504 F2d 586 Continental Illinois National Bank and Trust Company of Chicago v. United States

504 F.2d 586

74-2 USTC P 13,034

CHICAGO, as Executor of the Estate of Josephine W.
Speth, Deceased, Plaintiff-Appellee,
UNITED STATES of America, Defendant-Appellant.

Nos. 73-1967, 73-2077.

United States Court of Appeals, Seventh Circuit.

Argued April 16, 1974.
Decided Oct. 7, 1974.

Scott P. Crampton, Asst. Atty. Gen., David English Carmack, Atty., Tax Div., Dept. of Justice, Washington, D.C., James R. Thompson, U.S. Atty., and Gary L. Starkman, Asst. U.S. Atty., Chicago, Ill., for defendant-appellant.

Leonard R. Hartenfeld, David D. Rosenstein, Chicago, Ill., for plaintiff-appellee.

Before FAIRCHILD, PELL and STEVENS, Circuit Judges.

PELL, Circuit Judge.

view counter

The Continental Illinois National Bank and Trust Company of Chicago brought this action as executor of the estate of the decedent Josephine W. Speth to recover an alleged overpayment of estate tax and interest.1 After a bench trial, the district court entered judgment in favor of the taxpayer from which the Government has appealed.


The decedent Josephine Speth died on March 28, 1966, at the age of 75. Her sister, Margaret Speth, had died one month earlier on February 21, 1966. By her will, Margaret bequeathed to the decedent a life estate in certain property. At the time of Margaret's death, Josephine was suffering from cancer of the colon with metastasis to the liver.


The taxpayer, in computing the federal estate tax on Josephine's estate, claimed a credit under 2013 of the Internal Revenue Code of 1954 for the amount of federal estate tax paid by the estate of Margaret on the life estate transferred to Josephine. In calculating the amount of this credit, the taxpayer used the actuarial tables set forth in 20.2031-7(f) of the Treasury Regulations to value Josephine's life estate at the time of Margaret's death. The indicated table expectancy was approximately six years. The Commissioner disallowed the taxpayer's computation of the credit on the ground that, under Revenue Ruling 66-307, Josephine's actual life expectancy had to be utilized in determining the value of her life estate.2


The district court held, alternatively, that (1) the Revenue Ruling did not control this case, and (2) the Revenue Ruling was invalid. The Government contests both holdings on appeal.


* Section 2013 of the Internal Revenue Code was enacted 'to prevent the diminution of an estate by the imposition of successive taxes on the same property within a brief period.' Sen.Rep.No. 1622, 83rd Cong., 2dSess. 121 (1954), U.S.Code Cong. & Admin.News, 1954, pp. 4629, 4755. The section provides for a credit against estate tax where the decedent has received property in a transfer during the previous ten years which transfer was itself subject to estate tax.3 With certain adjustments and limitations not relevant here, the tax credit is the amount which bears the same ratio to the total estate tax paid by the transferor as the value of the property transferred bears to the total value of the transferor's taxable estate. If the transferred property his no value, the credit will, therefore, be zero. The percentage of credit allowed to the transferee's estate decreases ratably over the ten-year period.4


The credit applies to any beneficial interest in property received by the transferee. 2013(e). Thus, the credit is available for a life estate held by the decedent-transferee, even though the life estate will not be included in his estate and, therefore, cannot itself be the subject of double taxation. See Treas.Reg. 20.2013-1(a) and 20.2013-5(a).5


Section 2013(d) provides, in general, that 'the value of property transferred to the decedent shall be the value used for the purpose of determining the Federal estate tax liability of the estate of the transferor' subject to certain adjustments not relevant here. The Code itself does not, however, set forth the manner for computing the value of a life estate in property transferred to the decedent when the property was taxed in the estate of the transferor. The Treasury Regulations state, inter alia, at 20.2013-4(a), that:

view counter

'If the decedent received a life estate or remainder or other limited interest in property included in the transferor's gross estate, the value of the interest is determined as of the date of the transferor's death on the basis of recognized valuation principles (see especially 20.2031-7 and 20.2031-10).'6


Section 20.2031-7 of the Regulations contains the mortality tables for valuing life estates.7


The Government admits that, as a general rule, the 2013 credit on a life estate is to be calculated by finding the actuarial life expectancy, as determined by the tables in Treasury Regulation 20.2031-7, of the life tenant as of the transferor's death. The Government contends, however, that there is an exception to this general rule and this exception is embodied in Revenue Ruling 66-307.



Revenue Ruling 66-3078 is predicated upon a factual situation 'where it was known at the death of the transferor that the life tenant, afflicted with a ravaging and incurable disease of advanced state, could not survive for more than a year.' In this situation, the Revenue Ruling directs that, for the purpose of calculating a 2013 credit, the life tenant's actual life expectancy, rather than his actuarial life expectancy, as of the transferor's death, be used in valuing the life estate. The Revenue Ruling goes on to state that, as a general rule, such a departure from the mortality tables is proper, for the purpose of 2013, 'if it is known on the valuation date that a life tenant is afflicted with a fatal and incurable disease in its advanced stages, and that he cannot survive for more than a brief period of time.'


The Government contends that this Revenue Ruling covers the present case and that, therefore, Josephine's actual life expectancy as of her sister's death should be used in valuing her life estate. The taxpayer, on the other hand, argues that Revenue Ruling 66-307 is invalid and, even if it is valid, it does not cover the present case.


Forceful arguments have been made by both of the litigants with respect to the issue of the validity of the Revenue Ruling. We need not, however, resolve this matter in order to decide this case since, in our opinion, even assuming arguendo that the Revenue Ruling is valid, the present case falls outside the ruling.9


In reaching this conclusion, we are assuming that words used in the Revenue Ruling were employed precisely. Such an assumption is particularly warranted in the field of tax law where the decision is ordinarily determined by the language utilized and not necessarily by application of logical principles. We first note that the statement of the rule is in two-pronged form joined by the conjunctive 'and.' Thus, we conclude that it is essential not only that it be known that the disease is fatal and incurable in an advanced stage but also that the life tenant cannot survive for more than a brief time. It is at this point that we find ourselves deserted insofar as further guidance is concerned. Outside of the particular fact situation the Ruling presents, we are not advised as to the meaning of 'brief,' or whether it is an absolute or relative term.


The Government, in its brief, states that departure from the tables is proper if the actual life expectancy is 'considerably shorter' or 'far less' than that predicted in the mortality tables. Under the Government's interpretation of 'brief' as a relative term, a person with a fatal, incurable disease in its advanced stages who had an actuarial life expectancy of 25 years but an actual life expectancy of 5 years would apparently have only a 'brief' life span and, therefore, departure from the tables would be proper. As an absolute, as opposed to a relative matter, two years might be deemed by some to be a 'brief' life span. Thus, some might consider that a failing nonagenarian whose life expectancy, according to the mortality tables, was approximately two years and who had an actual life expectancy of only slightly less than this actuarial expectancy, had only a 'brief' life expectancy. Nevertheless, we doubt that the Government would contend, in this situation, that the 'brief' life expectancy (in absolute terms) should require departure from the actuarial tables.


Turning to source books such as Words and Phrases and Black's Law Dictionary, we find no definitions of 'brief' in the time sense.10 Webster tells us that 'brief' in its duration meaning is synonymous with 'short.'11 Accepting that definition we construe the Revenue Ruling to be applicable only to a situation in which it is known that death will follow closely after a crucial date. Moreover, although a court might well wish to consider the actuarial life expectancy in determining whether the life tenant's actual life expectancy is 'brief,' we construe 'brief,' as used in the Revenue Ruling, to be essentially an absolute, not a relative, term. We are fortified in this interpretation by the prefatory synopsis opsis of the Revenue Ruling which speaks in terms of 'where the death of a life tenant from known causes is predictable and imminent on the valuation date.' 'Imminent,' in the time sense, appears unlike 'brief' to have been the subject of considerable judicial attention. 20 Words and Phrases 215 (1959). Without citing individual cases, it is clear that 'imminent' refers to something which is threatening to happen at once, something close at hand, something to happen upon the instant, close although not yet touching, and on the point of happening. With this construction of the Revenue Ruling in mind, we turn to the evidence adduced in this case.


According to the actuarial tables, Josephine's life expectancy, at the time of her sister's death, was six years. The medicla evidence indicated, however, that, at that time, it was known that Josephine was suffering from cancer of the colon which had metastasized to the liver. There was no dispute over the fact that this was a fatal and incurable disease.12


The question of the decedent's actual life expectancy, due to this disease, however, was disputed at the trial. Dr Myles Cunningham, a cancer expert presented by the taxpayer, testified that, at the time of Margaret's death,13 Josephine did not have any of the clainical signs associated with liver metastasis which would indicate an imminent death. Absent such clinical signs of imminent death, a physician could not predict with reasonable certainty, according to Dr. Cunningham, how long a patient with liver metastasis would live. In particular, Dr. Cunningham noted, 'for one patient in the condition that Hosephine Speth was in in January of 1966, I don't believe you can place a time span on her expected lenght of life.' Dr. Cunningham further stated that there was a reasonable chance that Josephine could have lived a year or more after Margaret's death, that it was possible that she could have lived 18 months, and that, although it was unlikely, it was not impossible for her to live two years.14 The doctor noted that he himself had 'had at least one case who has lived six years now with liver metastasis.'15


Dr. Samuel Taylor, a cancer expert presented by the Government, testified that he would have expected Josephine to live four to six months after February 21, 1966. Dr. Taylor admitted, however, that she possibly could have lived for a year. He further noted that 'in medicine you can't be too didactic, because someone always surprises you.' Finally, Dr. Taylor agreed with Dr. Cunningham that spontaneous remission was a medical fact and could occur in any cancer patient prior to the emergence of clinical signs of imminent death.


The third doctor to testify, Dr. Kenneth Kettleson, was the decedent's physician. Dr. Kettleson was an internist but not a cancer expert. He stated that, in his opinion, Josephine's life expectancy at the time in question was between one and six months.


All three physicians agreed that, on the basis of the autopsy report, cancer may not have been the actual cause of Josephine's death but rather that her death could have been caused by a heart attack or a blood clot in her lung.


The district court found that the present case was distinguishable from the fact situation presented in the Revenue Ruling in that it was possible that, on February 21, 1966, Josephine could have lived for another year. Based on the testimony of Dr. Cunningham, we cannot say that this finding was clearly erroneous. Rule 52(a), Fed.R.Civ.P.; Commissioner v. Duberstein, 363 U.S. 278, 291, 80 S.Ct. 1190, 4 L.Ed.2d 1218 (1960). To the extent that the doctors disagreed about Josephine's actual life expectancy on February 21, 1966, this was a question of credibility to be determined by the trail judge. Zenith Radio Corp. v. Hazeltine Research, Inc., 395 U.S. 100, 123, 89 S.Ct. 1562, 23 L.Ed.2d 129 (1969); Aunt Mid, Inc. v. Fjell-Oranje Lines, 458 F.2d 712, 716 (7th Cir. 1972), cert. denied, 409 U.S. 877, 93 S.Ct. 130, 34 L.Ed.2d 131.


Moreover, we agree with the district court that the general rule which the Revenue Ruling purports to establish (i.e., that the taxpayer should use the actual life expectancy of the decedent where it is 'known' on the valuation date that the life tenant is afflicted with a fatal and incurable disease in its advanced stages, and that he cannot survive for more than a 'brief' period of time) does not cover this case. The testimony of the experts indicated that there was a range of longevity among patients with liver metastasis. According to Dr. Cunningham, this range could, for some patients, extend at least as far as six years. Where a particular patient fell within this range of longevity could not, according to Dr. Cunningham, be predicted, absent clinical signs of imminent death. In view of this testimony and the fact that the life expectancy predicted by the mortality tables for the decedent was itself six years, we cannot say that it was 'known' that Josephine's actual life expectancy was so 'brief' as to require a departure from the tables.


In addition to the Revenue Ruling, the Government relies on a number of cases (most of which concern Code sections other than 2013), in which departures from the actuarial tables have been permitted for the purpose of valuing life estates. The case before us is, however, readily distinguishable from these cases.


In the majority of these cases permitting departures from the tables, the life tenant's maximum actual life expectancy was one year or less. In Estate of Nellie H. Jennings, 10 T.C. 323 (1948), for instance, the Tax Court held that the life tenant's actual life expectancy should be used in valuing a charitable remainder where, at the time of the decedent-transferor's death, the life tenant was almost totally paralyzed, had a complete loss of memory and was described by his physician as 'just an individual protoplasm.' 10 T.C. at 327. The life tenant in Jennings had an actual life expectancy, according to the medical experts, of 'not more than one year' from the date of the transferor's death. 10 T.C. at 327. Similarly, in Estate of Nicholas M. Butler, 18 T.C. 914 (1952), also a charitable remainder case, a departure from the mortality tables was permitted to value a life estate where the life tenant had cancer and ' the facts in existence at the time of (the transferor's) death were such as to render it certain that (the life tenant) would not live more than one year after (the transferor's) death.' 18 T.C. at 919-920.16


In Estate of Lion .v Commissioner, 438 F.2d 56 (4th Cir. 1971), cert. denied, 404 U.S. 870, 92 S.Ct. 52, 30 L.Ed.2d 114, the Fourth Circuit applied the actual life expectancy, rather than the actuarial life expectancy, of a life tenant in determining a 2013 credit. Lion, however, involved a situation in which the transferor and the life tenant died simultaneously.


The cases permitting the substitution of actual life expectancy for actuarial life expectancy are typically characterized as 'exceptional' cases. Miami Beach First Nat'l Bank v. United States, 443 F.2d 116, 120 (5th Cir. 1971); Estate of Lion v. United States, supra at 62; Estate of Carl E. Weller,38 T.C. 790, 803 (1962). Absent 'exceptional' circumstances, the actuarial tables should be used since these tables 'afford a reasonable norm and some degree of certainty in ascertaining the value of property and the consequent tax liabilities of the beneficiaries thereof.' Miami Beach First Nat's Bank v. United States, supra at 119. See Simpson v. United States, 252 U.S. 547, 550, 40 S.Ct. 367, 64 L.Ed. 709 (1920). Although the tables prove to be accurate when used in large numbers of cases, discrepancies are bound to occur in individual cases.17 But, as the First Circuit has noted, 'the discrepancies may have to be suffered in the interest of a simplified overall administration of the tax laws.' McMurtry v. Commissioner, 203 F.2d 659, 667 (1st Cir. 1953). Departures from the tables are permitted, if at all, only where the result of using the tables is 'unrealistic and unreasonable,' Estate of Carl E. Weller, supra at 803; Estate of Chauncey Stillman, 24 T.C.M. 478, 497 (1965), or 'ignores . . . common sense.' Hall v. United States, 353 F.2d 500, 505 (7th Cir. 1965).


Moreover, although the taxpayer has the burden of showing the error in the Commissioner's determination, where the taxpayer proves that normally the actuarial tables provided in the Regulations would be applicable to his case, the burden is on the Government to show that the case is 'exceptional,' justifying a modification of or departure from the prescribed method. Estate of Carl E. Weller, supra at 803; Estate of Chauncey Stillman, supra at 497; Mercantile-Safe Deposit & Trust Co. v. United States, 368 F.Supp. 743, 746 (D.Md.1974).


The mere fact that the life tenant is suffering from an incurable and fatal disease is not necessarily sufficient to justify a departure from the actuarial tables. In Estate of Chauncey Stillman, supra, the taxpayer transferred a contingent remainder interest to a trust for the benefit of his daughters. In valuing the gift for gift tax purposes, the taxpayer determined the life expectancy of the life tenant by means of the actuarial tables provided in the Regulations. The Government contended, however, that the actual life expectancy of the life tenant should be used since the life tenant was afflicted, at the time of the gift, with multiple myeloma, a fatal and incurable type of cancer. The tax court noted the expert testimony that the disease, although fatal, was variable among patients and predictions as to indvidual longevity were therefore impossible. The court held in Stillman that departure from the tables would be improper in this situation.


'(The Commissioner) relies on the Jennings rule. Since the ultimate finding is that on the gift date, the life expectancy of the life tenant was indeterminate and could not be foretold, (the Commissioner's) method of valuation falls. On the evidence, this case is distinguishable from Jennings and the other cases in the Jennings group and the Jennings rule does not apply . . .. In this instance, departure from the method of valuation prescribed in section 25.2512-5(e) is neither justified nor required . . .. (The Commissioner's) determination was little better than a guess which cannot be approved.' 24 T.C.M. at 502-503.


See also, Miami Beach First Nat'l Bank v. United States, supra; Mercantile-Safe Deposit & Trust Co. v. United States, supra.


In the present case, the Government has failed to show that, assuming arguendo a departure from the prescribed actuarial tables may be proper in certain 2013 cases, the circumstances here justify such a departure from the prescribed method. Rather, as in Estate of Chauncey Stillman, supra, any determination of Josephine's actual life expectancy as of her sister's death would be 'little better than a guess.'18 The physicians who testified were candid in admitting their Sybyllic efforts of necessity lacked exactitude. In this situation, departure from the tables is not justified.




The taxpayer paid estate tax in the amount of $256,819.51. The District Director of the Internal Revenue Service assessed an estate tax deficiency in the amount of $33,950.52 plus interest. The taxpayer paid the assessed tax and interest and timely filed a claim for refund


'Actual life expectancy' means the life tenant's life expectancy at the valuation date as predicted on the basis of known facts concerning that individual rather than on the basis of actuarila tables. 'Actual life expectancy,' being predictive, may be, of course, more or less than the life tenant's actual life span


A credit is also allowed where the decedent predeceased his transferor by less than two predeceased his transferor by less than two years


One commentator has noted that 2013 'can . . . be characterized as the most complex and confusing provision of the estate tax law.' Rudick, The Estate Tax Credit for Tax on Prior Transfers, 13 Tax.L.Rev. 3 (1957). Without examining all the estate tax sections, we are inclined to agree


The Congressional rationale for permitting a credit for terminable interests was apparently that 'the terminable interest may have produced income (presumably during the life tenant's life time) that might be taxable in the decedent's estate. Sen.Rep.No.1622, supra at 122. A second possible rationale, suggested by one commentator, is that possession of the life estate may enable the life tenant to conserve other assets which otherwise he would have expended, and which subsequently are included in his gross estate. Kirby, Proposed Estate Tax Regulations, 96 Trusts & Estates 12 (1957)


In determining a 2013 credit, it is necessary to look at the dollar value of the property as it was taxed in the prior estate since the credit in the second estate is founded on the fact that the property was taxed at a specified dollar amount in the prior estate. However, unlike a fee simple interest in property, the value of a life estate is based not only on the value of the property on which the life estate is imposed but of necessity is also dependent upon the extent of the period of enjoyment. Some conceptual difficulties might arise as to the rationale for utilization, in 2013, of the predictive process as to the number of years the life tenant will live since in the case of the life estate, 2013 comes into play only after the transferee-life tenant has died, when the extent of enjoyment is a known fact

This is not situation in which the life expectancy of a transferee-life tenant must be predicted in order to calculate the estate taxes of the transferor (e.g., the charitable remainder cases). Nor is it a case in which the 'taxable event' clearly occurred at a particular date prior to the transferee's death and the transaction, for valuation purposes, 'closed' as of that earlier date (e.g., a transfer for full and adequate consideration: see United States v. Righter, 400 F.2d 344 (8th Cir. 1968)).

According to the Senate Report, 'to eliminate the tracing (of the transferred property) the credit is based upon the value of the property at the time of the death of the prior decedent.' Sen.Rep.No.1622, supra at 121, U.S.Code Cong. & Admin.News 1954, p. 4755. Under 812 of the Internal Revenue Code of 1939, the transferee was permitted a deduction from the gross estate for taxes paid by the transferor and the amount of the deduction was computed by tracing the transferred property in the hands of the transferee.

Regardless of the rationale, it is clear that in valuing a 2013 credit on a life estate, the value is to be determined as of the date of the transferor's death and that the valuation process is not aided by the hindsight provided by the actuality of duration.


Section 20.2031-10 of the Regulations covers the valuation of limited property interests for decedents dying after December 31, 1970


1966-2 Cum.Bull. 429


The decision of this court in Hall v. United States, 353 F.2d 500 (7th Cir. 1965), is not controlling with respect to the issue of the validity of the Revenue Ruling. In Hall, this court held that for the purpose of 2037(b) of the Internal Revenue Code of 1954 (pertaining to reversionary interests), taxpayers are not required to use the actuarial tables in all cases. Section 2037(b) provides, in part:

'The value of the reversionary interest . . . shall be determined . . . by usual methods of valuation, including the use of tables of mortality and actuarial principles, under regulations prescribed by the Secretary or his delegate.'

The wording of 2037 was construed in Hall to mean that the use of the actuarial tables was not mandatory. In the present case, however, neither the Code section nor the Regulation involved contains language similar to that involved in Hall.

Ignoring any dilemmatic impaling aspects arguably present, the Government in its brief, while citing Hall to support its position, states that the Commissioner does not agree with the result reached by this court in Hall, as 'Congress obviously contemplated the use of mortality tables to determine life expectancy under Section 2037.' The Government does not deem this obviousness to be present in Section 2013. In Hall, the use of the mortality tables would have been to the detriment of the taxpayer; in the present case, the converse is, of course, true.


The judicial definitions for the most part pertained to those documents filed in courts by lawyers. We scarcely need to observe that the word in this sense is frequently not associated with brevity


In Shippee v. Shippee, 122 N.J.Eq. 570, 195 A. 728 (1937), it was held that the death of a sister, 41 days after the death of the testatrix, did not occur 'shortly' after the latter's decease. In that case, the second decedent apparently had been suffering from a fatal disease for several months before the death of her sister; however, the will had been executed several years previously. We are not unmindful that terms of time must be viewed in the context in which the words are used. 'The expression 'a long time' would refer to a very different period or duration, and have a widely different meaning in measuring time when used by an archaeologist having reference to the period of existence of the Egyptian pyramids, than when used by Carolina Governors with reference to the time between drinks.' Wilson v. State, 2 Ala.App. 203, 56 So. 114, 115 (1911). We have viewed the word 'brief' in the context in which we find it


The Government agrees that the fact that the life tenant actually died within a relatively short time subsequent to the date of the transfer is, in itself, not a decisive factor. Ithaca Trust Co. v. United States, 279 U.S. 151, 49 S.Ct. 291, 73 L.Ed. 647 (1929)


The last medical examinations of Josephine, prior to her hospitalization immediately before her death in March 1966, were conducted in January 1966. All of the physicians who testified at the trial based their predictions as to Josephine's actual life expectancy on February 21, 1966 on these January reports


When asked to give his opinion as to the percentage of persons with liver metastasis who live one year or 18 months, Dr. Cunningham stated that approximately 10 to 15 percent survive one year and probably about 5 to 10 percent survive 18 months


Although it apparently played no part in the district court's decision, and we have given it no weight in our decision, a fact, of which the reading public cannot be unmindful, is that a lively discussion and considerable controversy had developed from the ability of modern medical science, by various sophisticated devices, to keep life in existence or 'vital functions (continuing) long after a patient's awareness has departed.' William D. Poe, 'Do We Need Restraint Medicine?' The Christian Century, September 19, 1973, at 914

Although we are unaware of any generic term for the particular subject, such as euthanasia in an allied area of discussion, 'prolongevity' might well be used.

The point at which death may be said to have occurred poses difficult problems with both legal and moral implications to which medical practitioners have addressed themselves, particularly as the situation bears on the transplanting of organs. 'AMA Passes 'Death with Dignity' Resolution,' Science News, December 15, 1973, at 375.

The record in this case did support the district court's conclusion, notwithstanding the Government's characterization of it as erroneous as a matter of law, that 'because of individual biological differences, it is possible to make a medically reliable prediction of longevity with relation only to a large group of cancer patients, but that such predictions are unreliable with reference to any individual patient.'


Other cases in the 'Jennings line' which have permitted departures from the mortality tables include: Estate of John P. Hoelzel, 28 T.C. 384, 388 (1957), life expectancy 'did not exceed 1 year'; Huntington Nat'l Bank, 13 T.C. 760, 765 (1949), life expectancy 'not greater than one year'; Estate of James W. Douglas, 12 T.C.M. 347 (1953), life expectancy of 'somewhere between six months and a year'; Estate of T. G. Hendrick, 9 T.C.M. 581, 583 (1950), life expectancy 'was not greater than one year.'

In Estate of John Halliday Denbigh, 7 T.C. 387 (1946) a departure from the mortality tables was permitted where the life tenant's actual life expectancy was 'a year or two'-- but this was compared with a 16 year life expectancy predicted by the mortality tables.


'The United States is in business with enough different taxpayers so that the law of averages has ample opportunity to work.' Gelb v. Commissioner, 298 F.2d 544, 552 (2d Cir. 1962)


The Government contends that the district court's decision suggests that to be able to depart from the actuarial tables, it is necessary to be able to predict the exact day upon which the decedent would die. The Government argues that a 'probable' prediction should be sufficient

Neither this decision nor those of other courts implies that, assuming departure is permitted in some cases, such departure is only justified upon a showing of the very day upon which the life tenant will die. The point is that in this case, as in Stillman, there was a range of longevity and there was no certainty where a particular patient would fall within this range, thereby rendering any determination of actual life expectancy uncertain.