights conferred on the preference shares (under new articles of association which were part of the arrangement) in that case.

    57.  For these reasons, and for the further reasons given by Lord Hoffmann and Lord Neuberger, I would dismiss this appeal.


My Lords,

    58.  Income tax was invented only decades after Blackstone (Commentaries, Book 1, p 442) had given his classic definition of the relationship between husband and wife at common law:

    "By marriage, the husband and wife are one person in law; that is, the very being or legal existence of the woman is suspended during the marriage, or at least is incorporated and consolidated into that of her husband; under whose wing, protection, and cover, she performs everything; and is therefore called in our law-French a feme-covert . . . "

At that date, a husband was absolutely entitled to his wife's earnings. He was not entitled to the income from property which had been settled to her separate use. Nevertheless, from its very beginnings the tax system decided to treat all her income, earned and unearned, as her husband's income. He was liable to declare it to the Revenue and he was liable to pay the tax on it. The 1799 Act which first introduced income tax provided that a married woman's income was to be stated and accounted for by her husband. In 1806, it was provided that the wife's profits should be deemed to be those of her husband. This remained the basic rule until 1990.

    59.  No doubt this was a convenient solution, avoiding the sort of difficult questions which this case has raised. It was so convenient that it continued after the Married Women's Property Act 1870, when wives became entitled to keep their earned incomes, and after the Married Women's Property Act 1882, when all a wife's property and income from any source was deemed to be her separate property. In some cases it also raised more tax, for example where the couple's combined incomes made them liable to surtax or to higher rates of tax to which they would not have been liable if separately taxed. In others, the effect might be offset by the higher married man's allowance (introduced in 1918) and the wife's earned income allowance (introduced in 1920).

    60.  Two Royal Commissions thought that the aggregation rule was not only convenient but principled: the taxable capacity of a married couple depended upon the total income coming into the household and not upon the chance of how it was owned between them (Royal Commission on the Income Tax, 1920, Cmd 615, para 259; Royal Commission on the Taxation of Profits and Income, Second Report, 1954, Cmd 9105, para 120). Others thought that it was inconsistent both with the principles of separate property and of equality between husband and wife. It was abandoned in stages.

    61.  The possibility of separate assessment was introduced in 1914. Either spouse could apply to be separately assessed. This did not affect the principle of aggregation or the total amount of tax payable. This was simply apportioned between them and the wife became solely liable for the part apportioned to her. She could, if she wished, complete her own tax return and thus maintain a degree of privacy from her husband.

    62.  The next stage, introduced in 1971, was to allow the spouses jointly to elect that the wife's earned income be separately taxed. This reduced their personal allowances to those of two single people but meant that they were less likely to reach the higher rates of tax. The husband remained liable to be taxed on the wife's unearned income. He also remained responsible for completing their tax returns unless either had also applied for separate assessment. The Revenue practice of writing to the husband about his wife's tax affairs (even if she had written to the Revenue) was only changed in 1978 (see The Taxation of Husband and Wife, 1980, Cmnd.8093, para25).

    63.  The final stage, reached when the relevant provisions of the Finance Act 1988 came into force in 1990, was to abolish the aggregation rule altogether and treat them as two separate individuals.

    64.  The provisions which we have to construe were part of the same package. The green paper on The Taxation of Husband and Wife (1980, Cmnd 8093, paras 50 and 88) had remarked (echoing the observations of the 1954 Royal Commission, Cmd 9105, para 119):

    ". . . there would be a potential tax advantage to be gained by transferring income from the better-off husband (or wife) to the other spouse. With earned income the opportunity for splitting income in this way hardly arises, except perhaps where the wife works for her husband. With investment income, however, the opportunities would clearly be much greater. In so far as this could lead to a more even distribution of the underlying capital between husbands and wives, this may be regarded as no bad thing. . . . But in any case . . . it would be necessary to give consideration to the prevention of artificial methods of reducing the tax bill by transferring income from husband to wife or vice versa."

    65.  Hence the new provisions in section 660A were closely modelled on those designed to stop fathers from taking advantage of the fact that their minor children were treated as separate individuals for income tax purposes by settling income-producing property upon them. The aim was to prevent artificial transfers of income from father to child or husband to wife. But, consistently with the policy of exempting outright transfers of property between spouses from capital transfer tax and inheritance tax, an exception was made for outright gifts. The policies are easy to state. But it is not easy to translate them into statutory language which exactly captures the promoters' intent. It is some comfort that the professional judges who have so far decided this case have reached such different conclusions and that we have reached different conclusions from them all.

    66.  I agree, for the reasons given by my noble and learned friends Lord Hoffmann, Lord Hope of Craighead, Lord Walker of Gestingthorpe and Lord Neuberger of Abbotsbury, that if this was a "settlement" for the purpose of section 660A(1) of the Income and Corporation Taxes Act 1988, it fell within the exception for "an outright gift" in section 660A(6). There is no need for me to say anything more about that. But I confess to having had rather more difficulty than they in concluding that this was a "settlement" at all.

    67.  When a husband and wife team set up a family business through the machinery of a private company in which they each have an equal number of shares, they may have clear expectations of what the future will hold or they may not. If combined with low salaries and high dividends, as things stand at present, the arrangement will always result in lower national insurance contributions and lower income taxation. But as I understand it, the Revenue is anxious to catch only one of the following examples: (i) they may expect that each will make a contribution to the company's earnings of roughly equal financial value; (ii) they may expect that each will make a contribution which is equal in terms of effort but (not least because of historic discrimination between the price of different kinds of services in the market place) unequal in terms of earnings for the company; (iii) they may expect that each will contribute what they can but that those contributions will vary over time, perhaps because of personal factors such as illness or child rearing, perhaps because of changes in the business and its market; or (iv), as here, they may expect that one will contribute the work which brings in the money from outside while the other will contribute the limited but necessary ancillary services to make that work possible but bring in no independent money from outside. There are many variations and permutations between these possibilities.

Continue  Previous