Freely Available British and Irish Public Legal Information
[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
United Kingdom House of Lords Decisions
You are here: BAILII >> Databases >> United Kingdom House of Lords Decisions >> Sun Insurance Office v. Clark [1912] UKHL 1038 (07 March 1912)
URL: http://www.bailii.org/uk/cases/UKHL/1912/49SLR1038.html
Cite as: [1912] UKHL 1038, 49 ScotLR 1038
[New search] [Printable PDF version] [Help]
SCOTTISH_SLR_House_of_Lords
Page: 1038↓
House of Lords.
(On Appeal from the Court of Appeal in England.)
(Before the
v.
Subject_Revenue — Income Tax Act 1842 (5 and 6 Vict. cap. 35), Sched. D — Insurance Company — Estimation of Profits — Deduction for Unexpired Risks.
Facts:
In estimating the average of profits of an insurance company assessable to income tax it is a question of fact
Page: 1039↓
which is the fairest method. General Accident Fire and Life Assurance Company v. Inland Revenue ( M'Gowan), [1908] AC 207, 1908 S.C. (H.L.) 24, 45 S.L.R. 681, lays down no fixed rule of law for such ascertainment. The appellants contended that in estimating yearly profits they were entitled to carry forward 40 per cent. as a reserve against unexpired risks. During the three years of which the profits were averaged for ascertainment of income tax this reserve had increased by £56,334, making a difference of £18,778 in the return for the year. Their contention was upheld by the Commissioners and Bray (J.), but the Court of Appeal held that in the case of the General Accident Fire and Life Assurance Company v. M'Gowan ( sup.) a rule was laid down for the assessment of such profits.
Held that no such rule was laid down in M'Gowan's case, and that the method of assessment pursued by the company was in the circumstances the fairest.
Appeal from the judgment of the Court of Appeal ( Cozens-Hardy, M.R., Fletcher Moulton, and Buckley, L.JJ.), reversing judgment of Bray (J.) upon a Case stated by the Income Tax Commissioners.
Their Lordships gave considered judgment as follows:—
When a fire insurance company is to be assessed for income tax, the main point to be ascertained will be what has been the gain or the loss upon the contracts of insurance which it has effected in each of the three years which are taken to make the average.
If it were practicable, the accurate way, I suppose, would be to add together the premiums which the company became entitled to receive in each year, say 1903, upon contracts made in that year, and then to add up the losses which the company became bound to pay upon those contracts made in 1903. The difference between these two sum-totals would show precisely what was the gain of the company or their loss in respect of the contracts made in 1903.
But this is impracticable, because contracts of fire insurance are made all through the year, from the 1st January to the 31st December, and most of them, or at all events many of them, are made to cover fire risks for a year, some we are told for five or six or seven years, from the date of their making. The premium is paid in advance. So the result in the way of gain or loss could not be ascertained as a fact until after the period of time had elapsed. Thus it appears that you cannot base the assessment of income tax upon the actual facts of the business done and the actual pecuniary results of it in the case of fire insurance companies who take single premiums to cover risks for a year or for more years. This is such a company, and I believe that nearly all companies are in the same position. If that be so, it follows that in assessing such fire insurance companies you must proceed wholly or in part by estimate.
An estimate being necessary, and the arriving at it by in some way using averages being a natural and probably inevitable expedient, the law as it seems to me cannot lay down any one way of doing this. It is a question of fact and of figures whether what is proposed in each case is fair both to the Crown and to the subject.
In M'Gowan's case, to which reference is made, three methods of estimating these gains or profits were before the House. I place first, merely for convenience sake and not for its importance or value, a faint suggestion which was made in that case, and as I understood it was as follows—It was suggested that each contract of insurance made during a particular year should be considered separately. If it had expired, then the actual result should be taken, whether profit or loss. If it had not expired, then an estimate should be made, having regard to the period unexpired and the degree of risk, which might be different (in summer and winter, for example) during that period. I do not imagine that either the Crown or the company would seriously desire such an inquiry. I do not know how many fire insurances are effected by a great company within a twelve-month, probably scores of thousands or even hundreds of thousands. Such a process, as to the unexpired contracts, would be minute and almost interminable. It was rejected because there was no evidence that it would be a reasonable way of ascertaining what was desired.
The second method suggested in that case was that of merely taking for each year the sum total of the premiums received and the sum total of the losses paid and subtracting the one from the other, without regard to the fact that the premiums cover risks running on into subsequent years, and the losses include losses arising out of contracts made in previous years. This method is, of course, not precise or scientific. It proceeds upon the view that when this is done for the three consecutive years indicated by the statute and the figures thus reached are averaged, a fair and reasonable conclusion is attained.
This method was adopted long ago, and has more than once been the subject of consideration in courts of law. I can conceive it being unfair either in the
Page: 1040↓
A third method, similar in principle to that advanced by the now appellants, was also considered by your Lordships in the case of M'Gowan. This method is to carry forward annually at the close of the year a percentage of the premium income in order to allow for unexpired risks. It has no pretensions to being precise. I can easily imagine cases in which an actuary could show that it was misleading. But if it comes nearer to the truth than any other method in a particular case, I do not understand why it should not be adopted. This third method, however, in its application to that case, would have meant that the insurance company was to pay income tax upon the footing of about £15,000 profits and gains in the three statutory years, whereas they had divided dividends of about £60,000 in those three years, and there was no evidence and no finding, nor could there honestly have been, that the third method worked fairly between the Crown and the subject.
In those circumstances this House rejected the third method, and adopted the second, but so far from laying it down as a rule of law, it was expressly pointed out that the second method was of itself imperfect, and though a good working rule generally, would not be applicable if in any case it appeared upon the facts to involve hardship. The headnote in the Law Reports is not accurate, and the view of Bray, J., of what the House decided is accurate.
After this preface, tedious but necessary in the circumstances, I come to the merits of the present appeal. Here again there is a competition between two methods of estimate. That which I have called the second is propounded by the Crown. That which I have called the third is propounded by the company, who deduct 40 per cent. of the premiums at the end of each year. The relevant facts are here the reverse of what they were in the case of M'Gowan. The third method has been examined by the Commissioners, and it is stated in the Special Case to be right in this case. It was in terms admitted by the surveyor of taxes before the Commissioners that the fair and reasonable allowance for this company to make, if entitled to make any allowance, was 40 per cent. The surveyor's point was that no allowance or deduction at all ought to be made, because he said that the proper method according to law was the second method. He did not prove, nor try to prove, that it was fair in this case, so that all the evidence and the finding in the present appeal was in support of the appellants' contention. In these circumstances it seems to me quite obvious that the third and not the second method must be applied here, for the plain reason that upon the materials before us it is the fair and only way presented to us by which the truth can be approximately attained.
In the hope that it may help to prevent future misunderstanding, I will recapitulate my own opinion. There is no rule of law as to the proper way of making an estimate. There is no way of estimating which is right or wrong in itself. It is a question of fact and figures whether the way of making the estimate in any case is the best way for that case. Experience seems to have satisfied courts of law for a considerable time that the method which I have described as the second is a useful working rule. But no one has said in this House that there is any constraint to accept it. It may be that the character or mode of carrying on this insurance business may alter or may have altered, and what was a good method once may become inaccurate or even obsolete.
I am equally anxious that your Lordships should not be supposed to have laid down that the method applied by the Commissioners in the present case has any universal application. If the Crown wishes in any future instance to dispute it they can do so by evidence, and it is not to be presumed that it is either right or wrong. A rule of thumb may be very desirable, but cannot be substituted for the only rule of law that I know of, viz., that the true gains are to be ascertained as nearly as it can be done.
I think that this appeal must be allowed.
The appellants carry on the business of fire insurance. It has been their practice to carry forward annually 40 per cent. of their yearly premium receipts in order to bring about the correct proportion as between year and year of the premium income which has to answer to the annual incidence of fire losses, on the balance of which income, over or under the losses or charges for the year, the profit or loss attaching to their business depends. If the premium income were stationary this would make but little difference as regards income tax, for a uniform amount deducted would be carried forward to profit in the ensuing year. But the business of the appellants and their premium income are
Page: 1041↓
It is plain that the question of what is or is not profit or gain must primarily be one of fact, and of fact to be ascertained by the tests applied in ordinary business. Questions of law can only arise when (as was not the case here) some express statutory direction applies and excludes ordinary commercial practice, or where, by reason of its being impracticable to ascertain the facts sufficiently, some presumption has to be invoked to fill the gap. Such a presumption was made in the appeal referred to, where, as pointed out in the judgment delivered by the Lord Chancellor, it was not shown that the percentage deducted represented the real value of the risks yet to run and was therefore unearned, and it appeared further that the company had actually divided the entire amount without deduction among its shareholders as part of the profit for the year. In the General Accident Company v. M'Gowan it was, therefore, in the absence of evidence to the contrary, presumed that the whole of the premium income received in the year represented income earned by bearing the risks of the year. It is difficult to see how, on the materials before the House, the decision could have been otherwise, but it seems to me equally clear that the reason of the decision has no application where, as here, it has been accurately ascertained that part of the premiums is not in the nature of profit earned. As Bray (J.) puts it, the case is analogous to one in which if goods are bought their value cannot be treated as profit without deducting the value of the liability to pay for them which the buyer has incurred. Such cases as Imperial Fire Insurance Company v. Wilson, 1 Tax Cas. 71, and Scottish Union and National Insurance Company v. Smiles, and Northern Assurance Company v. Russell, 2 Tax Cas. 551, 10 R. 461, 26 S.L.R. 330, are not, when carefully examined in the light of what appears to be the true principle, to be relied on as authorities for a proposition which would run counter to the practice and good sense of the commercial community. The Judges in the Court of Appeal appear to have thought themselves bound to hold that this House had laid down a rule of law so rigid that, although introduced for the purpose of supplying deficient evidence of fact, it must still apply even where there was no deficiency of proof. I have come to a different conclusion. I think that Bray (J.) has interpreted correctly the decision in question when he says that if in the earlier case such facts had been before this House as have been established in the present case, it would in his opinion have decided against the Crown. The present appeal ought, in my judgment, to succeed.
The facts are not in dispute. The appellants issue policies for periods of twelve months and possibly longer. The risks on these policies run, not from a fixed date, but from the dates named in the policies, which may be any day in the year. In consequence at the end of the financial year there are invariably a considerable number of policies still running in respect of which the appellants are still under liability to pay should any loss occur. The appellants alleged that, inasmuch as the premiums on such policies cover a period beyond the financial year, the total amount of premiums on such policies cannot be treated as ascertained profits, as the company may be called upon to pay losses under and by virtue of the contracts still existing.
It is not disputed that as a matter of principle the claim of the appellants is well founded. Premiums are not profits or gains. They are receipts which must be brought into account, and out of which, after proper deduction for losses, profits will accrue. If the actual amount of the premiums received in any financial year was substantially constant the assessment of gains and profits after the second year would not be affected; but the business of the appellants was an increasing business, and for the three years 1902, 1903, and 1904 the total increase of premium income averaged £18,788.
The only matter in dispute was the right of the appellants, in estimating the profits or gains, to make a deduction in respect of the outstanding premiums due to increased business during the three years under consideration. The matter was investigated by the Commissioners of Income Tax, who after hearing evidence were of opinion that the company was entitled to carry forward a percentage of fire premiums in one year in respect of unearned premiums which did not form part of the profits of the year. It was not disputed by the
Page: 1042↓
The question of what are profits or gains within the meaning of the Income Tax Act is prima facie a question of fact, and if the cases from Imperial Life Insurance Company v. Wilson in the year 1876 down to General Accident Company v. M'Gowan in 1908 be examined, it will be found that in every case the Courts have treated the question as one of fact, and have merely decided whether upon the facts before them the claim of the taxpayer to make a deduction in a particular way was justified. The Attorney-General contended that no estimate was permissible in arriving at the amount to be deducted, and he based this argument principally on the language of the judgment of the Court of Exchequer in Imperial Fire Insurance Company v. Wilson, and the language of the Lord President in the case of Scottish Union v. Inland Revenue ( 16 R. 461), cited in the judgment of Cozens-Hardy, M.R.; but in each case, if properly examined, it will be found that the Court were dealing with a question of fact, and were only approving that which in their judgment appeared tp be the best solution of that question, the method which had been approved in the case of Imperial Fire Insurance Company v. Wilson.
Coming now to the case upon which express reliance was placed, the respondents in their case state that it is governed by the decision of your Lordships' House in M'Gowan's case. In that case the mode of estimating the proper deduction put for ward by the then appellants was rejected by your Lordships' House as being absolutely unjustified upon the facts. See the judgment of the Lord Chancellor, who decided that the particular correction sought by the appellants in that case was indefensible upon the materials before the House, and that the method adopted by the Commissioners was a good working rule in the then present instance and generally. But he went on to say that if in any particular case an insurance company could show that it worked hardship, the rule ought to be modified so that the real gains and profits might be ascertained as near as may be. It is said that Lord Macnaghten in his judgment went further, but I have seen nothing to lead me to believe that he was purporting to lay down any rule of law; he was, in my opinion, only expressing very strongly his view upon the facts then before your Lordships' House.
I adopt the reasoning in the judgment of Bray (J.); who considered that, properly understood, the case of General Accident Company v. M'Gowan was a decision upon the facts and not on law, and I am unable to agree with the decision of the Court of Appeal that your Lordships intended to lay down any binding rule that as a matter of law no such deduction could be allowed. For these reasons, I am of opinion that the decision of the Court of Appeal should be reversed and the judgment of Bray (J.) restored.
Speaking for myself, I feel bound to say, with all respect, that I find great difficulty in seeing how, if that case be examined in a fair and reasonable though critical spirit, it can be supposed that in it any rule of law was laid down or anything decided as a matter of law. Two different practical methods for ascertaining the profits and gains of the appellants' business for the purposes of the Income Tax Acts were suggested and insisted upon by the respective parties in the case. The method insisted upon by the appellants led, on the facts, to grotesque results, while that insisted upon by the respondents, which was the same as that contended for by the Crown in the present case, led to comparatively accurate and just results. This latter was proved to be the usual method, and was considered to be simple, practical, and easily applied. A third method was suggested which was not practical. Your Lordships had to choose between the two practical methods, and you chose that which led to the just results.
That, however, is a very different thing from deciding as a matter of law that the method so approved of was the only legal method which could be adopted in such cases, or that deductions were never to be made from the premiums received on fire policies in respect of risks unexpired at the time when the yearly accounts of the company were made up. A little consideration of one of the most illuminating authorities in the books upon this question of the mode of ascertaining the taxable gains
Page: 1043↓
It is obvious that the entire premium cannot in every case be divided in the same proportions as those into which the accounting period divides the entire duration of the policy, because the risk is not the same for each calendar month or each season. On the contrary, the risk is admittedly greater in the winter months than in the summer months, because fires and artificial light are more used in the former than in the latter. For instance, if the 31st December be the accounting day and the policy be effected on the 1st July, it might be quite right to divide the premium into two equal portions, one-half being treated as earned in the year in which the policy was effected and the other half as to be earned in the succeeding half-year, but if the policy were effected on the 1st October, then it would obviously be unfair on account of this greater risk to treat only one-quarter of the entire premium as having been earned in the three winter months of October, November, and December. Having regard, therefore, to the fact that companies carrying on business of this kind are, under the decision of your Lordships' House, clearly entitled to object to their receipts being treated per se as their profits and gains without the proper deduction having been made of the cost of earning those receipts, it is obvious that the amount of the taxable profits and gains can only be ascertained by some system of averages or estimation, or by some other practical rule of thumb based upon experience and the facts of different cases.
As I understand, Fletcher Moulton, L.J., is of opinion that some method might be adopted by which the problem could be solved with scientific accuracy. Unfortunately he does not describe in detail what that method is. It may, for all that appears, involve such complicated calculations in reference to each particular policy as to be practicably unworkable, and certainly it does not appear to me that it is possible to adopt any method which does not involve recourse, at some stage of its processes, to estimates, averages, or such like things. If these matters be borne in mind, it does seem strange that it should have been supposed that it could ever have been laid down as a matter of law that one method, and one method alone, could legally be employed to solve the problem.
I now turn to a detailed examination of M'Gowan's case. If any meaning is to be given to words, no ingenuity can find in the judgment of the Lord Chancellor anything equivalent to a statement that as a matter of law only one rule or method could be applied to reach the desired result. On the contrary, he is reported to have used these words—“I think that the particular correction sought by the appellants in this case is indispensable upon the materials before us, and further, that the method adopted by the Commissioners is a good working rule in the present instance and generally. If in any particular case an insurance company can show that it works hardship, no doubt the rule ought to be modified so that the real gains and profits may be ascertained as near as may be. I am for dismissing the appeal.” That clearly means this, that the method adopted by the Commissioners is not the only method which can be legally adopted, but is a good working method for general use, to be departed from where it is shown to work injustice. Lord Ashbourne said that he concurred with the Lord Chancellor. That may mean that he merely concurred in the conclusion at which the Lord Chancellor had arrived, or that, in addition, he adopted every word which the Lord Chancellor had said. I understand that, according to the practice of your Lordships’ House, merely concurrence, as Lord Ashbourne expressed it, conveys the former meaning, not the latter. Lord Macnaghten then delivered judgment. It is contended that his judgment amounts to a decision that the rule adopted by the Commissioners is the only rule which can legally be adopted in such cases to ascertain
Page: 1044↓
I do not myself think that a fair and reasonable construction of his language leads to any such conclusion, but if I am mistaken in that view then it cannot be disputed that the conclusion at which he arrived (quite unnecessary for the decision of the case) is the very contrary of that at which the Lord Chancellor arrived. The one noble Lord says in effect that the rule is not a rule of law at all, but is generally a good working rule, to be departed from when it is shown to work injustice, and the other lays it down that it is the only legal rule, is to be invariably applied, and can never be departed from whether it works injustice or not. No two statements could be more irreconcilable. Lords James of Hereford and Robertson and I myself then stated that we concurred—but concurred with what or in what? I know that their minds were too logical, and I hope that my own mind was too logical, to agree at the same moment with each of two contradictory and irreconcilable propositions. We must, therefore, either have considered that the two judgments must have meant the same thing, or our concurrence must have been limited to a mere concurrence in the result arrived at by both our colleagues without adopting the precise language of either. It is, I think, plain that Lord Collins, in his judgment delivered after our concurrence had been expressed, did not use any language from which it could be fairly or legitimately inferred that in his view the rule approved of by the Lord Chancellor was the only rule which can be legally applied in such cases.
I am, therefore, clearly of opinion that this case of M'Gowan is no authority whatever for the proposition that the rule adopted so reluctantly by the Court of Appeal is the only legal rule which can be applied for the purpose indicated. If the Court of Appeal considered, as apparently they did consider, that the case was an authority for such a proposition, binding upon them, they were, in my view, with all respect, entirely mistaken. It was, I think, quite open to them to approve of a different method of ascertaining which would lead to what they considered true and just results. In this particular case it is shown upon the evidence that the method contended for by the appellants is, in this instance at all events, a better method, juster and more satisfactory in its application and results than that insisted upon by the Commissioners. It should, therefore, I think, be approved of and applied in preference to the latter. I accordingly am of opinion that this appeal should be allowed with costs.
Order appealed from reversed. Respondent to pay to the appellants their costs here and below.
Counsel for the Appellants— Danckwerts, K.C.— Bremner. Agents— Dawes & Sons, Solicitors.
Counsel for the Respondent—The Attorney-General ( Sir R. Isaacs, K.C.)—The Solicitor-General ( Sir J. Simon, K.C.)— W. Finlay. Agent—Solicitor for Inland Revenue.