November 17, 1949.
Samuel Taylor, Edgar Sinton, San Francisco, Cal., for petitioner.
Theron Lamar Caudle, Asst. Atty. Gen., Ellis N. Slack, L.W. Post and Harry Baum, Sp. Assts. to Atty. Gen., for respondent.
Before HEALY, BONE and ORR, Circuit Judges.
This case involves federal estate taxes. The question, affirmed by the taxpayer and denied by the Commissioner, is whether the Tax Court was in error in holding it proper to compute the remainder interest in a trust in conformity with existing treasury regulations.
In 1922 the decedent, Koshland, created a trust of securities in favor of his wife for the period of her life, the income to be paid her quarterly or oftener. In event the income in any year might prove less than $15,000, the trustees were authorized to resort to the principal. A son of the couple was designated remainderman. Koshland died in 1944, aged 75 years, at which time his wife was 66 years of age and in good health. At the time of the Tax Court hearing she was still living and was expected by her physician to live out her normal life expectancy. The fair market value of the trust as of the death of the decedent was $231,524.64.
Concededly the value of the remainder interest was includible in the gross estate; and in arriving thereat the Commissioner subtracted from the value of the trust property the value of the wife's life estate, calculating the latter in conformity with Table A, appearing in Treasury Regulations 105, sec. 81.10(1). This table is based on the Actuaries' or Combined Experience Table of Mortality, and is the result of experience of 17 British life insurance companies covering a period from 1762 to 1837. It shows a life expectancy in Mrs. Koshland of 10.46 years at age 66. On the basis of this table, the use of which the Tax Court held not improper, the value of the life estate was determined to be $114,530.93, leaving a value of $116,973.71 for the trust remainder. Naturally, the greater the value given the life estate, the less the value of the taxable remainder; and the taxpayer claims that another table, showing a much greater life expectancy, should have been used.
At the hearing before the Tax Court the taxpayer produced as a witness a Mr. Waites, an experienced actuary. Waites testified that the mortality table prescribed by the regulations is obsolete. It was his opinion that the 1937 Standard Annuity Table more accurately reflects present day life expectancies. Since Waites was the only witness testifying on the subject, it is contended that the Tax Court was obliged to accept and apply his conclusions. However, it is necessary to evaluate the witness' testimony as a whole, including qualifying statements and admissions on cross-examination; and a study of his testimony leaves us unconvinced, as it did the Tax Court, that the 1937 table should have been used. Nor do we think it was shown by the actuary that the employment of the mortality table set out in the regulations was arbitrary or unreasonable.
Waites testified that the 1937 Standard Annuity Table has been used by insurance companies and actuaries as a basis for determining annuities and things of that nature since 1937. The table is used for both male and female lives except that the age of the female is taken at an age several years younger than the male life, experience having demonstrated that women tend to live longer than men. For evaluating annuities it is the most current table in use. Under it a female of age 66 is given a life expectancy of 16.90 years. The Actuaries' or Combined Experience Table of Mortality is not now employed by insurance companies in computing annuities. The 1937 table is used only for annuities. Annuitants, as a rule, are a self-selected group, and the experience that insurance companies have in respect of their mortality rate is not the same experience that they have with the average run of persons. From this we understand that annuity mortality tables do not purport to reflect the general mortality experience but only the experience of a more or less limited group who, so to speak, bet against the insurance companies that they will outlive the normal expectancy.
Insurance companies, according to this witness, do not use annuity mortality tables in determining life insurance premiums or in calculating life insurance reserves. For premium purposes they use their own mortality tables based upon their individual experience. The mortality table currently in use in a majority of the states for purposes of reserves and the like is the Commissioners' 1941 Standard Ordinary Table, which is adjusted for possible epidemics and catastrophies. This is a development or reflection of the actual mortality that insurance companies experienced in the years preceding the table's adoption. Under it the life expectancy of Mrs. Koshland at age 66 was 11.01 years, as compared with 10.46 years under the table in use by the internal revenue people. Another mortality table currently used in a number of states (in some instances prescribed by law) is the American Experience Mortality Table, which would give Mrs. Koshland an expectancy of 10.54 years. These tables, it is seen, vary but slightly.
The word "Commissioners" in this table has reference to the Insurance Commissioners of the states.
The burden of proving the unsoundness of the table long in use by the bureau was necessarily on the taxpayer. The Tax Court observed that the table advocated by the latter's actuarial expert might be worthy of more consideration if the question at issue were the cost of an annuity from a commercial insurance company; and it remarked upon the absence of any showing that the mortality of inheritors or donees is closely comparable with that of purchasers of annuities. These observations, the taxpayer insists, fly in the face of the actuary's testimony. If that be true we must confess to the same obtuseness from which the Tax Court suffered.
A point of somewhat minor consequence remains to be considered, namely, the contention that the Tax Court failed to take account of the proper factor for quarterly payments of income to which Mrs. Koshland is entitled under the trust. The factor provided by the table prescribed by the regulations is 1.01488, to be multiplied by the annuity value of the annual payments to the beneficiary.
Both parties agree that a life estate where, as here, the tenant is to be paid more frequently than annually has a higher value than one where the payments are annual. The actuary testified that the correct factor to be used in the present instance was .375 added to the factor for annual payments. The Tax Court rejected this view, saying: "Even greater weakness pervades petitioner's argument as to the proper factor for quarterly payments. The actuarial expert testified that the factor respondent used was proper if only an annuity for a term certain were involved, but was not correct if the annuity were for life. He testified further that the value of a life annuity, payable quarterly, is less than the value of an annuity certain, payable quarterly, for a term equal to the annuitant's life expectancy. Yet the factor petitioner urges and the method of its application lead to a higher value for a life annuity. This discrepancy could not be adequately explained by petitioner, nor was there any significant evidence as to the derivation of the factor it sought to have us apply. Petitioner's view cannot be sustained."
Here again the taxpayer expresses amazement at the failure of the Tax Court to comprehend the actuary's testimony. Yet, in our understanding of the record, the testimony of the expert does display the weakness noted by the Court. Possibly the witness himself was confused. His testimony seems opposed to the end result of his computation. In this condition of the record we feel obliged to accept the Court's disposition of the point.