Dr. Melvin A. Yarlott, Jr., appeals from a decision of the United States Tax Court, 78 T.C. 585, that stipends received during the academic phase of a graduate degree program in surgery are taxable income.1 Taxpayer argues that stipends received during the academic phase should be considered separately from stipends received during the clinical phase of the program, and that the former should not be taxable because they were paid for studies and not as compensation for services. We affirm.

Taxpayer was enrolled in the University of Minnesota Graduate School Surgery Program. This is a seven-year program, the completion of which fulfills the requirement for a Ph.D. degree in surgery and a traditional residency program for certification in a medical specialty. The surgery program is divided into three segments. During the first two years the candidate is engaged in the clinical phase of the program. An academic phase usually begins in the third year and continues through the fourth year and sometimes into or through the fifth year of the program. The remaining years are devoted to the completion of the clinical phase. The issue in this case is taxpayer's federal income tax liability for the taxable years 1974 and 1975. During these two years taxpayer was engaged solely in the academic phase of the program. He received a stipend of $11,200.08 in 1974 and $12,150 in 1975. The tax deficiencies determined for those years were $4,189.47 and $4,839.76, respectively.

Taxpayer argues that the two years in which he was involved solely in the academic phase of the program should be considered separately from the program as a whole and that the stipends for those years should qualify for exclusion from gross income pursuant to 26 U.S.C. Sec. 117.

Taxpayer has been before this court on an earlier occasion with respect to stipends received during the residency period, and has taken a contrary position. Rockswold v. United States, 620 F.2d 166 (8th Cir.1980), aff'g 471 F. Supp. 1385 (D.Minn.1979). In that case taxpayer and other parties plaintiff agreed that the advanced degree program must be viewed as a whole and that it would be improper to view part of the stipends as taxable and part as scholarships. 471 F. Supp. at 1392 n. 2.

The question of the taxability of funds received by medical students in residency programs has been before this and other courts on numerous occasions.2 The tax court recognized that the research and clinical phases of the Minnesota advanced degree program are more clearly demarcated than the typical residency program, and counsel for the Commissioner concedes that the research and clinical phases are more readily divisible, on the basis of calendar years, than in other cases. Thus, the question whether taxpayer may exclude funds during the two years is a close one.

Our review of the tax court's decision is guided by the following principles. First, the tax court must consider the evidence as a whole. Rockswold v. United States, 620 F.2d at 168; Leathers v. United States, 471 F.2d 856, 861 (8th Cir.1972), cert. denied, 412 U.S. 932, 93 S. Ct. 2754, 37 L. Ed. 2d 161 (1973). And second, whether payments may be excluded from income under 26 U.S.C. Sec. 117 is a question of fact to be resolved by the finder of fact and should not be disturbed on appeal unless found to be clearly erroneous or not supported by substantial evidence. Rockswold v. United States, 620 F.2d at 169; Leathers v. United States, 471 F.2d at 858. With these principles in mind we review the tax court's determination that stipends received for years involved solely in the academic phase of the program are taxable income because the seven-year program represents an indivisible whole and because the stipends are a quid pro quo for services rendered by taxpayer.

The facts concerning the residency phase of the program are set out in detail in the earlier Rockswold cases. 620 F.2d 167-68; 471 F.Supp. at 1386-87.

With respect to the academic phase of the program, the tax court found that taxpayer was required to undertake original investigative research as part of a thesis requirement and to complete formal course work in a minor. He chose his own research project in tumor immunology and was not confined to a prescribed list of approved projects. He was not paid additional moneys for research conducted during evening and working hours and was not required to work any particular number of hours. He had freedom in the conduct of his research but was subject to supervision on a periodic basis by a research adviser. The tax court further found:

During the years in question, petitioner did not treat any patients or participate in patient rounds. He did not perform any routine patient care related activities such as ordering prescriptions, writing treatment orders for patient care, taking patients' physical and medical histories, etc. He did not diagnose patients, perform any medical or surgical procedures or teach junior medical fellows or medical students. During the academic phase, petitioner was never on call and did not work in any emergency room. Moreover, like other students participating in the academic phase of the Surgery Program, petitioner did not experiment with human beings at any time during the course of his research.

The tax court also found that students in the surgery program do not conduct contract research that might be underway at the university. Research training grants awarded by various private and federal sources fund the direct costs of research which include the payment of stipends to the student. The stipends are kept commensurate with amounts paid to residents in other programs throughout the Midwest, and the floor amount is based on the perceived needs of a typical or average graduate medical student and his family during the period of enrollment. The individual needs of the recipient are not examined.

In determining whether the surgery program is a divisible or unified program, the tax court referred to the Rockswold decisions, supra, and to Johnson v. United States, 507 F. Supp. 663 (D.Minn.1981), which involved the same degree program, but dealt with a taxable year that was spent partly in the academic phase and partly in the clinical phase of the program. The tax court observed that in Johnson the advanced degree program in surgery was held to be not divisible for purposes of establishing an exclusion under 26 U.S.C. Sec. 117.

The tax court continued in this analysis and made the following findings:

The Surgery Program is a unified program leading to an advanced degree. Both the clinical and academic portions of the program must be completed in order to satisfy the degree requirements. The academic phase builds upon the knowledge and experience gained during the initial clinical years and in turn the final clinical years are augmented by the knowledge and experience picked up during the academic phase. The complementary nature of the requirements for obtaining the advanced degree blurs any clear division between the two phases that petitioner contends is present. We believe that each succeeding year of the program builds upon the prior year, that fulfillment of the two components of the Surgery Program indivisibly coalesce to bring the medical fellow to a level of skill and education sufficient to qualify him for the advanced degree.

The tax court observed the varying length of the academic phase of the program and commented further:

Thus, from the very outset, the fifth-year medical fellow is assured precisely the same stipend benefit regardless of whether such year finds him engaged in the clinical phase of the program, the academic phase, or a combination of the two. The fact that the stipend payments are set from the beginning is indicative of a unified approach to the awarding of stipend payments. Such approach carries with it a unified purpose that, on these facts, does not vary according to whether the medical fellow is involved in research or clinical activities.

After concluding that the program must be viewed as an indivisible whole, the tax court went on to examine whether the stipends were a quid pro quo for services rendered, as it was required to do under Bingler v. Johnson, 394 U.S. 741, 751, 758 n. 32, 89 S. Ct. 1439, 1445, 1449 n. 32, 22 L. Ed. 2d 695 (1969). The tax court observed that indirect benefits accrued to the hospitals as a result of a medical fellow's training during the academic phase. It also observed that the activities of the fellow during such time conferred direct benefits on the university, such as the right to share in any royalties produced from patentable research, the requirement of acknowledgment in any publications, and the general enhancement of the university's reputation. Based on these and other factors and the substantial services involved in the program as a whole, the tax court concluded that the primary purpose underlying the stipend payments was to compensate taxpayer for services rendered and therefore that such payments were includable in gross income.

Taxpayer argues that the tax court followed a wooden approach based on Leathers v. United States, 471 F.2d 856 (8th Cir.1972), cert. denied, 412 U.S. 932, 93 S. Ct. 2754, 37 L. Ed. 2d 161 (1973), and erred as a matter of law in concluding that it must consider the program as a whole rather than in segments. We do not so regard the decision of the tax court. Whether the seven-year program should be considered as a unified whole or whether the academic and clinical phases should be treated separately for determining whether the payments received in the specific years were taxable was a question of fact to be determined by the tax court. The tax court made findings of a factual nature in determining that the program was an indivisible whole and must be considered as such. Although the finding of indivisibility may be a finding of an ultimate fact, Pullman-Standard v. Swint, 456 U.S. 273, 287, 102 S. Ct. 1781, 1789, 72 L. Ed. 2d 66 (1982), makes plain that the clearly erroneous rule applies to findings of fact whether they be "ultimate" facts or "subsidiary" facts. Although the issue is a close one, and although the evidence would have sustained a finding either in favor of or contrary to the taxpayer, we cannot conclude that the findings of the tax court are clearly erroneous.

Taxpayer argues that his case is distinguishable from other reported decisions because the undisputed evidence establishes a clear distinction between the stipends paid for the residency or clinical activities and the stipends paid for academic study. He argues that stipends were paid to persons entering the academic phase who had completed residency programs at other institutions, although the tax court found that this consideration was not a significant one. He further argues that he has been paid seven years of stipends for four years of clinical residency work. These arguments essentially raise factual issues which the tax court determined adversely to taxpayer, and which in our judgment are insufficient to support a conclusion that the tax court's findings are clearly erroneous.

Perhaps the most troublesome issue raised by taxpayer is the apparent inconsistency in the Government's position in Revenue Ruling 73-368, which involved a two-year doctorate program. In that case the student spent the first year teaching and the second year engaged in purely academic activities. The Commissioner ruled that stipends paid while teaching were payments for compensation whereas stipends paid for academic work were excludable. Although the circumstances presented in this ruling are extremely similar to, and nearly indistinguishable from, those involved in our case, the tax court has made detailed factual findings on indivisibility that are not clearly erroneous and that are consistent with taxpayer's earlier position.

As we have said, the evidence in this case would fully support a finding in favor of taxpayer. The findings, however, that the advanced degree program is indivisible and that the stipends for the academic phase are taxable income are factual findings that we conclude are not clearly erroneous. Accordingly, we affirm the decision of the tax court.