Davis, Judge,
delivered the opinion of the court:
The stipulated facts in this tax refund suit thrust upon us a narrow but knotty issue of estate tax law under Section 2036(a)(2) of the Internal Revenue Code of 1954.1 In 1961 Marian B. Farrel established an irrevocable trust with a corpus of various securities and her grandchildren as beneficiaries. Two individuals were named as trustees. They were given discretionary power to pay or apply all or part of the net income or principal to or for the benefit of any one or more of the beneficiaries (and their issue). The instrument also provided for a "time of division” when the [624]*624corpus was to be divided into various portions, each of which (according to specified circumstances) was either to be paid over immediately to a specified beneficiary, or held in a new trust with the trustees having discretionary power to make payments to or for the benefit of specified beneficiaries until a later time when required payments were to be made. No provision was made in the trust for any distribution to Mrs. Farrel in any circumstances.
The trust called for two trustees at all times, and provided for Mrs. Farrel to appoint a successor trustee if a vacancy occurred in that position through death, resignation or removal by a proper court for cause. However, neither the instrument nor Connecticut law (which governed the trust) permitted Mrs. Farrel to remove a trustee and thereby create a vacancy. The trust was silent as to whether Mrs. Farrel could appoint herself as a successor trustee in the event of a vacancy, but neither the trust instrument nor Connecticut law would have prevented her from doing so.
Two vacancies occurred in the office of trustee during Mrs. Farrel’s life. In 1964 a named trustee died and Mrs. Farrel appointed a third person as successor trustee. In 1965 that successor trustee resigned and Mrs. Farrel, as settlor, appointed another individual to succeed him.
Mrs. Farrel died in October 1969. Her estate, plaintiff here, filed in 1971 a federal estate tax return which did not include the trust property in the gross estate, and paid the tax shown on the return.2 In 1973 the Internal Revenue ■Service assessed a deficiency on the ground that the trust property should have been included in the gross estate under Section 2036(a)(2) of the 1954 Code.3 Plaintiff paid [625]*625the deficiency, filed a timely refund claim, and after the appropriate waiting period instituted the present refund suit.
Both parties agree that (a) the trustees had "the right, either alone or in conjunction with any person, to designate the persons who shall possess or enjoy the property or the income therefrom” within the meaning of Section 2036(a)(2);4 (b) Mrs. Farrel, the decedent-settlor, could lawfully designate herself (under the trust and Connecticut law) as successor trustee if a vacancy occurred during her life; (c) the occurrence of a vacancy in the office of trustee was a condition which Mrs. Farrel could not create and which was beyond her control; and (d) Mrs. Farrel had the opportunity, before her 1969 death, to appoint a successor trustee only during the two periods in 1964 and 1965 mentioned above. The legal conflict is whether the right of the trustees (as to who should enjoy or possess the property or income) should in these circumstances be attributed to the decedent under § 2036(a) for any of the three periods designated in that statutory provision — her life; any period not ascertainable without reference to her death; any period which does not in fact end before her death. The Government’s answer is yes and the plaintiff of course says no.
Only Section 2036(a) is now before us but, since taxpayer’s presentation emphasizes a comparison of that provision with Section 2038 (a cognate but separate part of the estate tax), it is important to set out, at the beginning, the relevant aspects of the latter, as we do in the margin.5 [626]*626Plaintiffs primary point is that (i) it is now and has long been settled that Section 2038 does not cover a power or right subject to a conditional event which has not occurred prior to and does not exist at the decedent’s death, such as a discretionary power to distribute, income or principal under specified conditions which have not occurred before the death, and (ii) the same rule has been and is applicable to Section 2036(a).
There is no question that taxpayer is correct as to the construction of Section 2038. That slant was given by the courts to the provision’s predecessor under the 1939 Code (see Jennings v. Smith, 161 F.2d 74, 77-78 (2d Cir. 1947); Estate of Want v. Commissioner, 29 T.C. 1223 (1958), rev’d on other grounds, 280 F.2d 777 (2d Cir. 1960); Estate of Kasch v. Commissioner, 30 T.C. 102 (1958)), and the Treasury has itself adopted the same interpretation for the 1954 Code as well. Treasury Regulations on Estate Tax (1954 Code), Section 20.2038-1(a) and (b); see also Rev. Rul. 55-393, 1955-1 Cum. Bull. 448.
The initial and fundamental question we have to face is whether this settled understanding of Section 2038 necessarily governs Section 2036(a), as it now stands. We think not for two reasons which we shall consider in turn: first, that the critical points-of-view of the two provisions differ, and, second, that the regulations governing the two sections take diametrically opposed positions on the narrow issue of contingent rights and powers of the kind involved here.
The two separate provisions appear to diverge sharply in their perspective — the point from which the pertinent powers and rights are to be seen. Section 2038(a) looks at the problem from the decedent’s death — what he can and cannot do at that specific moment. Excluded are contingent rights and powers (beyond the decedent’s control) which are not exercisable at that moment because the designated contingency does not exist at that time. Section 2036(a), on [627]*627the other hand, looks forward from the time the decedent made the transfer to see whether he has retained any of the specified rights "for his life or for any period not ascertainable without reference to his death or for any period which does not in fact end before his death.” This language makes the transferor’s death one pole of the specified time-span but the whole of the time-span is also significant. Because of the statute’s reference to the time-span, differences of interpretation are quite conceivable. It is possible, for instance, to hold the words to mean that the retained right has to exist at all times throughout one of the periods, but it is also possible to see the language as covering contingencies which could realistically occur at some separate point or points during the designated periods — always including the moment of decedent’s death. We take it (from the argument’s insistence on the parallel to 2038) that that taxpayer would not stand on the former ("at all times”) interpretation if a vacancy in the trusteeship existed and had not been filled at Mrs. Farrel’s death. But under the language of 2036(a) there is no compelling reason why the moment of death has to be exclusively important.
Free access — add to your briefcase to read the full text and ask questions with AI
Davis, Judge,
delivered the opinion of the court:
The stipulated facts in this tax refund suit thrust upon us a narrow but knotty issue of estate tax law under Section 2036(a)(2) of the Internal Revenue Code of 1954.1 In 1961 Marian B. Farrel established an irrevocable trust with a corpus of various securities and her grandchildren as beneficiaries. Two individuals were named as trustees. They were given discretionary power to pay or apply all or part of the net income or principal to or for the benefit of any one or more of the beneficiaries (and their issue). The instrument also provided for a "time of division” when the [624]*624corpus was to be divided into various portions, each of which (according to specified circumstances) was either to be paid over immediately to a specified beneficiary, or held in a new trust with the trustees having discretionary power to make payments to or for the benefit of specified beneficiaries until a later time when required payments were to be made. No provision was made in the trust for any distribution to Mrs. Farrel in any circumstances.
The trust called for two trustees at all times, and provided for Mrs. Farrel to appoint a successor trustee if a vacancy occurred in that position through death, resignation or removal by a proper court for cause. However, neither the instrument nor Connecticut law (which governed the trust) permitted Mrs. Farrel to remove a trustee and thereby create a vacancy. The trust was silent as to whether Mrs. Farrel could appoint herself as a successor trustee in the event of a vacancy, but neither the trust instrument nor Connecticut law would have prevented her from doing so.
Two vacancies occurred in the office of trustee during Mrs. Farrel’s life. In 1964 a named trustee died and Mrs. Farrel appointed a third person as successor trustee. In 1965 that successor trustee resigned and Mrs. Farrel, as settlor, appointed another individual to succeed him.
Mrs. Farrel died in October 1969. Her estate, plaintiff here, filed in 1971 a federal estate tax return which did not include the trust property in the gross estate, and paid the tax shown on the return.2 In 1973 the Internal Revenue ■Service assessed a deficiency on the ground that the trust property should have been included in the gross estate under Section 2036(a)(2) of the 1954 Code.3 Plaintiff paid [625]*625the deficiency, filed a timely refund claim, and after the appropriate waiting period instituted the present refund suit.
Both parties agree that (a) the trustees had "the right, either alone or in conjunction with any person, to designate the persons who shall possess or enjoy the property or the income therefrom” within the meaning of Section 2036(a)(2);4 (b) Mrs. Farrel, the decedent-settlor, could lawfully designate herself (under the trust and Connecticut law) as successor trustee if a vacancy occurred during her life; (c) the occurrence of a vacancy in the office of trustee was a condition which Mrs. Farrel could not create and which was beyond her control; and (d) Mrs. Farrel had the opportunity, before her 1969 death, to appoint a successor trustee only during the two periods in 1964 and 1965 mentioned above. The legal conflict is whether the right of the trustees (as to who should enjoy or possess the property or income) should in these circumstances be attributed to the decedent under § 2036(a) for any of the three periods designated in that statutory provision — her life; any period not ascertainable without reference to her death; any period which does not in fact end before her death. The Government’s answer is yes and the plaintiff of course says no.
Only Section 2036(a) is now before us but, since taxpayer’s presentation emphasizes a comparison of that provision with Section 2038 (a cognate but separate part of the estate tax), it is important to set out, at the beginning, the relevant aspects of the latter, as we do in the margin.5 [626]*626Plaintiffs primary point is that (i) it is now and has long been settled that Section 2038 does not cover a power or right subject to a conditional event which has not occurred prior to and does not exist at the decedent’s death, such as a discretionary power to distribute, income or principal under specified conditions which have not occurred before the death, and (ii) the same rule has been and is applicable to Section 2036(a).
There is no question that taxpayer is correct as to the construction of Section 2038. That slant was given by the courts to the provision’s predecessor under the 1939 Code (see Jennings v. Smith, 161 F.2d 74, 77-78 (2d Cir. 1947); Estate of Want v. Commissioner, 29 T.C. 1223 (1958), rev’d on other grounds, 280 F.2d 777 (2d Cir. 1960); Estate of Kasch v. Commissioner, 30 T.C. 102 (1958)), and the Treasury has itself adopted the same interpretation for the 1954 Code as well. Treasury Regulations on Estate Tax (1954 Code), Section 20.2038-1(a) and (b); see also Rev. Rul. 55-393, 1955-1 Cum. Bull. 448.
The initial and fundamental question we have to face is whether this settled understanding of Section 2038 necessarily governs Section 2036(a), as it now stands. We think not for two reasons which we shall consider in turn: first, that the critical points-of-view of the two provisions differ, and, second, that the regulations governing the two sections take diametrically opposed positions on the narrow issue of contingent rights and powers of the kind involved here.
The two separate provisions appear to diverge sharply in their perspective — the point from which the pertinent powers and rights are to be seen. Section 2038(a) looks at the problem from the decedent’s death — what he can and cannot do at that specific moment. Excluded are contingent rights and powers (beyond the decedent’s control) which are not exercisable at that moment because the designated contingency does not exist at that time. Section 2036(a), on [627]*627the other hand, looks forward from the time the decedent made the transfer to see whether he has retained any of the specified rights "for his life or for any period not ascertainable without reference to his death or for any period which does not in fact end before his death.” This language makes the transferor’s death one pole of the specified time-span but the whole of the time-span is also significant. Because of the statute’s reference to the time-span, differences of interpretation are quite conceivable. It is possible, for instance, to hold the words to mean that the retained right has to exist at all times throughout one of the periods, but it is also possible to see the language as covering contingencies which could realistically occur at some separate point or points during the designated periods — always including the moment of decedent’s death. We take it (from the argument’s insistence on the parallel to 2038) that that taxpayer would not stand on the former ("at all times”) interpretation if a vacancy in the trusteeship existed and had not been filled at Mrs. Farrel’s death. But under the language of 2036(a) there is no compelling reason why the moment of death has to be exclusively important. Unlike Section 2038, this provision seems to look forward from the time of transfer to the date of the transferor’s death, and can be said to concentrate on the significant rights with respect to the transferred property the transferor retains, not at every moment during that period, but whenever the specified contingency happens to arise during that period (so long as the contingency can still occur at the end of the period).
There is nothing unreasonable about this latter construction, which accords with Congress’ over-all purpose to gather into the estate tax all transfers which remain significantly incomplete — on which the transferor still holds a string — during his lifetime. It is hard to believe, for instance, that, whatever may be true of 2038, 2036(a) would have to be seen as failing to cover a trust where the trustee, with discretionary powers, could be removed by the settlor, and the settlor substituted as trustee, whenever economic conditions fell below a stated level (e.g., a designated level on a certain stock exchange index or a [628]*628level of earnings of the trust) even though fortuitously that condition did not happen to exist at the time of death. In a case like that, the lifetime link between the decedent and the trust property (and income) would be so strong as plainly to measure up to both the letter and the spirit of 2036(a) if the Treasury chose to see it that way. This case, though perhaps less clear, falls into the same class of a continuing substantial tie.
The other element which leads us to reject plaintiffs attempt to equate 2036(a) with 2038, for this case, is that the Treasury has affirmatively chosen to separate the two sections — there is a Treasury regulation under the former § 20.2036-1 which, to our mind, clearly covers this decedent’s situation (in contrast to the regulation under 2038 which excludes it).6 Taxpayer urges us to read the regulation otherwise, and if we cannot to hold it invalid.
The regulation says flatly (see footnote 6, supra) that it is immaterial "(ni) whether the exercise of the power was subject to a contingency beyond the decedent’s control which did not occur before his death (e.g., the death of another person during the decedent’s lifetime).” This would seem on its surface to blanket this decedent’s position under her trust, but plaintiff would read it very literally and narrowly to apply only where the contingency relates [629]*629to the "exercise” of an already existing power, and conversely, to be inapplicable where the power only springs into existence when a trustee vacancy occurs.7 Similarly, taxpayer sees in the broad sweep of the last sentence of § 20.2036-1(b)(3) (fn. 6) the implied negative pregnant that a restricted power in the decedent to appoint herself a substitute trustee only in the event of a vacancy lies outside 2036(a). We cannot accept these strained (if not casuistic) analyses of the regulation because they go directly counter to its apparent purpose to cover just such contingencies as we have here. If proof of that objective is needed it is fully supplied by the companion regulation under 2038 (Treasury Regulation on Estate Tax (1954 Code), § 20.2038-l(b)) which declares in coordinate terms that "section 2038 is not applicable to a power the exercise of which was subject to a contingency beyond the decedent’s control which did not occur before his death (e.g., the death of another person during the decedent’s life). See, however, Section 2036(a)(2) for the inclusion of property in the decedent’s gross estate on account of such a power” (emphasis added).
We are required, then, to consider whether § 20.2036-1(b)(3) should be overturned as invalid. Recognizing the deference due Treasury Regulations (Commissioner v. South Texas Lumber Co., 333 U.S. 496, 501 (1948); Bingler v. Johnson, 394 U.S. 741, 749-51 (1969)), we cannot take that step. We have pointed out that 2036 is not the same as 2038 in its wording or in the viewpoint from which it appraises the decedent’s link to the transferred property. We have also said that it is not unreasonable to regard 2036(a), in the way the Treasury does, as a blanket overall sweeping-in of property over which the decedent still has at [630]*630death some significant, though contingent, power to choose those who shall have possession or enjoyment.
To this, plaintiffs ultimate response is double-barreled: (i) the regulation was not contemporaneous with the adoption of the predecessor of 2036 but came years later, and (ii) for taxable years before the regulation was adopted courts applied the same rule to the predecessors of 2036 as they have to 2038 and its predecessors. Both points may be technically correct. The relevant regulation was promulgated in 1958 many years after the first predecessor of 2036 was enacted. And in 1947 Jennings v. Smith, supra, seemed to use, as an alternative holding, the same rule for the forerunner of 2036 that it applied to the predecessor of 2038. To the same effect was Estate of Kasch v. Commissioner, supra (1958) (predecessor section).8 Nevertheless the Treasury was not foreclosed, in our view, from taking the position it did in 1958. As we have stressed, neither the text nor the purpose of 2036 demands that it be treated the same as 2038. The judicial observations lumping the two predecessor provisions responded to Government arguments treating the two sections as fully parallel on this point.9 Moreover, Jennings v. Smith was the only appellate decision on the point which was outstanding when the 1958 regulations were promulgated, and its statement was, at best, a rather summary alternative holding10 not truly necessary to the ultimate decision. Thus, the judicial interpretation of Section 2036’s predecessor was neither so strong nor so encrusted that Congress may be thought to [631]*631have embodied it in Section 2036 when it adopted the 1954 Code. On the contrary, we think the Treasury was free to take a new look under that new Code, as it evidently did. In sum, the current regulation under 2036 is a reasonable reading of the text and objective of the statute and the previous contrary judicial interpretation was not such that the Treasury was forced to bow to it despite the prima facie acceptability of the administrative stance.
We end by noting that the contingent right of Mrs. Farrel to make herself a trustee in the event of a vacancy— unlike the de facto "powers” involved in United States v. Byrum, 408 U.S. 125 (1972) and in Estate of Tully v. United States, 208 Ct. Cl. 596, 528 F.2d 1401 (1976)—was a legally enforceable right, in effect imbedded in the trust instrument, which bore directly on the designation of the persons to possess or enjoy the trust property or income. That the exercise of this right was foreseeable when the trust was created — that it was a real right, neither insignificant nor illusory — is shown by the fact that Mrs. Farrel had two opportunities to exercise it in eight years and, if she had lived, may well have had more.
CONCLUSION
For these reasons, we hold that plaintiff is not entitled to recover and the petition is dismissed.