(dissenting).
I dissent because in my opinion the decision of the majority is directly contrary to the law as enunciated by the Supreme Court in Ithaca Trust Co. v. United States, 1929, 279 U.S. 151, 49 S.Ct. 291, 73 L.Ed. 647 and also opposed to long-recognized principles of estate valuation.
The decedent died on July 8, 1952, at which time his estate was burdened with an obligation to pay his ex-wife $40 a week until her death or remarriage. Based upon her life expectancy this obligation required $27,058.30 to meet it. As of the date of death the decedent’s estate was worth $27,058.30 less than its value would have been had the obligation imposed by the separation agreement and divorce decree not survived decedent’s death. The Commissioner disallowed the deduction; the Tax Court allowed it primarily on the authority of the Ithaca Trust Co. decision and Estate of Pompeo M. Maresi, 6 T.C. 582, affirmed 2 Cir., 1946, 156 F.2d 929. In disallowing the deduction, the Commissioner apparently and the majority here rely upon the fortuitous circumstances that the estate tax return was not filed until July 28, 1953 and Mrs. Shively had remarried in June 1953. Therefore my colleagues say in effect: why resort to statistical probabilities when the facts are known, Mrs *376Shively actually received $2,079.96 (the $40 weekly payments) from the estate; remarriage terminated the obligation; therefore, no further claim can be made. The word “fortuitous” is used because had the tax return been filed and audited prior to June 1953, there would have been no basis to challenge the actuarially computed deduction. The same result would have obtained had Mrs. Shively not remarried for a few years after final audit. Thus only in the event that the courts by decision in effect enact a tax statute providing that in any estate in which the determination of values depends upon future contingencies, such estate shall be reopened for purposes of recomputation of estate taxes upon the actual happening of the contingency, can the tax ever be calculated on known facts related back to the date of death.
I had assumed the law, at least since 1929, to have been clear and definite in its relation to the problem here presented. In Ithaca Trust Co., supra, the decedent gave his estate to his wife for life, the residue to charities. The value of the taxable net estate could only be computed after ascertaining the value of the life estate because “as those gifts were subject to the life estate of the widow, of course their value was diminished by the postponement that would last while the widow lived ” (279 U.S. at page 154, 49 S.Ct. at page 291). It so happened there, as here, that the widow died before the period for filing the return had expired. The legal question was identical, namely, “whether the amount of the diminution, that is, the length of the postponement, is to be determined by the event as it turned out, of the widow’s death within six months, or by mortality tables showing the probabilities as they stood on the day when the testator died” (279 U.S. at page 155, 49 S.Ct. at page 291). Mr. Justice Holmes, writing the opinion, had the initial reaction (as almost all would) that “The first impression is that it is absurd to resort to statistical probabilities when you know the fact” (279 U.S. at page 155, 49 S.Ct. at page 291). “But [he continued] this is due to inaccurate thinking” because “The estate so far as may be is settled as of the date of the testator’s death” (279 U.S. at page 155, 49 S.Ct. at page 291). He, therefore, concluded that “Tempting as it is to correct uncertain probabilities by the now certain fact, we are of the opinion that it cannot be done, but that the value of the wife’s life interest must be estimated by the mortality tables” (279 U.S. at page 155, 49 S.Ct. at page 292).
Had the law, as expounded by the majority here, been applied, the Supreme Court would have valued the widow’s life interest on the basis of her actual life span (six months). Though tempted it decided otherwise. And wisely so because otherwise there could be no finality in estate administration. Since Mr. Justice Holmes and his brethren on the Supreme Court, after fully recognizing the problem, were able to resist temptation, fortified by this knowledge and support, and in contrast to my more yielding colleagues, I believe that the principles of the Ithaca Trust case remain sound law. In the field of estate tax law it is particularly important that there be as much certainty as possible and that precedents be followed instead of overturned. Wills and trusts are drafted and estates are administered in reliance upon these precedents. They should be honored.
The argument made that section 812 (b) permits deduction only of claims allowed by the laws of the jurisdiction under which the estate is being administered (Connecticut) is wholly irrelevant to the problem before us. The widow was under no duty to present a claim in the Connecticut probate proceeding. Her rights were fixed before the decedent’s death by contract and decree. The only question was how much should the administrator deduct to fulfill that commitment. Mortality tables indicated $27,-058.30. The widow was not forced to prove her claim for the number of weeks *377between decedent’s death and her remarriage. It was then currently being met.
“Deductibility is not conditioned on a claim’s allowance by a local court, but rather upon its enforceability under local law” (Smyth v. Erickson, 9 Cir., 1955, 221 F.2d 1, 3). No suggestion is made by the government that the estate’s obligation was not enforceable in Connecticut. As pointed out in the Smyth case, supra, “The use of the term ‘allowed’ in section 812(b) (3) is not to be construed as meaning that unless a claim against the estate has been allowed by the state court no deduction therefor will be permitted” (at page 3). The majority in effect concedes that the obligation had “vitality as of date of death” but substitutes the uncertain date of the filing of the return as the determinative date for a re-examination as to vitality. Furthermore, I am at a loss to understand the majority’s conclusion that “We hold that where, prior to the date on which the estate tax return is filed, the total amount of a claim against the estate is clearly established under state law, the estate may obtain under Section 812(b) (3) no greater deduction than the established sum, irrespective of whether this amount is established through events occurring before or after the decedent’s death.” There is nothing whatsoever in the record that a claim for $2,079.96 or any other amount was “established under state law.” Nor is there any basis in law for the proposition that the value of an obligation created by contract and decree during lifetime may upon death be established by “events occurring before or after the decedent’s death.” Certainly the cases cited in support of such a principle do not so hold.
For these reasons I would affirm the decision of the Tax Court.