Commissioner of Internal Revenue v. Harmon/Dissent Douglas

Mr. Justice DOUGLAS, with whom Mr. Justice BLACK concurs, dissenting.

The federal income tax law makes a discrimination in favor of the community property states. In 1937 the Secretary of the Treasury pointed out that 'A New York resident with a salary of $100,000 pays about $32,525 Federal income tax; a Californian with the same salary may cause one-half to be reported by his wife and the Federal income taxes payable by the two will be only $18,626. The total loss of revenue due to this unjustifiable discrimination against the residents of 40 States runs into the millions.' That discrimination has become increasingly sharp as surtax rates have increased. The source of that discrimination is to be found in decisions of this Court.

Those decisions are best illustrated by Poe v. Seaborn, 282 U.S. 101, 51 S.Ct. 58, 75 L.Ed. 239, which involved the community property system of the State of Washington. They held that the husband need pay the federal income tax on only one-half of his salary and other income from the community, since the other half of those earnings from their very inception belonged to his wife. The collector had argued that the control exercised by the husband over the community was sufficient to make him liable for the tax on the full amount. That result had indeed been indicated by Mr. Justice Holmes speaking for the Court in United States v. Robbins, 269 U.S. 315, 327, 46 S.Ct. 148, 149, 70 L.Ed. 285. And it has been strongly urged that our recent decisions-such as Helvering v. Clifford, 309 U.S. 331, 60 S.Ct. 554, 84 L.Ed. 788, and Harrison v. Schaffner, 312 U.S. 579, 61 S.Ct. 759, 85 L.Ed. 1055-make for the same result. But in Poe v. Seaborn and related cases the Court discarded that test. It was more concerned with legal doctrine than it was with economic realities. It held that the wife's interest in the community (including the husband's salary) was 'vested' and that therefore the husband need pay the federal income tax on only half of that income.

One dubious decision does not of course justify another. But if Texas can reduce the husband's income tax by creating in his wife a 'vested' interest in half his salarly and other income, I fail to see why its neighbor, Oklahoma, may not do the same thing. The Court now concedes that once established, the community property status of Oklahoma spouses is at least equal to that of man and wife in any community property state. How then can Oklahoma be denied the same privilege which other community property states enjoy?

It is said that the elective feature of the Oklahoma statute causes it to run afoul of Lucas v. Earl, 281 U.S. 111, 50 S.Ct. 241, 74 L.Ed. 731, which held that an assignment of income to be earned or to accrue in the future was ineffective to render the income immune from taxation as that of the assignor. But the Court was not troubled with Lucas v. Earl in Poe v. Seaborn. It disposed of that argument by saying that in Lucas v. Earl the 'very assignment' was bottomed on the fact that 'the earnings would be the husband's property, else there would have been nothing on which it could operate. That case presents quite a different question from this, because here, by law, the earnings are never the property of the husband, but that of the community.' 282 U.S. at page 117, 51 S.Ct. at page 61, 75 L.Ed. 239. (Italics added.) By the same reasoning we should say that Oklahoma has made these earnings the 'property' of the community once the written election has been filed and that income which accrues thereafter never becomes the sole 'property' of the husband. Indeed we have the word of the Supreme Court of Oklahoma that such a transfer was effected by the written election filed by the husband and wife in this case. Harmon v. Oklahoma Tax Commission, 189 Okl. 475, 118 P.2d 205. There is no suggestion that the transfer of 'property' interests in this case is any less genuine or effective than it was in Poe v. Seaborn. The written election once filed is irrevocable. Only death or a decree of absolute divorce can alter it. Okl.Stat.Ann.1941, Title 32, § 51. If as Poe v. Seaborn holds the crucial circumstance is whether the income as it accrues is the 'property' of the community, it should make no difference for federal income tax purposes that the transfer from the husband to the community was effected by the act of filing a written election rather than by the act of marriage. If, 'by law, the earnings are never the property of the husband, but that of the community' (Poe v. Seaborn, supra, 282 U.S. at page 117, 51 S.Ct. at page 61, 75 L.Ed. 239), the husband should fare no better in Washington or Texas or California than in Oklahoma. The source of the 'law' which determines whether or not that result obtains is the same in each case-the legislature and the judiciary of the particular state. If they declare that the husband has lost and the wife acquired a 'property' interest by a certain act (whether by marriage, or by the filing of a paper), it is the 'law' though it is a recent pronouncement and not an 'inveterate' and long standing rule of that particular state. The consequence under the federal income tax statute is of course for us to decide. My only point is that if that is the formula for some states it should be the formula for all. We should apply it equally and without discrimination or we should discard it completely.

But it is said that the filing of a written election under the Oklahoma statute is an 'anticipatory arrangement' for the disposition of income under the rule of Lucas v. Earl; that a 'consensual' community will not be recognized for federal income tax purposes but that a 'legal' community will. As the Tax Court, however, pointed out (1 T.C. 40, 49) such a distinction will not stand scrutiny. Community property created by marriage is the effect of a contract. It is the result of a consensual act. The same is true where husband and wife agree to leave Oklahoma and establish their domicile in Texas so as to gain the advantages of a community property system. I can see no difference in substance whether the state puts its community property system in effect by one kind of contract or another. One is as much 'legal' as another. The agreement to marry or the agreement to move from Oklahoma to Texas is as 'consensual' as the act of filing a written election under the Oklahoma statute.

But if a distinction is taken between a 'legal' and a 'consensual' community, it cannot be consistently maintained for federal income tax purposes. In the first place, even the distinction which the Court seeks to take between this case and Poe v. Seaborn vanishes when after-acquired property is considered. Let us assume there is property first acquired in Oklahoma after the written election has been filed and in Washington after marriage. How are we justified in saying that Lucas v. Earl makes the written election but not the marriage an anticipatory arrangement affecting the income from that after-acquired property? Oklahoma is as explicit as Washington in saying that property so acquired by the husband 'shall be deemed the community or common property of the husband and wife and each, subject to the provisions of this Act, shall be vested with an undivided one-half interest therein.' Okl.Stat.Ann.1941, Title 32, § 56. In both cases the husband never was and never could be the sole owner of that property if local law is to be the guide. His 'status' under Oklahoma law is as fixed and irrevocable as it is under Washington law. How can it be said that after-acquired property is governed by 'status' in one case and by 'contract' in the other? If such a distinction is drawn, we are indeed making income tax liability turn on 'elusive and subtle casuistres'. Cf. Helvering v. Hallock, 309 U.S. 106, 118, 60 S.Ct. 444, 450, 84 L.Ed. 604, 125 A.L.R. 1368. In the second place, the Tax Court pointed out in this case that the difference 'between a community property law which is operative only when expressly invoked and one which operates unless expressly revoked' (1 T.C. at page 46) has no practical basis. There may be a 'consensual' community within a so-called 'legal' community. In some of the so-called 'legal' community property states separate property of one spouse may be converted by contract or deed into community property or vice versa. Volz v. Zang, 113 Wash. 378, 194 P. 409; State ex rel. Van Moss v. Sailors, 180 Wash. 269, 39 P.2d 397; Kenney v. Kenney, 220 Cal. 134, 136, 30 P.2d 398. But see Kellett v. Kellett, 23 Tex.Civ.App. 571, 56 S.W. 766; McDonald v. Lambert, 43 N.M. 27, 85 P.2d 78, 120 A.L.R. 250. And it has been supposed since Poe v. Seaborn that income from that type of community property was not thereafter to be treated as the separate property of the spouse who originally owned it. See 3 Mertens, The Law of Federal Income Taxation (1942) § 19.29. That has been the consistent view both of the courts (Black v. Commissioner, 9 Cir., 114 F.2d 355) and of the Tax Court. Shoenhair v. Commissioner, 45 B.T.A. 576, 579; Harmon v. Commissioner, 1 T.C. 40, 46, 47. And that has been the Treasury position. G.C.M. 19248, Int. Rev. Bull., Cum. Bull. 1937-2, p. 59. If Poe v. Seaborn states the correct rule, that view seems irrefutable. Community property is no less created 'by law' whether it was created by the contract of marriage or by a post-nuptial agreement.

But are we now to understand that postnuptial agreements in all community property states are ineffective for federal income tax purposes because they are 'consensual'? Or is the Court willing to give income tax effect to such contracts only within the established community property states? If it is the former, then we are overriding settled administrative construction on which great reliance was placed in Poe v. Seaborn, 282 U.S. at page 116, 51 S.Ct. at page 61, 75 L.Ed 239. If it is the latter, then we can hardly say that the difference between the Oklahoma system and the Washington system is that Washington has created its system 'as an incident of matrimony' while Oklahoma has not. In that event we make unmistakably plain the discrimination against Oklahoma-we give income tax effect to a post-nuptial agreement between spouses in eight states and deny effect to a similar agreement in Oklahoma. The only apparent basis for such discrimination is that the community property systems in the eight states are traditional; that those eight states have a well settled policy; that Oklahoma merely gives its citizens a choice to get under or stay out of its community property system. Yet how can we say that the state which allows husband and wife to revoke or alter its community property system by contract has a more 'settled' policy towards community property than a state which gives husband and wife the choice to invoke its community property system or to keep their marital property on a common law basis? The truth is that there is a wide range of choice in each. But the fact that there is a choice should not be deemed fatal when Oklahoma's case comes before the Court and irrelevant when Washington's case is here.

The distinctive feature of the community property system is that the products of the industry of either spouse are attributed to both; the husband is never the sole 'owner' of his earnings; his wife acquires a half interest in them from their very inception. 1 de Funiak, Principles of Community Property (1943) § 239. That was the test which Poe v. Seaborn adopted. If Oklahoma meets that test, then she should be treated on a parity with her sister states. The fact that her system is new-born does not make it any the less genuine.

I do not mean to defend Poe v. Seaborn. I only say that if we are to stand by it, we should not allow it to become a 'vested' interest of only a few of the states. The truth of the matter is that Lucas v. Earl and Helvering v. Clifford on the one hand and Poe v. Seaborn on the other state competing theories of income tax liability. Or to put it another way, Poe v. Seaborn has been carved out as an exception to the general rules of liability for income taxes. If we are to create such exceptions we should do so uniformly. We should not allow the rationale of Poe v. Seaborn to be good for one group of states and for one group only. If we are to abandon the rationale of Poe v. Seaborn, we should do so openly and avowedly. If the practical consequences of applying the rationale of Poe v. Seaborn to other situations would be disastrous to federal finance, it is time to reexamine the case. The rule which it fashions is the rule of this Court. We have the responsibility for its creation. If we adhere to it, we should apply it without discrimination. If we are not to apply it equally to all states, we should be rid of it. This is the time to face the issue squarely.