Consolidated Investment and Enterprises Limited
Appellant
v.
The Commissioner of Income Tax
Respondent
Appeal from the Supreme Court of Mauritius
Composition of the Board:
Lord Goff of Chieveley
Lord Browne-Wilkinson
Lord Lloyd of Berwick
Lord Nicholls of Birkenhead
Sir Michael Hardie Boys
Judgment delivered on the 19th February 1996
by Lord Browne-Wilkinson
______________________________________________________________
(1) Administrative law - Taxation - Dividends - Exempt income - Allowable deductions - Allocation of interest charges
(2) Privy Council jurisdiction - New course of argument before the Judicial Committee - Various questions on which parties desire decisions in principle from the Judicial Committee
___________
Legislation referred to in judgment
Income Tax Act 1974, sections 2, 4, 7, 19, 20, 26, 55, 57
The following judgment was delivered by the Board:
This is an appeal by Consolidated Investment and Enterprises Limited ("the taxpayer company") against a decision of the Supreme Court of Mauritius (Glover C.J. and Yeung Sik Yeun J.) dismissing the appeal of the taxpayer company from the dismissal by the Tax Appeal Tribunal of an appeal from a determination of the respondent, the Commissioner of Income Tax ("the Commissioner") claiming additional tax of Rs680,074 in respect of the year 1989/90. The appeal is unusual since the arguments advanced before their Lordships were materially different from those urged before the courts below which, it is now accepted, cannot be sustained for the reasons given by them. Their Lordships were asked by both parties to determine certain new issues in principle. Since the taxpayer company and other companies in Mauritius are affected by the issues now raised, their Lordships agreed to adopt that course.
The taxpayer company is an investment and holding company for a group of 28 subsidiaries. In the relevant year, some subsidiaries paid dividends, others did not. The income of the taxpayer company for the year in question came from three different sources, viz.
(1) dividends paid by subsidiaries, which dividends constituted "exempt income" in the hands of the taxpayer company;
(2) dividends paid by subsidiaries out of the subsidiaries' profits, which profits had been charged to tax in the hands of the paying company ("A dividends");
(3) other income.
In earning such income, the taxpayer company incurred management expenses and interest charges on borrowed capital.
Before turning to the statutory provisions directly in issue, namely section 55(2) of the Income Tax Act 1974, it is convenient to set out how that Act approaches the taxation of a company's profits apart from that section.
Section 4 provides that income tax is assessable and payable on all income "other than exempt income". "Exempt income" is income of a type specified in section 7 of the Act. It is common ground that the taxpayer did receive income of that description. Such exempt income is not central to the main issue which arises on this appeal, and is to be sharply distinguished from the A dividends to which their Lordships will refer later.
Section 4 further provides that income tax shall be calculated on the taxpayer's "chargeable income" which is defined by section 2 in relation to a company taxpayer to mean the amount of income determined in accordance with Part VI of the Act. Section 55(1) (which is in Part VI) provides that "the chargeable income" of the company is the "net income" of the company for the relevant purpose. "Net income" is defined by section 2 to mean the amount remaining after deducting from the gross income "allowable deductions".
"Allowable deductions" are defined as meaning the sums deductible under Part IV of the Act. Under section 19(1) any expenditure exclusively incurred in the production of "gross income" is deductible. But section 20(1)(c) expressly provides that expenditure incurred in the production of "exempt income" is not deductible. As to interest charges, section 26(1) provides, so far as relevant, that expenditure on interest shall not be deductible except "interest... on capital employed... for the production of gross income".
Still ignoring for the time being section 55(2), the impact of these provisions on the income of the taxpayer company in the relevant year would be as follows:-
(1) The dividends which are "exempt income":
Since "exempt income" is not included in "gross income" it is not included in the net income of the company which under section 55(1) is the income chargeable to tax.
(2) The A dividends:
These are the dividends paid to the taxpayer company out of profits of the subsidiaries which have already borne tax. As will be seen sub-section (2) of section 55 is expressly directed to such A dividends. But, apart from the provisions of that sub-section, the A dividends received by the taxpayer company would form part of the taxpayer company's gross income and, subject to any allowable deductions, would form part of the company's net income chargeable to tax.
(3) Other income:
Other income of the company would be chargeable under section 55(1), subject to allowable deductions, as part of the taxpayer company's chargeable income.
As to allowable deductions, no management or interest charges incurred in earning the exempt income would be deductible: section 20(1)(c). The management and interest charges preferable to the A dividends would be deductible in computing the taxpayer's chargeable income.
Therefore, apart from further provision, the taxation of the A dividends in the hands of the recipient of the A dividends (i.e. the taxpayer company) would involve an element of multiple taxation. If the subsidiary pay the dividends out of the subsidiary's taxed profits, the same profits will be taxed twice, once in the hands of the subsidiary paying the dividend and again in the hands of the recipient, the taxpayer company.
There is no doubt that section 55(2) is directed to eliminating this multiple charge to tax. It provides as follows:-
"(2) (a) Where in any income year a resident company pays dividends to another resident company, including a company to which section 57 applies and the dividends are paid out of income of the company which has been charged to tax for any year of assessment beginning with the year of assessment 1984-85, the company receiving the dividends shall not, to the extent that the dividends received by it are not directly or indirectly attributable to the allowable management and interest expenses incurred by it, be chargeable to income tax on such dividends.
(b) Where a resident company, including a company to which section 57 applies, has received dividends which are not chargeable to income tax under paragraph (a) and that company pays dividends to another resident company, the company receiving the dividends shall not, to the extent that the dividends received by it are not directly or indirectly attributable to the allowable management and interest expenses incurred by it, be chargeable to income tax on such dividends.
(c) Where a resident company receives dividends form another resident company and only part of those dividends received is not chargeable to income tax under paragraph (a) or (b), the company receiving the dividends shall be liable to income tax on the excess.
(d) (i) Subject to subparagraph (ii), the amount of dividends chargeable to income tax for the purposes of paragraphs (a) and (b) shall be the lower of the allowable management and interest expenses incurred or a percentage thereof determined as follows -
dividends received (exclusive of the excess under paragraph (c))
100x
net income from all sources before deduction of management and interest expenses
(ii) Where for a trading company, the percentage is subparagraph (i) is 25 per cent or less, the dividends received (exclusive of the excess under paragraph (c)) shall be regarded as not chargeable to income tax.
(e) ..."
Three questions arise on these provisions which their Lordships will deal in turn.
A. How does section 55(2) work ?
It is common ground that the purpose of section 55(2) is to prevent the taxation of the A dividends in the hands of the taxpayer company if, but only if, those dividends have borne tax in the hands of the paying subsidiary i.e. have been paid out of the taxed income of the subsidiary. It achieves this first purpose by the words "where in any income year a resident company pays dividends to another resident company... and the dividends are paid out of income of the company which has been charged to tax... the company receiving the dividends shall not... be chargeable to income tax on such dividends".
If the relief had been granted in full on the whole of the A dividends, the relief would have been too great. The management and interest expenses incurred by the taxpayer company in earning the A dividends would be allowable deductions and be set against the other gross income of the company. The draughtsman could have conteracted this by providing that such expenses referable to the A dividends should not be deductible. But unhappily he sought to meet the point by clawing back the amount of the expenses so allowed as deductions by limiting the amount of the A dividends to which relief was to be afforded. The draughtsman achieved this second, claw back, purpose in paragraphs (a) and (b) of section 55(2) by providing that the relief is to be only "to the extent that the dividends received by it are directly or indirectly attributable to the allowable management and interest expenses incurred by it". That phraseology, although not very happy, is at least comprehensible. As was common ground, the amount of dividends payable out of the taxed profits of the subsidiaries which would otherwise benefit from the relief in section 55(2) is to be limited so as to exclude an amount of dividends equal to the management expenses and charges referable to the A dividends paid out of income. The part of the dividend so excluded from the relief given by section 55(2) will therefore remain within the general charge to tax.
However, paragraph (d) of subsection (2) produces chaos. At first sight, paragraph (d) sets out the formula by reference to which one has to calculate the amount of the A dividends which are not relieved from tax by paragraphs(a) and (b) but remain chargeable to tax under paragraph (c): "the excluded dividends". But as a matter of language it is difficult to give subparagraph (d) such an effect. If paragraph (d) sets out a formula for calculating the "excess" chargeable to tax under (c), how is one to work the formula in paragraph (d) ? The words in brackets in the numerator of the fraction in paragraph (d) "(exclusive of the excess under paragraph (c))", require one to have discovered what is "the excess under paragraph (c)" before one can work out the formula in paragraph (d). Given that difficulty, the first question raised on the appeal in how section 55(2) is to operate in practice.
Mr. Baker, in his admirable submissions for the taxpayer company, accepted that the common sense of the matter suggests that paragraph (d) simply provides a machinery for calculating the excluded dividends i.e. that part of the A dividends which are not to be relieved from tax under paragraphs (a) and (b) so as to reflect the fact that expenses incurred in earning A dividends remain allowable deductions against the gross income of the company. But, he submitted, the words of the statute do not permit that construction. He therefore submitted that paragraph (d) operates as a separate provision, affording an exemption from tax for A dividends additional to that afforded by subparagraphs (a) and (b).
In outline he submitted that the section operates as follows. At stage (1), it is necessary to identify the excluded dividends by construing the words between the commas in paragraph (a) i.e. "the dividends received by it" which are "directly or indirectly attributable to the allowable management and interest expenses". Such identification of the excluded dividends is achieved by identifying the allowable management and interest expenses incurred in the production of the A dividends. The excluded dividends are equal to the amount of those expenses. Such allowable expenses are to be a rateable allocation of the expenses i.e. the proportion of the company's total allowable management and interest expenses corresponding to the ratio between the total A dividends and the gross income of the company. After this has been done, one has determined what is "the excess" for the purposes of paragraph (c) and it is possible to proceed to stage (2) by applying paragraph (d). The latter paragraph must now be understood to operate as a further limit on the amount of dividends chargeable to income tax for the purposes of paragraphs (a) and (b). The further allowable management and interest expenses incurred in the production of A dividends or a percentage of such expenses computed by the formula. Applying the formula one then reduces "the excess under (c)" by a further amount represented by the operation of the formula in (d).
This is a most ingenious construction but one which their Lordships reject. First, it produces a wholly irrational result, affording to the taxpayer double relief from tax on the part of the expenses incurred in the production of the A dividends. On Mr. Baker's argument, the whole of the expenses incurred in the production of the A dividends are not clawed back and brought back into tax. Under his construction the total expenses referable to the production of the A dividends have been excluded from relief at stage 1 and form part of "the excess under paragraph (c)". At stage 2 he then operates the formula under paragraph (d) so as further to reduce the amount of expenses excluded from relief under paragraphs (a) and (b). Thus Mr. Baker's construction fails to achieve what, it is common ground, was intended to be achieved, namely the claw back into tax of the amount of the expenses referable to the production of the A dividends.
Mr. McCall, in his cogent argument for the Commissioner, has demonstrated that the language of section 55(2) does not force their Lordships to accept Mr. Baker's construction. As Mr. McCall points out, in relation to any given A dividend received by the taxpayer company there are two factors which may reduce the amount of the dividend relieved from tax: first, the proportion, if any, of such a dividend which is paid out of capital and has therefore not borne tax in the hands of the paying company; second, the expenses of earning the A dividend. Mr. McCall submits that the reference in paragraph (c) to "the excess" is a reference only to the part of the dividend which is excluded from relief by reason of it having been paid out of capital and therefore not borne tax in the hands of the paying subsidiary. If this is correct, there is no difficulty in operating paragraph (d), since the "excess under paragraph (c)" does not refer to the amount by which the relief is to be reduced so as to claw back the amount allowed for expenditure but refers only to the capital element in any A dividend.
Although Mr. McCall's construction involves some distortion of the language of the section, their Lordships are satisfied that it is correct. Not only does it produce the only common sense result but any other construction is impossible to reconcile with paragraph (d)(ii). Paragraph (d)(ii) applies when the operation of the formula in paragraph (d)(i) produces the result that less than 25% of the allowable expenses are attributable to the A dividends; in that case the A dividends "exclusive of the excess under paragraph (c)" are wholly relieved from tax by section 55(2). This de minimis provision requiring management and interest charges to be lest out of account and the whole dividend (other than the capital element) afforded relief from tax simply cannot operate if the dividend relieved from tax is to exclude the excess identifies by paragraph (c) as including a deduction on account of such charges. This factor demonstrates that the words "the excess under paragraph (c)" refer only to that part of the A dividend which is not relieved from tax because it was not paid out of income.
Section 55(2) therefore operates in a logical and explicable manner which both relieves the A dividends from double taxation (to the extent that it has borne tax in the hands of the paying company and claws back only such part of the management expenses as are referable to the A dividends. The A dividend is relieved from tax subject to two deductions, viz.
(1) that part of the dividend which was not paid out of profits which have borne tax in the hands of the paying company; and
(2) an amount equal to the management expenses and interest charges referable to the A dividend.
The amount at (2) is to be calculated in accordance with paragraph (d)(i) being either the whole of such expenses (e.g. if there is no other income of the recipient company against which the expenses can be set) or, if lower, the proportion of such expenses corresponding to the ratio between the total amount of the A dividends (excluding the capital element) and the total chargeable income of the company before deduction of management and interest expenses.
B. Exempt income.
The operation of section 55(2) requires one to identify what are "the dividends" referred to, what is concluded in the words "net income from all sources" in the denominator of the fraction in paragraph (d) and what are "the allowable management and interest expenses" referred to in paragraphs (a), (b) and (d). The Commissioner contends that exempt income is included in "the dividends" and "net income from all sources" and that allowable management and interest expenses include the expenses incurred in producing "exempt income".
Their Lordships have no hesitation in rejecting the Commissioner's submissions. As has been said "gross income" is defined so as to exclude exempt income; accordingly exempt income can form no part of the "net income" of the taxpayer company for any purpose of the Act. Moreover, management and interest expenses incurred in producing "exempt income" are expressly made non-deductible. Therefore exempt income is excluded from chargeable income of a company brought into tax by section 55(1). Subsection (2) provides a relief from tax in relation to income otherwise chargeable to tax. Therefore, prima facie references to "dividends" and income in subsection (2) will not include exempt income. Although the word "dividends" is not defined so as to exclude dividends which constitute exempt income, the whole structure of the relief afforded by subsection (2) of section 55 shows that exempt income cannot be relevant for any purpose of the subsection.
Say that one half of the dividend constituting exempt income of the recipient taxpayer company was paid out of the capital of the paying subsidiary company. Were it not for section 55(2), the whole of such dividend would not be subject to tax because the whole dividend would be exempt income under section 7. But if the Commissioner's argument is correct and such dividend being exempt income is to be included in the dividends under consideration in section 55(2), the one half of such dividend paid out of capital will be brought into tax: that half paid out of capital would not be relieved under paragraph (a), would form part of the excess under paragraph (c) and as much would be charged to tax by paragraph (c). This is obviously an impossible result. Moreover to the extent that part of the purpose of section 55(2) is to claw back allowable deductions otherwise claimable in respect of the relieved dividends, the section cannot be held to operate sensibly if it claws back the expenses incurred in producing exempt income, since such expenses cannot have been deducted in the first instance. There is nothing to be clawed back.
For these reasons, in their Lordships' view both "exempt income" and the expenses referable to the production of exempt income are to be left wholly out of account in applying section 55(2).
C. Allocation of interest charges.
In order to operate the provisions of section 55(2), it is necessary to allocate the management expenses and interest charges as between exempt income, A dividends and other income of the taxpayer company. The parties are agreed that it is impossible to identify specifically the management expenses attributable to, for example, exempt income. Such expenses are incurred generally for the purposes of running all aspects of the company's business. Therefore it is agreed that the total management expenses have to be allocated between the various types of income pro rata e.g. in the same ratio as the exempt income bears to the total income of the company from all sources.
But there is a dispute as to how the interest charges are to be allocated between the various types of income. The taxpayer company led evidence, which was not challenged, that it could identify certain borrowings as having been incurred for the specific purpose of producing a specific source of income, for example a borrowing to acquire shares in a company the dividends from which constitute "exempt income". The taxpayer company submits that in allocating the interest charges between the various categories of income, the interest payable on any specific borrowing should be allocated to the income earned by the asset acquired or improved by means of such borrowing. The Commissioner, on the other hand, contends that the interest charges should be dealt with in the same way as management expenses, i.e. the total interest charges of the company should be allocated pro rata between the various sources.
The relevant provisions of the Act are section 19(1) and 26(1)(a) which provide as follows:-
"19 (1) Any expenditure... shall be deductible from the gross income of a taxpayer in the income year in which it was incurred to the extent to which it is exclusively incurred in the production of his gross income for that year...
26 (1) Notwithstanding section 19, no deduction shall be allowed of any expenditure incurred on interest in an income year except,... expenditure incurred on interest -
(a) on capital employed in the course of any business or other income earning activity for the production of gross income derived in that income year;"
In order to be an allowable deduction, the interest has to be on capital employed in the course of any business "for the production of gross income". Prima facie in the case of a composite business having a number of activities one would expect the total allowable interest charges from the whole of the company's business to be set against the total income of the whole company business chargeable tax. If this were not done, interest charges incurred in relation to the start up of a new activity not yet showing a profit would be deductible. The interest referable to capital expenditure on a non-income bearing asset could not be shown to be interest incurred "for the production of gross income derived in that income year" in relation to that specific asset. If the taxpayer is right, interest charges referable to shares in non-profitable companies would not be capable of being set against the dividends payable by profitable subsidiaries.
If it is correct, in the ordinary case, to set the total interest charges against the total income of the composite business, it must follow in their Lordships' view that the same principle is applicable when interest charges have to be allocated between different sources of income. It cannot be correct, as their Lordships understand to be the taxpayer's contention, to allow all interest charges attributable to loss making activities to be set against the company's general income but in calculating the interest charges attributable to exempt income or A dividends to allocate the interest charges by reference to the actual interest charges paid in relation to the assets producing the exempt income or the A dividends as the case may be. The same method of treating interest charges must be applicable throughout.
Moreover, in the ordinary case (including the present) the company will have general borrowings on overdraft or loan account to finance the general conduct of the company's business. The interest on such borrowings is incapable of being specifically appropriated to any given source of income. If specific interest is to be allocated to specific income, how is interest on such general borrowing to be allocated ?
Although there may be exceptional cases, in the ordinary case of a company carrying on a composite business, interest charges must be allocated between different sources of income pro rata in the same way as management expenses.
Conclusion.
Counsel formulated for their Lordships a number of questions on which the parties desired decisions in principle. For the purposes of exposition, their Lordships have dealt with these questions in a different order. But to avoid any doubts as to the correct answers their Lordships now set out those questions and answers which they give:
(1) Whether as the taxpayer contends in subsection (2) must be identified on the basis that exempt dividends are excluded or as the Commissioner contends include all dividends ?
Answer: Excluded: see B. above
(2) Whether as the taxpayer contends "net income from all sources" in paragraph (d) must be identified on the basis that exempt income in excluded or as the Commissioner contends that it is included ?
Answer: Extended: see B. above.
(3) Whether as the taxpayer contends paragraph (d) imposes a limit on the amount of A dividends chargeable to tax as computed for the purposes of paragraphs (a) and (b) or as the Commissioner contends paragraph (d) lays down a method of making the computations required for the purposes of paragraphs (a) and (b) ?
Answer: As the Commissioner contends: see A. above.
(4) Whether as the taxpayer contends the words in paragraph (d) "the excess under paragraph (c) denote the amount computed under paragraphs (a) or (b) or as the Commissioner contends the amount of any capital dividends received by the taxpayer company ?
Answer: As the Commissioner contends: see A. above.
(5) Whether as the taxpayer contends references to the allowable management and interest expenses in paragraphs (a) and (b) are references to such expenses after deducting the amount of such expenses attributable to exempt income or without such a deduction as the Commissioner contends ?
Answer: After deduction of expenses attributable to exempt income: see B. above.
(6) If the answer to question (5) is in the former sense, whether as the taxpayer contends the expenses attributable to exempt income may be identified by a rateable apportionment of management expenses and a specific attribution of interest expenses or as the Commissioner contends must be identified by a rateable apportionment of both ?
Answer: As the Commissioner contends: see C. above.
(7) Whether as the taxpayer firstly contends the allowable amount of expenses in paragraph (d) means such expenses attributable to the production of A dividends or as the Commissioner contends has the same meaning as in the answer to question (5) ?
Answer: As the taxpayer contends: see B. above.
(8) If the answer to question (7) is in the former sense, whether as the taxpayer contends the expenses attributable to the A dividends are to be computed on the basis of a rateable apportionment of management expenses and to specific attribution of interest expenses or as the Commissioner contends on a rateable apportionment of both ?
Answer: As the Commissioner contends: see C. above.
Their Lordships will accordingly allow the appeal and remit the matter to the Commissioner for the determination of the tax liability of the appellant in accordance with the views they have expressed. There will be no order as to costs, here or below.
*
* *
