Opinion
No. 129.
March 3, 1930.
Appeal from the District Court of the United States for the Southern District of New York.
Action by New York Life Insurance Company against Frank K. Bowers, Collector of Internal Revenue for the Second District of New York, to recover back capital stock taxes paid by the plaintiff during the years ending June 30, 1919, 1920, 1921, and 1922. From the judgment [ 34 F.2d 60], both parties appeal.
Affirmed.
James H. McIntosh, of New York City, for plaintiff.
Samuel C. Coleman, of New York City, for defendant.
Before L. HAND, AUGUSTUS N. HAND, and CHASE, Circuit Judges.
The plaintiff is a large mutual life insurance company with its principal place of business in New York. In common with other such institutions it had before 1906 issued policies in consideration of equal yearly premiums, the accumulated dividends from which it retained for a fixed period, and divided among those who survived and had kept their policies alive when the period expired. These are semi-tontine policies. Besides these it did the ordinary business of life insurance under which also the insured paid equal yearly premiums, in consideration of a payment should he die within a given period, or a sum at the end of the period if he survived; or merely of a payment whenever he did die. Upon such policies the insured receives a dividend in the year following the payment of each premium. To provide against contingencies all premiums are computed at more than the estimated cost of the insurance, and at the end of each year there is a remainder, called a "dividend," to which under his contract the insured is entitled, either absolutely, or, in the case of semi-Tontine policies, contingently upon his survival. In calculating the cost of insurance the insurer not only includes the actual losses which occur during the year, but "reserves," determined on an actuarial basis and universally required by statute. These are to accumulate sufficient sums to pay future losses, and thus to keep the class for any year actuarially solvent as a whole. Besides these, the necessary expenses of the business must be paid.
The tax in question was levied by virtue of section 1000(c) of the Revenue Act of 1918 (40 Stat. 1126), which provided that in the case of mutual companies it should be based upon "the sum of its surplus or contingent reserves maintained for the general use of the business and any reserves the net additions to which are included * * * under the provisions of Title II, as of the close of the preceding accounting period used by such company for purposes of making its income tax return." During the years in question in its returns made to the New York superintendent of insurance, the plaintiff carried three items, numbered thirty-five, thirty-six and thirty-seven, respectively, which are the subject of this action. The first was designated "Dividends declared on, or apportioned to, annual dividend policies payable to policy holders to and including December 31" (of the following year), "whether contingent upon the payment of renewal premiums or otherwise." This was the amount to be declared by the plaintiff, as the remainder or profit upon those premiums received during the past year upon insurance other than semi-tontine, after paying any losses in that year, setting aside the necessary "reserves," and deducting the proper share of the expenses.
Item thirty-six read as follows: "Dividends declared on, or apportioned to, deferred dividend policies payable to policy holders to and including December 31," of the following year. This represented the accumulation of all dividends throughout the existence of such policies, including the year in question, which would be paid in the following year upon the semi-tontine policies which fell due in that year. Item thirty-seven was like item thirty-six, except that it represented the accumulated dividends throughout their existence of all semi-tontine policies which would fall due at some time later than the following year. In the case of all such dividends they were declared by resolution of the plaintiff's board of directors during the month of January in the following year.
In estimating the tax in question the Commissioner included all of items thirty-five, thirty-six and thirty-seven for each current year, and this is the source of the plaintiff's complaint. It contends that the phrase, "surplus and contingent reserves maintained for the general use of the business," includes only those sums, the result of the general operation of the business, which have accumulated above legal reserves and declared dividends. The District Court held that, as the dividends were not set apart, but remained in the plaintiff's business for all purposes, the items were "maintained for the general purposes of the business," and that they were part of "the surplus and contingent reserves."
In the case of ordinary stock companies the statute (section 1000(a)(1) levied a tax upon the value of the capital stock of the corporation, including its surplus and undivided profits, and this would ordinarily be reached merely by subtracting the liabilities from the assets. To insurance companies of any sort this was inappropriate. During the earlier years of a group of coeval policies the premiums are not enough to answer the payments, treated as absolute, though future, liabilities; it is only as the policies mature that the accumulated premiums become large enough to answer the later payments. It was therefore necessary to adopt some other rule. The law of 1916 taxed only stock insurance companies, and excluded from capital "such deposits and reserve funds as they are required by law or contract to maintain or hold for the protection of or payment to or apportionment among policyholders." Revenue Act of 1916, § 407 (39 Stat. 789). It apparently presupposed that policies were not liabilities, though it did not say so. In other respects the capital was to be calculated like that of an ordinary business stock company. A question might arise whether "dividends," annual or deferred, were included within this exemption, but we have not that before us, except indirectly, because in 1918, while section 1000(b), 40 Stat. 1126, was retained, Congress by section 1000(c) expressly provided for mutual companies. In these, though all the policy holders are collectively a debtor to each one, all are also members of the company, and therefore collective owners of any profits which it may make. While it is true that each policy provides for the declaration of those profits in the form of dividends, it is rather as a charter — also a contract — provides for the declaration of dividends to shareholders in a stock company. The policy holders are therefore at once associates in the business of life insurance, as to which they are investers, and creditors of the group as a whole. Penn Mutual Ins. Co. v. Lederer, 252 U.S. 523, 40 S. Ct. 397, 64 L. Ed. 698.
When Congress in 1918 undertook to supply the casus omissus left in 1916, it defined the amount to be taxed as a "sum" of two factors (section 1000(c). The first was the "surplus or contingent reserves maintained for the general use of the business"; the second, those "reserves the net additions to which" were included in the income tax of the company. We agree that "dividends," whether annual or accumulated, should not be classed as "reserves," a word which has an accepted meaning in insurance, and which covers primarily those funds required by law, though in some cases added funds required by contract. Therefore the items in suit are not included within the second factor of the "sum" to be taxed. We think that they are part of the surplus, and our reasons are that they are more nearly profits than the other sums which the section included. It expressly taxed "contingent reserves," that is, limited funds accumulated at the option of the insurer out of abundant caution (e.g., section 87 of the N.Y. Insurance Law [Consol. Laws, c. 28]). These are withdrawn from what would otherwise be declared as dividends; they are funds temporarily at any rate taken out of profits and devoted to the insurer's contract obligations. Also it taxed those "reserves," stricti juris, whose net additions entered into the income tax. Section 234 (10), 40 Stat. 1079, allowed insurance companies to deduct from their gross income the "net addition required by law to be made * * * to reserve funds." Additions to all other "reserves" were taxed as income, though they were such in the most limited sense. The second factor of the "sum" defined in section 1000(c) thus comprehended any "reserves" required by contract as contrasted with those "required by law."
It appears to us incredible that Congress should have meant to exempt dividends which are taxable as income and to include both "contingent reserves" and those other reserves which we have just considered. The general purpose appears to have been to make the test whether the taxed funds were made up of sums whose annual accretions were subject to the income tax, no matter how they were allocated in the corporate accounts. At any rate as between dividends and such "reserves," contingent or absolute, the preference in exemption would naturally be for the second, and a tax which covers them must cover the first. Furthermore, whatever may be said as to taxing any "reserves" whatever, to tax the dividends of a mutual company is at least consistent with the distinction between its obligation as insurer and the profit of the members as a group. The profits do not even come altogether from premiums; in part at any rate they result from the financial operations of the whole group, which thus makes profits or losses as such. There is more than a formal reason for treating separately what it may engage to pay according to the tenor of its contracts, and what it distributes, even though these very contracts stipulate that there shall be a distribution.
That this results in inequality between mutual and stock companies is true. As to "reserves" there can be no doubt; the stock companies escape all such taxes, the mutual do not, unless they hold no more than those "required by law." As to dividends, the inequality is not so clear. To be sure they would seem to fall within the phrase, "such deposits * * * as they are required by * * * contract to maintain for * * * apportionment among policy holders," especially when "apportionment" is contrasted with "protection" and "payment." Yet it is at least open to argument that dividends, though the policy may secure them, are not "required by contract," and that the same distinction was intended between payments as such and profits, as appears in section 1000(c). For it is possible to regard the policy holders even of a stock company as also associates in the enterprise, and perhaps that is the juster view. But even if it is not, and if the dividends of such policy holders are exempt, Congress may have intended to insist upon the legal distinction between them and shareholders. Whatever the resulting inequalities, section 1000(c) was certainly intended to be the sole measure of the tax on mutual companies, and its language bears evidence of being deliberately chosen. We cannot temper the result without disregarding the plain sense.
Finally, unless we are to suppose that mutual companies were taxed on dividends, there was very little which could be reached at all. Not all such companies have "contingent reserves," and the "surplus," after the deduction of reserves of all sorts and dividends, may be, as here it would be, a very small matter, or nothing at all. The phrase, "maintained for the general use of the business," might indeed lend color to the plaintiff's position, were it not affixed not only to "surplus," but to "contingent reserves" as well. It is however impossible to see in what sense such reserves are maintained for the general use of the business which does not equally apply to dividends. If the clause means that the surplus must be unallocated for any purposes, it is not appropriate to "contingent reserves."
There remains the second question raised by the appeals, which is whether the taxes levied for the years beginning July 1, 1920, and July 1, 1921, had the support of any statute. Section 1000(a)(1) of the Act of 1918 levied a "special excise tax with respect to carrying on or doing business," upon the "fair average value" of the capital "for the preceding year ending June 30," and subdivision (c) provided that it should not apply "to any corporation which was not engaged in business * * * during the preceding year." The District Court decided that the tax was levied for the preceding year, but that that due July 1, 1921, was in force notwithstanding the Act of 1921. The tax for the following year was not however in force, because the statute had been repealed before July 1, 1922. Each side appeals from this ruling.
As to the fourth year — July 1, 1921, to June 30, 1922 — we have no doubt. If the tax is levied for the preceding year it is conceded that by section 1400(a) of the Act of 1921 (42 Stat. 320) section 1000 of the Act of 1918 (40 Stat. 1126) was repealed to take effect as of January 1, 1922, and there was therefore no statute in effect in the following July. We think that the same is true, even though the tax was payable in advance, as the defendant maintains. Section 1400(b) of the Act of 1921 provided: "In the case of any tax imposed by any part of the Revenue Act of 1918 repealed by this Act, if there is a tax imposed by this Act in lieu thereof, the provision imposing such tax shall remain in force until the corresponding tax under this Act takes effect." Section 243 of the Act of 1921 (42 Stat. 261) imposed "in lieu of the taxes imposed by sections 230 and 1000" a tax upon "the net income of every life insurance company" "for the calendar year 1921" and thereafter. Section 1000 of the Act of 1921 re-enacted the capital stock tax of section 1000 of the Act of 1918 which was to be "in lieu of" it. It did not comprise insurance companies. While the words, "section 1000," in section 243 of 1921 refer to section 1000 of that act, and verbally the income tax was therefore only in lieu of the capital stock tax of 1921, it was nevertheless also "in lieu of" the tax of 1918, because it took the place of the capital stock tax of 1921, which in turn took the place of the preceding tax of 1918. Hence the passage quoted from section 1400(b) applied to it, and the tax in suit could not be levied on July 1, 1921, even though payable in advance. The income tax levied by section 243 of the Act of 1921 — "the corresponding tax" — had already "taken effect" on that day; it included the calendar year 1921, although collected only in 1922. It could not have been the intent of Congress to impose two taxes for the period between July 1, 1921, and June 30, 1922, when one was "in lieu" of the other. Since the Act of 1921 was enacted after July 1, 1921, this indeed results in the remission of the capital stock tax already levied on July 1, 1921, but the language of section 1400(b) seems to us to demand that result.
As to the tax for the year before, the case is more troublesome. There is no difficulty, if that tax is payable in advance, because the income tax had not "taken effect" on July 1, 1920. If however it was payable for the preceding year, the tax due July 1, 1921, in part covered a period during which the income tax had already "taken effect"; it did not any longer "remain in force." Owing to the difference in the period of taxation, the inevitable alternative is between levying a double tax upon the plaintiff for the first six months of 1921, or allowing it to escape all such taxation for the previous half year. The answer must depend upon whether taxes under section 1000 of the Act of 1918 were levied in advance or for the preceding year.
They were construed in the former sense by the Treasury (Reg. 50, art. 1), and so had been the tax of 1916, which was almost literally the same. Such an interpretation, already made when the Act of 1918 was passed, is strong evidence of intent. National Lead Co. v. U.S., 252 U.S. 140, 145, 146, 40 S. Ct. 237, 64 L. Ed. 496; Heiner v. Colonial Trust Co., 275 U.S. 232, 235, 48 S. Ct. 65, 72 L. Ed. 256. Moreover, these are "special taxes," long known to our fiscal system, and generally payable in advance (section 186, Title 26, U.S. Code (26 USCA § 186). It is true, as the District Court observed, that section 1000(c) of 1918 provides that "the taxes imposed * * * shall not apply in any year to any corporation which was not engaged in business * * * during the preceding year"; and section 1000(a)(1) computes them on "the fair average value" of the capital for the preceding year. Of course it is impossible to calculate such a tax in the case of a company that had done no business during any part whatever of the preceding year. But section 1000(c) does not say that the company must have been in business "during the whole of the preceding year," but only "during the preceding year." The same applies to the use as a basis of the "fair average value for the preceding year." It was entirely possible to compute the tax in the case of a company which had been in business for a part of the preceding year, and this satisfies all that the words demand. The Court of Claims (Washington Water Power Co. v. U.S., 56 Ct. Cl. 76) has decided in a case arising under the Act of 1916 that the tax is payable in advance, and this seems to us the proper construction. Finally, as we have said, the tax had its origin in the Revenue Act of 1916, which imposed it for the year beginning January 1, 1917. While it is possible that Congress meant to levy it for a period during which the act was not in effect, the language is "on and after January first, nineteen hundred and seventeen." It is more likely that such an excise, in accordance with the usual theory, was laid prospectively upon an activity to be engaged in, than upon its pursuit during a time when there was no tax in view.
Judgment affirmed.