CCA 200212029: Counsel advises denying examined resident FICA refund claims

Plain-English summary

Chief Counsel advised on FICA refund claims from hospitals and universities after a court decision concerning medical residents. It examined whether residents were students under the student FICA exception in IRC § 3121(b)(10). Counsel considered selected residency programs, resident work hours, classroom and clinical activities, supervision, employment terms, and the value of residents' services to hospitals. It concluded that the examined residents' services were primarily structured on-the-job training, rather than work incidental to a course of study, and advised denying the examined refund claims. The memo says resident status depends on the facts and circumstances; it does not announce that every medical resident is ineligible for the exception.

Ruling snapshot

• Question: Do the examined medical residents qualify for the student FICA exception?
• Outcome: Advice given: deny the examined refund claims.
• Key authorities: IRC § 3121(b)(10); Treas. Reg. § 31.3121(b)(10)-2; Rev. Rul. 78-17; Rev. Proc. 98-16.

Full text (IRS public release)

DEPARTMENT OF THE TREASURY
I N T E R N AL R E V E N U E S E R V I C E WASHINGTON, D.C. 20224

CHIEF COUNSEL
January 24, 2002
Number: 200212029
Release Date: 3/22/2002
UIL: 3121.02-10
CC:TEGE:EOEG:ET2

MEMORANDUM FOR VICTOR PICHON
EO Technical Advisor
Health Care/Colleges and Universities

FROM: Mary Oppenheimer
Assistant Chief Counsel (CC:TEGE:EOEG)

SUBJECT: Medical Resident FICA Refund Claims

This memorandum supplements our earlier memoranda dated April 19, 2000, August 23, 2001, and December 14, 2001. You have asked for our advice on handling FICA refund claims stemming from the Minnesota v. Apfel decision. The April 19, 2000, memorandum described medical residency programs generally and discussed the relevant facts to be developed in examining medical resident FICA refund claims. The August 23, 2000, memorandum provided our views on whether a teaching hospital could be considered a school, college or university (“S/C/U”) for purposes of the student FICA exception. This analysis was supplemented by our memorandum dated December 14, 2001, on whether a teaching hospital is a related § 509(a)(3) organization. The remaining question therefore is whether medical residents are students–the subject of this memorandum.

Fact Development

As a result of Minnesota v. Apfel, 151 F3d 742 (8th Cir. 1998), the Service received many refund claims from hospitals and universities. Our office assisted the EO division in selecting what we believe are a representative sample of cases with the hope that the findings from these cases could be applied to the remaining claims. We selected cases that exhibit each of the common organizational structures identified in Exhibit 2 of our April 19, 2000, memorandum. In addition, we identified several common residency programs, as well as two subspecialty programs, that we believed were representative of the different types of residency programs. We hoped to gather information in order to compare and contrast the various specialty and subspecialty programs.1 Finally, our

1
The specialty programs selected for examination are diagnostic radiology, family practice, general surgery, internal medicine, obstetrics/gynecology (OB/GYN), pediatrics and radiation oncology. The subspecialty programs selected are neurological surgery and thoracic surgery.

2

office assisted the EO division in developing uniform Information Document Requests (IDRs) in order to ensure consistency in factual development.2

The examining agents have substantially finished their audit work and prepared reports summarizing their findings. The agents gathered information from IDRs, resident and attending physician interviews, and from public sources such as the Internet and the American Medical Association’s Graduate Medical Education Directory (the “Greenbook”). Attorneys from the Office of Chief Counsel reviewed the reports and met with the agents to discuss their findings.

Applicable Law

Section 3121(b)(10) excepts from the definition of “employment” “service in the employ of a [S/C/U] or an organization described in section 509(a)(3) . . . if such service is performed by a student who is enrolled and regularly attending classes at such [S/C/U]” (the “student FICA exception”).

Although we concluded in our memorandum dated August 23, 2001, that teaching hospitals are not S/C/Us, it is nevertheless necessary to determine whether medical residents are students because some teaching hospitals may be part of the same legal entity as a university (see the April 19, 2000, memorandum) and it is possible that teaching hospitals could be § 509(a)(3) organizations in relation to a S/C/U (see the December 14, 2001, memorandum).

Section 31.3121(b)(10)-2(c) provides:

The status of the employee as a student performing services shall be determined on the basis of the relationship of such employee with the organization for which the services are performed. An employee who performs services in the employ of a [S/C/U] as an incident to and for the purpose of pursuing a course of study at such [S/C/U] has the status of student in the performance of such services.

2
For sponsoring institutions, the standardized IDRs asked for information regarding: (1) whether its specialty and subspecialty programs were accredited; (2) who supervised patient care activities at participating hospitals; and (3) resident characteristics such as whether residents were enrolled, paid tuition and registered for credit. For each specialty program, the standardized IDRs asked for information regarding: (1) the general characteristics of the individual specialty program, such as the number of residents and whether the residency was accredited by the ACGME; (2) participating institutions, such as the amount of time residents spent at each participating institution; (3) reimbursement arrangements with participating institutions; (4) who supervised the day-to-day activities of the residents, such as whether regular faculty or clinical faculty supervised residents; and (5) documentation required to meet ACGME standards with respect to educational activities.

3

Under § 31.3121(b)(10)-2(b), if the employee has the status of student, “the amount of remuneration for services performed by the employee in the calendar quarter, the type of services performed by the employee, and the place where the services are performed are immaterial.”

Rev. Rul. 78-17.3 Rev. Rul. 78-17, 1978-1 C.B. 306, considered whether services performed by employees in three situations were excepted from employment under the student FICA exception. The Rev. Rul. describes the facts as follows:

Situation 1. A is enrolled in a Master of Education program at the university. During the current academic term A is registered for four courses totaling 12 points of credit. The maximum course load in all programs is 18 points of credit. A is also employed 15 hours per week by the university.

Situation 2. B is enrolled in a Master of Education program at the university. During the current academic term B is registered for two courses totaling 6 points of credit. B is also employed 40 hours per week by the university.

Situation 3. C is enrolled in a Doctor of Education program at the university and has completed the requisite course work. C's dissertation topic has been approved and C is currently conducting the research and experimentation needed for the dissertation. During the current academic term C is registered at the university for dissertation advisement under the supervision of a committee of faculty members. C is also employed 6 hours per week by the university.

For Situation 1, the Service held that because A is (1) enrolled and is regularly attending classes, (2) taking a substantial course load, and (3) employed only on a part- time basis, A’s services are excepted from employment.

For Situation 2, the Service held that although B is enrolled and regularly attending classes, because B is employed on a full-time basis and is taking only two courses worth 6 points of credit (a full-time course load is 15 points), B’s employment is not incident to and for the purpose of pursuing a course of study. Thus, B’s services are not excepted from employment.

For Situation 3, the Service noted that C is enrolled and registered for dissertation advisement. The Service recognized that a certain amount of non-classroom study may be necessary to obtain an academic degree. Thus, C’s pursuance of a regular course of study necessary to receive the desired degree, in accordance with the

3
Rev. Proc. 98-16 did not revoke any earlier Service guidance.

4

requirements of the school, satisfies the requirement of regularly attending classes. Further, the Service held that C’s employment on a part-time basis is incident to and for the purpose of pursuing a course of study.

Thus, the Service held the following factors to be relevant in determining student status: (1) whether the employee is enrolled and regularly attending classes; (2) the extent of the employee’s course load; and (3) whether the employee is employed on a part-time or full-time basis. Further, the “regularly attending classes” requirement may be met if the employee is conducting research and experimentation required by a S/C/U to earn an academic degree.

Rev. Proc. 98-16. Revenue Procedure 98-16, 1998-1 C.B. 403, sets forth generally applicable objective standards for determining whether services performed by an employee of certain institutions of higher education are excepted from FICA tax under
§ 3121(b)(10). However, the objective standards do not apply to, inter alia, medical residents “because the services performed by [medical residents] cannot be assumed to be incidental to and for the purpose of pursuing a course of study.” This does not mean that medical residents could not possibly be students; rather, whether medical residents are students depends upon the facts and circumstances in each case.

The Rev. Proc. substantially modified the holdings in Rev. Rul. 78-17. For example, the Rev. Proc. modified the holding in Situation 1 by providing an objective course load standard and eliminating the requirement that the employment be on a part-time basis, and modified the result in Situation 2 both by applying the course load standard and by applying an analysis of whether the employee was a “career employee” as opposed to whether he or she was a full-time employee. While Rev. Proc. 98-16 does not provide controlling standards for medical residents, we note that its approach is similar to Rev. Rul. 78-17. Both consider (1) whether the employee is enrolled in classes for credit; (2) the extent of course load; and (3) the nature of the employment relationship.

The Student FICA Exception as Applied to Medical Residents

Social Security Ruling 78-3. In Social Security Ruling 78-3, the Social Security Administration (SSA) considered, inter alia, whether medical residents performing services at Maricopa County General Hospital in Arizona during the years 1970 through 1974 were students and thus excluded from the State’s § 218 agreement. Apparently the State had elected to exclude students from coverage under the § 218 agreement. The SSA noted that the 1965 Amendments to the Social Security Act (the “Act”) provided that effective January 1, 1966, the services of medical and dental interns would no longer be excluded from coverage. The SSA concluded that medical interns were not excluded from social security coverage under the State’s § 218 agreement. The SSA found support for this position in St. Lukes Hospital Assoc. v. United States, 333 F.2d 157 (6th Cir. 1994) (see Appendix).

The University of Minnesota Decisions. In Minnesota v. Chater, 1997 U.S. Dist. LEXIS 7506, (D. Minn. 1997), the State of Minnesota sought a redetermination of an

5

SSA determination that medical residents employed in the University of Minnesota (“University”) medical residency programs during 1985 and 1986 were covered under social security pursuant to the State’s § 218 agreement.4 The State argued that medical residents were not included in a coverage group under the § 218 agreement, or, alternatively, that the residents were excluded under the general student exclusion that the State elected to place in the agreement.5 The court concluded that University residents were not included in a coverage group, but, in any event, it determined them to be students within the meaning of § 210(a)(10) of the Act, and thus excepted from coverage under the § 218 agreement.6

In concluding that the medical residents were students, the court noted the following facts:
•

• University medical residents were enrolled at the University, paid tuition and registered for approximately 15 credit hours per semester.
• Although the residents did provide patient care, this was a necessary part of their medical education. “A future physician cannot adequately develop skills if not permitted to perform procedures on real patients.” The court found significant that residents are subject to varying levels of supervision depending upon their

4
State of Minnesota involved the tax years 1985 and 1986. Under the Omnibus Budget Reconciliation Act of 1986, Pub. L. No. 99-509, the IRS became responsible for determining liability for social security taxes under a § 218 agreement with respect to remuneration for services paid after December 31, 1986. See the April 19, 2000, memorandum for a discussion of coverage under § 218 agreements.
5
Section 218(c)(5) of the Act (42 U.S.C. § 418(c)(5)) provides:

Such agreement shall, if the state requests it, exclude (in the case of any coverage group) any agricultural labor, or service performed by a student, designated by the State. This paragraph shall apply only with respect to service which is excluded from employment by any provision of section 210(a) other than paragraph (7) of such section and service the remuneration for which is excluded from wages by subparagraph (B) of section 209(a)(7).
6
The statutory language is the same under § 210(a)(10) of the Act and § 3121(b)(10) of the Code. Section 210(a)(10) of the Act (42 U.S.C. § 410(a)(10)) excepts from employment:

Service performed in the employ of (A) a school, college, or university, or (B) an organization described in section 509(a)(3) of the Internal Revenue Code of 1986 if the organization is organized, and at all times thereafter is operated, exclusively for the benefit of, to perform the functions of, or to carry out the purposes of a school, college, or university and is operated, supervised, or controlled by or in connection with such school, college, or university, unless it is a school, college, or university of a State or a political subdivision thereof and the services in its employ performed by a student referred to in section 218(c)(5) are covered under the agreement between the Commissioner of Social Security and such State entered into pursuant to section 218;
if such service is performed by a student who is enrolled and regularly attending classes at such school, college, or university.

6
experience and skill level. The court noted that “like students in other disciplines, residents are evaluated on their performance.” • In addition to the clinical component, other educational experiences included daily rounds, lectures and formal didactic courses.

• Residents in their first year are not eligible to be licensed in the State and residents beyond their first year are not required to be licensed because the State considered them to be students.
• The University classified residents as holding “student/professional training” positions.
• Residents have been characterized as “students” in other contexts in Minnesota.
  For example, for workers’ compensation purposes, the Minnesota statute defined the term “employee” to include “students enrolled and regularly attending the medical school of the [University] in the graduate school program [(M.D. program)] or postgraduate program.”
  • Finally, failure to make satisfactory progress could result in the dismissal of the resident from the program.

On appeal, the Eighth Circuit upheld the district court’s determination that the medical residents were students, citing only that the residents were (1) enrolled at the university, (2) paid tuition, and (3) were registered for approximately 15 credit hours per semester. 151 F.3d at 748. The court also noted its finding in Rockswold v. United States, 620 F.2d 166, 167 (8th Cir. 1980), that the medical residency program “is designed to educate and train physicians so that they can pursue careers in academic medicine and medical research.” 151 F.3d at 747-48. The court refused to grant deference to Social Security Ruling 78-3, finding it contrary to the social security regulations, which require a case-by-case examination of the facts. Id. at 748.

In response to the Minnesota decision, the SSA issued Acquiescence Ruling 98-5 (8), 63 F.R. 58444. Ruling 98-5 applies only to employers located in the 8th Circuit (Minnesota, the Dakotas, Nebraska, Iowa, Missouri and Arkansas). The ruling provides that, in applying the student services exclusion within the 8th Circuit, SSA will make a case by case examination of the relationship of medical residents with the employer S/C/U to determine whether the residents meet the statutory criteria of being enrolled and regularly attending classes. In evaluating the relationship, the SSA will consider all the facts and circumstances.7

Analysis of the Facts and Circumstances

A. The Legal Standards

Following the Eighth Circuit decision, there are two key legal standards against which we must analyze the facts to determine student status for purposes of § 3121(b)(10):

7
Unlike the SSA, the Service applies a facts and circumstances approach in all circuits.

7

• The employee must be enrolled and regularly attending classes;
• The relationship between the employee and S/C/U must be examined to determine whether the services were incident to and for the purpose of pursuing a course of study.

Under regulations § 31.3121(b)(10)-2(c), the status of an employee as a student performing the services is determined on the basis of the relationship of the employee with the organization for which the services are performed. In determining the nature of the relationship, we believe it is appropriate to examine the relationship from both the perspective of the employer and the employee. In examining a relationship to determine student status, it is not practical to determine a person’s subjective reasons for engaging in certain activities. Instead, the more practical and reliable approach is to examine the objective facts. We believe the true nature of a relationship is manifested by the activities of the parties.8 Thus, our analysis concentrates on what the residents do; how the institutions structure the activities of the residents; and what activities predominate in terms of time spent and in terms of relative priority. We have also found it useful to compare residents to other types of students and employees in order to determine whether their relationship with the sponsoring and participating institutions most resembles that of a student/employee or a non-student/employee.

This memorandum discusses certain aspects of the relationship between medical residents and teaching hospitals that we believe have the greatest force in determining the nature of the relationship. The facts discussed were developed by Service agents in examining the FICA refund claims. In addition, where appropriate, the memorandum discusses the legislative history of the student FICA exception.

B. The Minnesota Factors

The appellate court cited the following facts in concluding that residents were students: (1) the residents were enrolled at the university, (2) paid tuition and (3) and were registered for approximately 15 credit hours per semester. It was our sense that these facts were uncommon among residency programs before the examinations began, and the agents’ findings confirmed this general impression. Indeed, in none of the examined cases did the residents pay tuition or register for course credit at a university. The residents did not receive a university degree upon completion of the program, but instead received a certificate of completion.

8
The Social Security Regulations at 20 C.F.R. § 404.1028(c) state: “Whether you are a student for purposes of this section depends on your relationship with your employer. If your main purpose is pursuing a course of study rather than earning a livelihood, we consider you to be a student and your work is not considered employment.” We note the Employment Tax Regulations, which look to whether the services were “incident to and for the purpose of pursuing a course of study,” is more clearly an objective standard. The Employment Tax Regulations provide the controlling standards in these cases, and, moreover, the court in Minnesota looked to the objective facts, although apparently it was presented with limited facts upon which to judge the relationship between the University and the medical residents.

8

The Tax Court noted the unique nature of the University of Minnesota residency program in a 1982 decision considering whether stipends paid to University of Minnesota residents were excludable under § 117 of the Code. “Unlike the typical medical residency program, the residency program in the instant case combined an academic phase with the traditional clinical phase, the two of which in combination satisfied the requirements for an advanced degree.” Yarlott v. Commissioner, 78 T.C. 585, 597 (1982), aff’d, 717 F.2d 439 (8th Cir. 1983). See also Rockswold, 620 F.2d at 167 (noting that the University of Minnesota program combined a clinical phase with an academic phase leading to an advanced degree).

Another unusual fact cited by the district court in Minnesota is that first year medical residents could not be licensed, and residents beyond their first year of training need not obtain a license to practice medicine because they were treated as students under State law. By contrast, the agents found that medical residents generally obtained provisional licenses to practice medicine within an accredited residency program during their internship year (first year) and thereafter residents were generally eligible to become fully-licensed, and, indeed, were responsible under their employment contracts to become licensed to practice medicine in the state.9

Yet another fact that appears to be unique to Minnesota is that residents were characterized as “students” under the state’s workers’ compensation laws. The Minnesota statute specifically treated as employees for workers’ compensation purposes “students enrolled and regularly attending the medical school of the
[University] in the graduate school program or postgraduate program.” In none of the cases examined by our agents did the state find it necessary to specifically include medical residents within the definition of employees covered for purposes of workers’ compensation. Nor are we aware of other states that specifically bring medical residents within the workers’ compensation statute or that refer to medical residents as “students.” Most states cover all employees under their workers’ compensation laws unless specifically excepted. In reported cases involving injured interns and residents, coverage for purposes of workers’ compensation is assumed to exist without question or discussion. The issues in dispute deal with other matters. John Doe v. Yale University, 252 Conn. 641 (1999) (whether a joint venture is the “employer” of the resident under the Connecticut workers’ compensation act); Gedon v. University Medical Residents Services, et al., 677 NYS 2d 397 (1998) (whether the injury sustained by the resident arose out of and in the course of his employment); Mermelstein v The City of New York and New York City Health and Hospitals Corporation , 571 NYS 2d 261 (1991) (who was the employer of residents responsible for paying workers’ compensation benefits). Perhaps the unique nature of Minnesota

9
After completing a period of graduate medical education (GME) (typically one year, as determined by the state) and passing part three of the U.S. medical licensing exam, a resident is eligible to become fully-licensed to practice medicine. A resident must complete two years of GME to be licensed in Connecticut, Michigan, New Hampshire, New Mexico, Pennsylvania, South Dakota, Utah and Washington. Nevada requires three years of GME to be licensed. Longer periods apply to graduates of medical schools outside of the United States and Canada. See 2000-2001 Greenbook, page 1260.

9

statutes derives from the unique nature of the University’s graduate medical education program.

C. The Primacy of Patient Care

When we seek to apply the statute and regulations, the key issue is whether the services (patient care) performed by the residents are incident to and for the purpose of pursuing a course of study. The facts developed by the agents consistently demonstrate that although residency programs have a significant educational component, patient services are not incident to a course of study. Instead, patient care is the paramount activity in the relationship between the resident and the employer.

Hours Worked. In examining the nature of the relationship between residents and teaching hospitals, the agents identified a number of facts indicating the primacy of patient care. The most significant fact is the number of hours that residents work. The agents confirmed our understanding that residents in most cases work in excess of 80 hours per week (including on-call time). Some residents reported routinely working in excess of 100 hours per week. For example, residents in surgical and OB/GYN residencies reported working in excess of 100 hours per week. In addition, certain rotations, such as rotations in an intensive care unit, require in excess of 100 hours per week. Diagnostic radiology residents reported working fewer hours (approximately 50 to 60 hours per week). There is also some variation based upon the year of the residency. Interns worked the most hours with the hours decreasing as the resident progressed through the residency. In addition, there appeared to be some variation based upon the size of the hospital and the area in which the hospital is located. Large hospitals in metropolitan areas seemed to require longer hours by residents in order to meet their patient care needs.10

The agents’ findings indicate somewhat longer hours than reported by the American Medical Association (AMA) in the following chart.11

10
Recent newspaper articles have reported on the long hours that residents work and the effect that fatigue has on their ability to function properly. See N.R. Kleinfield, Life, Death, and Managed Care, New York Times (November 14-17, 1999); S. Jauhar, Medical Residents, Yes, But Workers, Too, New York Times (April 18, 2000); Abigail Trafford, Sweatshop Conditions Can’t Give Quality Care, Washington Post, page HE5 (March 27, 2001); Low Experience, High Expectations, Washington Post, page HE12 (March 27, 2001); A Day (and a Half) in the Life of an Intern, Washington Post, page HE16 (March 27, 2001). In addition, an OSHA complaint was recently filed requesting that residents’ hours be limited to 80 hours per week. The complaint was filed by the consumer group, Public Citizen; the Committee for Interns and Residents, a union representing 11,000 residents; and the American Medical Student Association, which represents 30,000 medical students. This complaint is summarized in, Ted Rohrlich, Curbs Urged on Interns’ Workweek, Los Angeles Times, page A16 (May 1, 2001).
11
The statistical information in the above chart was obtained from a document published by the American Medical Association entitled Characteristics of Graduate Medical Education Programs and Resident Physicians By Specialty, Table 12 (1999).

10

The Average Hours on Duty Per Week During the First Year of Selected ACGME-Accredited Residency Program (1998-1999)

Specialty/Subspecialty
Hours on Duty Per Week

Family Practice
64.1

Internal Medicine
66.1

Ob/Gyn
74.8

Pediatrics
71.4

Radiology
51.1

General Surgery
79.9

Thoracic Surgery
73.1

One Radiation Oncology resident said there were no set hours; rather, residents were allowed to go home only after all the patients had been seen. In response to an agent’s request for information on the work schedules of surgical residents one taxpayer stated: “When the OR schedule runs late, as often happens, the resident and the involved teaching physician mutually decide when the resident may leave or be relieved. The guiding principles are patient care and resident education (in that order).”

The Service has held that the hours worked by an employee are relevant in determining whether the employee’s services are incident to and for the purpose of pursuing a course of study. See Rev. Rul. 78-17. Rev. Proc. 98-16 did not revoke Rev. Rul. 78-17 and its holdings are not inconsistent with this view. The Rev. Proc. provides that if its objective standards do not apply because an employee’s services cannot be presumed to be incident to and for the purpose of pursuing a course of study, then student status is determined based upon the facts and circumstances. In a facts and circumstances analysis, we believe that hours worked continue to be a relevant factor.

The Social Security Act Amendments of 1939. The legislative history and other authorities support an analysis that looks to hours worked as a relevant factor in determining student status. The student FICA exception was enacted by the Social Security Act Amendments of 1939 (“SSA of 1939"), Pub. L. No. 76-379, § 606. Rev. Rul. 78-17 cited the following legislative history in support of its holding:

In order to eliminate the nuisance of inconsequential tax payments the bill excludes certain services performed for fraternal benefit societies and other nonprofit institutions exempt from income tax and certain other groups. While the earnings of a substantial number of persons are excluded by this recommendation, the total amount of earnings involved is undoubtedly very small. . . . The intent of this exclusion is to exclude those persons and those organizations in which the employment is part-time or intermittent and the total amount of earnings is only nominal, and the payment of tax is inconsequential or a nuisance. The

11
benefit rights that are built up are also inconsequential. Many of those affected, such as students and the secretaries of lodges, will have other employment which will enable them to build up insurance benefits. This amendment, therefore, should simplify the administration for the worker, the employer, and the government.

H.R. Rep. No. 728, 76th Cong. 1st Sess. (1939), 1939-2 C.B. 538, 543 (emphasis added). The Senate Report uses similar language. S. Rep. No. 734, 76th Cong. 1st Sess. (1939), 1939-2 C.B. 565, 570.

This legislative history was also the basis for the standards set forth in Revenue Ruling 85-74, 1985-1 C.B. 331, dealing with the student nurse exception under § 3121(b)(13) of the Code. This section excludes from the definition of employment “service performed as a student nurse in the employ of a hospital or a nurses’ training school by an individual who is enrolled and is regularly attending classes at a nurses’ training school chartered or approved pursuant to State law.” Based upon the legislative history to the SSA of 1939 set forth above, and the statutory language providing the exception from employment for “service performed as a student nurse,” the Revenue Ruling promulgated the following three-part test for determining whether a nurse’s services are excepted from FICA under the student nurse exception:

(1) The employment is substantially less than full-time,

(2) The total amount of earnings in nominal, and

(3) The only services performed by the student nurse for the employer are incidental parts of the student nurse’s training toward a degree which will qualify him or her to practice as a nurse or in a specialized area of nursing.

Rev. Rul. 85-74 was challenged in Johnson City Medical Center v. United States, 999 F.2d 973 (6th Cir. 1993). The court, applying a Chevron12 analysis, first held that the statute was not unambiguous. Id. at 976. Continuing its Chevron analysis, the court found that the standard to be applied with respect to a revenue ruling is whether “it conflicts with the statute it supposedly interprets or with the statute’s legislative history or if it is otherwise unreasonable.” The court concluded that the Rev. Rul. reflected the legislative history, and was not unreasonable in any way. Thus, the court accorded deference to the Rev. Rul. based upon the standards set forth in Chevron. Id. at 977. In addition, the court upheld the district court’s determination that the remuneration received by the nurses was more than nominal. Id. at 977-78.

Thus, we believe that consideration of hours worked in determining student status reflects Congress’ intent in enacting the student FICA exception. We believe that the

12
Chevron U.S.A, Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 842-43 (1984).
Under Chevron, if an agency has been explicitly authorized by Congress to promulgate regulations, such regulations are given controlling weight unless they are arbitrary and capricious. If the authority to provide rules is not express, but rather implied, then a court may not substitute its own statutory construction if the agency’s interpretation is reasonable.

12

long hours worked by residents strongly suggests that their services are not incident to and for the purpose of pursuing a course of study.

Didactic Activities Secondary. The agents generally found that didactic activities were secondary to patient care. Some conference time was considered “protected,” meaning that others were responsible for covering patient care during that time. However, if a patient needed a resident’s attention, didactic activities generally gave way to patient care needs. Residents reported that it was not uncommon to be “beeped out” in order to handle a patient emergency. This is consistent with the Greenbook, which states:

The education of resident physicians relies on an integration of didactic activity in a structured curriculum with diagnosis and management of patients under appropriate levels of supervision and scholarly activity aimed at developing and maintaining life-long learning skills. The quality of this experience is always related to the quality of patient care, which is always the highest priority (emphasis added).13

House Staff Manuals. The house staff manuals and handbooks also demonstrated that the overarching purpose of the relationship between a resident and a teaching hospital is to provide patient care. The house staff manual in one case stated that the primary responsibility of the medical resident is to participate in the care of the patient under appropriate supervision.

Supervision of Patient Care. Much of a resident’s day-to-day work is conducted without direct participation by a faculty member/attending physician. The residents reported that an attending physician was typically actually present about two hours per day. Instead, much of the supervision is done by phone or pager and records are later signed by the attending physician.14 For example, medical residents may do the paper- work in connection with an admission or a discharge, with the attending physician signing-off on the admission or discharge after the fact. Third and fourth year residents develop treatment plans which are presented to the attending physician. In addition, third year residents are in charge of code cases with the attending physician only providing support. The attending physician was always advised as to medication changes and patient condition or care changes, but attending physicians often post- review these actions. Most routine tests, prescriptions, lab work, blood work up and CT scans are requested without the prior approval of an attending physician. However, radiation oncology residents may not prescribe radiation treatment without the prior approval of an attending physician, and the attending physician must be present during the “key portion” of a surgical procedure. One resident stated that after the first six months, the medical residents made the decisions, with the attending physician simply approving the plan of care. Indeed, the residents indicated that they were supposed to

13
2000-2001 Greenbook, page 31.
14
However, the agents found that emergency rooms and intensive care units generally have attending physicians present at all times.

13

demonstrate independence, and thus it was not viewed favorably if a resident asked for too much assistance.

Instead, the residents relied heavily on the senior residents and the chief resident for guidance and supervision. Residents are expected after their first year to demonstrate leadership skills. Chief residents are responsible for preparing work schedules and for insuring adequate patient coverage at all times. The chief resident plans the schedules and draws up house staff teams consisting of medical students, interns, residents and a senior resident. The chief resident may supervise three house staff teams. In hospital wards, the senior residents are typically in charge of patient care overnight with no supervision from attending physicians. A PGY-5 surgical resident is expected to manage his own surgical team, including assigning surgical cases to the more junior surgical residents.

The residents themselves provide much of the training on specific procedures. The popular saying in residency programs is “see one, do one, teach one.” After a resident becomes “certified” or obtains “privileges” for a particular procedure, the resident may perform that procedure and supervise another resident who is not certified in performing that procedure. All procedures must be performed a certain number of times before a resident can be certified in that procedure. These procedures range from the relatively simple such as a blood draw to more complex procedures such as a colonoscopy.

Indeed, in some cases we learned that the residents are designated as faculty in supervising the services performed by medical students and are given faculty appointments. In addition, fellows15 in some cases were regular or voluntary faculty members and at the same time residents. In one house officer’s agreement, the house officer was appointed as an adjunct faculty member, and was granted the privileges associated with such status.

Attending Physicians Often Not Hired as Teachers. The primary purpose of the relationship between teaching hospitals and attending physicians/faculty members in many cases also seemed to be patient care. Although there is an expectation that staff physicians at a teaching hospital will train residents, medical staff are generally selected based upon their medical knowledge and training, not whether they are effective teachers, or even whether they have any teaching experience. Although there was significant variation in the examined cases, the agents found that most of the “faculty” at the teaching hospitals were “clinical faculty.” Clinical faculty are physicians who are not members of the regular teaching staff of a medical school, but instead are staff physicians, or physicians with hospital privileges who have volunteered to train residents. Clinical faculty are generally not compensated for training residents. The

15
The ACGME defines a “fellow” as “[a]n individual undertaking post-residency training in a field of research that is not accredited by the ACGME. Some specialities also use ‘fellow’ to designate resident physicians in subspecialty GME programs.” The agents found that residents in accredited subspecialties were referred to as “fellows.” The ACGME prefers that individuals in subspecialties be referred to as “residents.” See 2000-2001 Greenbook, page 1270.

14

agents found that appointment to the faculty of an affiliated medical school was automatic upon becoming a member of the physician staff, and likewise faculty status automatically ended upon ceasing to be a staff member of the hospital. An agent noted that in one reaccreditation application it was stated that other aspects of the faculty member’s responsibilities such as research, paper writing, grant proposals and meetings must not interfere with insuring the proper functioning of the ward service.

Residents Meet a Hospital’s Operational Needs. As would be expected, during these long hours residents provide substantial patient care services. The agents typically found that medical residents fulfilled a substantial portion of the operational needs of a hospital. In one case, out of beds, approximately beds were under the care of house staff teams. In another case, the average census per internal medicine intern was eight to ten patients per day with an average stay of seven days. In another case, a second year radiation oncology resident reported performing approximately ten consults, 10 -15 follow ups and caring for 20 treatment patients per week. A fourth year radiation oncology resident saw approximately 40 patients per week plus 20 patients undergoing treatment. Surgical residents reported seeing between 10 and 40 patients on a daily basis.16 Thus, from the perspective of the institution responsible for patient care, it is difficult to describe the patient care provided by the resident as a mere incident to an educational program.

D. Educational Activities

Curriculum. The requirements that residents be “enrolled and regularly attending classes” and “pursuing a course of study” suggest that the learning must be more structured than simply the experiences arising from the treatment of whatever patients happen to be admitted to the hospital. We believe this language requires that a curriculum exist. In practice, the curriculum presented to our agents generally consisted of the rotations that residents perform. The ACGME establishes the rotations that must be performed in each specialty.17 For example, a few of the “focused experiences” for family practice residents are: (1) human behavior and mental health; (2) adult medicine; (3) maternity and gynecological care; (4) care of the surgical patient; (5) sports medicine; and (6) care of neonates, infants, children and adolescents. For internal medicine residents, a few of the “major learning experiences” are: (1) ambulatory patients, (2) continuity ambulatory patients, (3) hospitalized patients, (4) emergency medicine patients, and (5) critical care patients.18

16
In Boston Medical Center v. National Labor Relations Board, 330 N.L.R.B. No. 30, 1999 NLRB Lexis 821, *31 (1999), the American Public Health Association filed an amicus brief for the petitioner. It noted that “house staff provide the bulk of physician-type services to the traditionally underserved in hospital emergency rooms and clinics, and that it would perhaps be insulting, if not disquieting, to the under served to be told that their care is being provided not by ‘doctors’ but by ‘students.’ ” 17
The larger institutions had many nonaccredited fellowship programs. These nonacredited programs were in subspecialities that have not been recognized by the ACGME. Examples of nonaccredited residency programs include breast imaging, a subspecialty of diagnostic radiology; gynecological oncology, a subspecialty of OB/GYN; and neurophysiology, a subspecialty of neurology.
18
Greenbook, pages 82-85 (family practice), pages 96-99 (internal medicine).

15

Although residents generally progressed from simpler to more demanding responsibilities, our agents were not presented with evidence of a curriculum of the type expected in an academic program. Instead, what residents learned depended upon what the patients presented. We believe that this type of training is a form of structured on-the-job training that might be found in other professions or trades requiring highly- skilled and highly-specialized workers. We note that as in any on-the-job training, the specific learning experiences depend upon the job at hand, which, in the case of medical residents, depends upon the patients. We do not believe that structured on- the-job training equates to a “course of study” for purposes of the student FICA exception.

“Classes.” Assuming for the sake of discussion that a curriculum exists, it must be determined whether the medical residents were “regularly attending classes.” We do not believe “classes” should be interpreted so narrowly to include merely traditional lecture/discussion and lab sessions. Instead, a variety of events and activities, whether or not in a classroom, including lectures, demonstrations, tutorials, and teaching rounds, at which a faculty member plays a leadership role in furthering the objectives of an established curriculum, may be considered classes for purposes of the student FICA exception. The frequency of events such as these determines whether the medical resident may be considered to be regularly attending classes.

The agents identified certain activities that were didactic in nature and that could reasonably be considered “classes” for purposes of the student FICA exception. For example, teaching rounds, certain lectures and conferences and supervised research projects in many cases could reasonably be considered classes. The agents found that these activities were primarily intended to educate medical residents on a specific topic. These activities typically arose in the context of a case that presented an opportunity for learning about a particular subject matter.

However, taxpayers in many of the cases asserted that any supervised activity should be considered a class. We believe that it is not enough that an activity was supervised, because attending physicians are responsible for supervising all patient care. As we have discussed in our memorandum, attending physicians have dual roles, namely, (1) he or she is the attending physician of record having ultimate responsibility for patient care, and (2) generally he or she is considered a “faculty” member of an affiliated medical school or the teaching hospital itself with responsibility for overseeing the residents’ training. We believe that the existence of a curriculum and activities furthering that curriculum under the supervision of a faculty member distinguish a class that is part of a course of study from on-the-job training or work experiences.

In addition, it is necessary to distinguish between industry quality control standards and classes that are part of a course of study for residents. The agents noted that in many cases the activities which the taxpayer claimed were “classes” within the meaning of
§ 3121(b)(10) were activities that would have occurred based upon the quality control systems in place at the hospital. For example, mortality/morbidity conferences and tumor conferences would occur regardless of whether the hospital participated in a GME program. The agents found that certain activities, like the review of a radiology

16

film by the attending physician, could be viewed as a quality control measure, because in many institutions two physicians review films. One diagnostic radiology resident noted that when moonlighting another physician reviews his slides in accordance with standard quality control practices.

The agents also learned that activities such as grand rounds and certain other conferences qualify for Continuing Medical Education (CME) credit for state licensing purposes. We question the appropriateness of considering these activities to be “classes” within the meaning of § 3121(b)(10) since the activities appear to merely be the continuing education in which all professionals generally participate.

It is also necessary to distinguish between activities that are primarily patient care and activities that are part of a medical resident’s course of study. For example, the agents found that “work rounds” or “morning rounds” should not be considered “classes.” The purpose of work rounds is to check on the patient’s condition overnight. In most cases, the work rounds were not supervised by attending physicians, but rather the more senior residents conducted the rounds in which junior residents and medical students participated. During work rounds, the patients’ statuses are reviewed and discussed and a plan of treatment is formulated. These rounds typically last about two hours. A pediatrics residency intern described his duties in connection with work rounds as taking vitals, temperature, pulse, fluid information, performing a physical exam and making assessments. This information is provided to the senior resident and then to the attending physician. These rounds are primarily intended to ensure that patients receive good care. While work rounds provide valuable work experience, they should not be considered classes for purposes of the student FICA exception.19

Finally, a large portion of a typical residents’ day is spent performing routine activities that could not reasonably be argued were “class-type” activities. For example, writing orders, blood draws and IVs and other so-called “scut work” are common tasks, particularly among junior residents. In addition, residents answer patient questions and document patient histories. The residents stated that they perform these relatively routine tasks to assist the attending physician so he or she does not get bogged down. The residents reported that this type of work tapered off by the fourth year. In addition, the amount of scut work varied among specialties and among institutions. These tasks are generally learned in medical school. Repeating these types of activities is not educational and certainly cannot be considered classes.

19
The ACGME distinguishes among “teaching rounds,” “work rounds” and “management rounds.” Teaching rounds are “patient-based sessions in which a few cases are presented as a basis for discussion of such points as interpretation of clinical data, pathophysiology, differential diagnosis, specific management of the patient, the appropriate use of technology, and disease prevention.” The purpose of management rounds is for the attending physicians to “interact at intervals with his or her patients and to communicate effectively and frequently with the resident staff participating in the care of these patients.” “Work rounds are rounds in which a senior resident supervises a junior resident’s patient care activities, without an attending physician present.” 2000-2001 Greenbook, page 97. Thus, “teaching rounds” are intended to be for educational purposes, whereas “management rounds” and “work rounds” are primarily for the purpose of ensuring adequate patient care.

17

Teaching Rounds. The agents found that some activities could more reasonably be considered to be “classes.” For example, there often times are “attending rounds” or “teaching rounds” that follow the house staff team’s rounds. The purpose of these rounds is to discuss the plan of treatment for each patient. A resident may be responsible for “presenting” patient cases to the attending physician and the team several times during a year. When presenting a case, the resident is expected to be prepared to field questions from the attending physician. We believe it is reasonable to consider teaching rounds to be classes for purposes of the student FICA exception.

The agents noted some variations for radiation oncology residencies. Teaching rounds and morning conferences occurred twice per week for about one hour. At weekly “chart rounds,” attending physicians, residents, fellows, therapists, and physicists all attend to discuss treatment. In addition, radiotherapy planning conference is held daily during which all new patients are presented. During these conferences, the presenting resident is questioned by faculty and other medical residents in attendance. Residents report spending many hours preparing for a case presentation. These too are reasonably considered classes.

Noon Conferences. Common among the programs were lectures scheduled every day around noon while the residents were eating lunch. The subject matter of noon conferences typically centered on current patients. The residents generally stated that attendance was considered mandatory, but in most cases attendance was not taken. Attendance was typically 60 - 70 percent of the residents. There were also so-called “pizza conferences” in another case that appeared to serve the same purpose as noon conferences. It is not unreasonable to also consider noon conferences to be classes for purposes of the student FICA exception.

Grand Rounds and Other Conferences. Grand rounds generally occur at all teaching hospitals. Grand rounds might have a large audience including attending physicians and other staff members. In one case, diagnostic radiology grand rounds were held once per month and included an invited speaker. In addition, from time to time attending physicians give lectures or talks on complex or interesting topics or cases. In one family practice residency, residents were expected to attend a lecture once per week for four hours. In addition, while rotating through a department, a family practice resident was expected to attend the department’s conferences.

The agents found that attendance at grand rounds and other conferences is sporadic. If a resident misses a lecture or conference due to patient care requirements there is no make-up available. Residents on rotations outside the hospital will not attend conferences. In addition, the agents found that attending physicians and other medical staff members receive CME credit for grand rounds and some of the other conferences that medical residents attend. Thus, we believe that grand rounds and other conferences in many cases are not reasonably considered classes. However, conferences or lectures held specifically for the purpose of educating residents are reasonably considered classes.

18

Diagnostic radiology and radiation oncology residencies seemed to have more didactic activities. In one diagnostic radiology residency, conferences consisting of lecture and film review were held twice per day at 12:30 and 4:00. In addition, fourth year radiology residents were required to take a physics class. Radiation oncology residents were required to take physics courses at an affiliated institution. These activities are reasonably considered classes. However, the relevance of course work outside the hospital is questionable because the employee must be enrolled and regularly attending classes at the institution where the employee is employed.

Morbidity/mortality Conferences and Tumor Conferences. Generally all staff members participate in morbidity/mortality conferences. These conferences appear to be quality control measures. In addition, the agents learned that states generally allow CME credit for these conferences. Thus, we believe that these conferences are not reasonably considered classes.

Journal Club. Journal club was common in all the cases. In journal club, residents discuss recent medical journal articles. Journal club typically met once per week for one or two hours, and may have been held in a bar or restaurant. One resident said that there was usually a designated topic but invariably the discussion turned to another topic. It was our impression from the agents’ reports that faculty members generally did not participate in journal club. Whether a journal club meeting is reasonably considered a class depends upon whether a faculty member participates in the meeting.

Research Projects. Research projects are encouraged, but generally are not mandatory. Rather, as in any professional field, research and writing is viewed as a career enhancer. For some residencies, a research project was required to be completed during the residency. In these cases, residents were given time to complete the project, but some patient care was required because Medicare does not reimburse for a resident who is only conducting research. (See the discussion of Medicare below). In some cases, on-call time was used to satisfy the patient care requirement for Medicare purposes. For radiation oncology residencies, a research project of publishable quality was expected to be completed during the final year of the residency. Weekly meetings were held with a faculty member to discuss the project. Again, in a case where a faculty member is involved in supervising a research project, this activity is reasonably considered a class.

Testing. It would be expected that a student would be tested to determine whether he or she has satisfactorily mastered the subject matter of the class. Residents’ knowledge of the material covered during teaching rounds, grand rounds and other conferences is not tested. The evaluation of residents is not focused primarily on cognitive knowledge; rather, they are primarily evaluated by their performance on-the- job. (See discussion below of performance appraisals.) Although residents may take a yearly standardized in-service examination, the agents learned that in-service testing is used in large part to measure a program’s strengths and weaknesses and to prepare residents for board examinations. For certain specialities, such as radiation oncology and general surgery, residents maintain a log of the procedures they have performed.

19

But again, this record is required so that the resident can sit for board exams, not to measure whether a resident has mastered the subject matter of a class.

Conclusion. The agents found that didactic activities, including teaching rounds, and lectures and conferences, were generally between four and twelve hours per week. The agents found generally that the time spent in didactic activities did not change much from year to year. Even if the most liberal definition of “classes” is used, including the time spent in work rounds, residents spent only 10 to 20 percent of their time in didactic activities. Thus, even assuming for the sake of discussion that a “course of study” or a curriculum existed, the time spent in patient care activities compared to the time spent in didactic activities indicates that the services performed by residents were not incident to and for the purpose of pursuing a course of study. Rather, the educational activities appear to have been incidental to the patient care services.

E. Structure of the Relationship

The agents found residents’ employment contracts, job descriptions, performance appraisals and compensation and benefits to be instructive in determining the nature of the relationship. The agents found that the structure of the relationship between a resident and his or her employer reflected the preeminence of the service aspect of the relationship.

Employment Contracts. The employment contracts/appointment letters resembled an employment contract of a typical non-student/employee. Residents entered into annual contracts with the employer setting forth their respective responsibilities and compensation and benefits. The annual contract typically stated that the teaching hospital will provide malpractice insurance; the resident agrees to supervise the services of other residents and medical students; and the resident is responsible for obtaining the required permits and licenses to practice medicine in the state. Annual contracts stated the bases upon which residents would be evaluated, such as competence and demeanor and behavior.

Job Description. The job descriptions also resembled what might be expected in the case of a non-student/employee position. For example, the responsibilities of a general surgery resident were listed as: 1) assessment of patients; 2) forming and carrying out a problem-based plan of care; 3) communicating promptly, respectfully and accurately with patients, family members and supervising physicians; 4) keeping pertinent, legible and timely records; 5) performing selected procedures with the appropriate level of supervision; and 6) educating patients.

Compensation and Benefits. The compensation and benefits that residents receive is typical of a non-student/employee relationship. Medical residents are typically provided health insurance, eligibility to participate in salary deferral arrangements under
§ 403(b), housing assistance or allowances, short term disability, workers’ compensation coverage, flexible spending accounts (cafeteria plans), discounts on auto/homeowners/renters insurance, employee assistance programs, meal allowances

20

while on call, free parking and uniforms. Residents also accrue sick leave and vacation time, receive unpaid leave under the Family Medical Leave Act (FMLA) and receive maternity leave.

The question arises whether the amount of compensation that residents receive is significant in determining student status. Residents generally received in the range of
$35,000 to $50,000 per year in salary plus benefits.20 Regulations § 31.3121(b)(10)- 2(b) provides that “the amount of remuneration for services performed by the employee in the calendar quarter . . . [is] immaterial. . . . The statutory tests are (1) the character of the organization . . . and (2) the status of the employee as a student . . . .”

The language that “the amount of remuneration for services performed by the employee in the calendar quarter . . . [is] immaterial” is best explained by the quarterly limit that existed when the student FICA exception was first enacted. Before 1950, services performed by a student enrolled and regularly attending classes for a S/C/U not exempt from income tax were not “employment” to the extent the remuneration for these services did not exceed $45 in a “calendar quarter”; however, remuneration for student services performed for a S/C/U exempt from income tax were not subject to a dollar limit per calendar quarter.21 In 1950, the quarterly limit on remuneration paid to an employee/student of a non-exempt S/C/U was eliminated and the separate student exclusion provisions for exempt and non-exempt entities were combined.22 We believe the regulations’ curious reference to the “amount of remuneration . . . in the calendar quarter” was intended to clarify that the $45 limit per quarter for services performed by students for non-exempt S/C/U is no longer in effect; it does not mean that the amount of remuneration is wholly irrelevant in determining student status. This view is consistent with Service position in Rev. Proc. 98-16. At § 3.04, the Service stated, “If the employee does perform services as an incident to and for the purpose of pursuing a course of study and, therefore, has the status of student, the amount of remuneration for services performed by the employee . . . [is] immaterial.”

20
According to the AMA, the average compensation of residents for 1998 was $38,177.61.
Characteristics of Graduate Medical Education Programs and Resident Physicians By Specialty, Table 11 (1999).
21
Social Security Act Amendments of 1939, Pub. L. No. 76-379, §§ 201, 606, 53 Stat. 1360, 1374-75, 1384-85 (1939). Section 1426(b)(10)(A) of the Code excepted from employment “[s]ervice performed in any calendar quarter in the employ of any organization exempt from income tax . . . if . . . (iii) such service is performed by a student who is enrolled and regularly attending classes at a school, college, or university.”

Section 1426(b)(10)(E) of the Code excepted from employment

[s]ervice performed in any calendar quarter in the employ of a school, college, or university, not exempt from income tax under section 101, if such service is performed by a student who is enrolled and is regularly attending classes at such school, college, or university, and the remuneration for such services does not exceed $45 (exclusive of room, board, and tuition).
22
Social Security Amendments of 1950, Pub. L. No. 81-734, § 104(a), 64 Stat. 477, 497, 531 (1950).

21

Indeed, concluding that the amount of remuneration is wholly irrelevant to student status ignores the legislative history which provides:

The intent of this exclusion is to exclude those persons and those organizations in which the employment is part-time or intermittent and the total amount of earnings is only nominal, and the payment of tax is inconsequential or a nuisance. The benefit rights that are built up are also inconsequential. Many of those affected, such as students and the secretaries of lodges, will have other employment which will enable them to build up insurance benefits. This amendment, therefore, should simplify the administration for the worker, the employer, and the government.

H.R. Rep. No. 728, 76th Cong. 1st Sess. (1939), 1939-2 C.B. 538, 543 (emphasis added).

In Rev. Rul. 85-74, the Service cited this language in concluding that whether “the total amount of earnings is nominal” determines, in part, whether an employee’s services are excepted from employment under the student nurse exception under § 3121(b)(13). As discussed, the Sixth Circuit in Johnson City Medical Center found Rev. Rul. 85-74 to be a reasonable reflection of the legislative history upon which we rely.

Medical residents’ compensation and benefits certainly are not “nominal,” and the FICA taxes on their salaries is more than “inconsequential.” Moreover, the benefits rights that are built-up are not inconsequential.23 While residents’ compensation is much less than they now earn, or will earn upon completion of their residencies, this is similar to many professionals and other highly skilled workers who earn a fraction in the early years of their career compared to their earnings after they have gained some valuable work experience. We do not want to overemphasize this argument, however, because we believe that while the amount of compensation and benefits has some relevancy, other factors have more force in determining the nature of the relationship, namely, the primacy of patient care in relation to the educational aspect of the relationship.

Performance Appraisals. A medical resident’s performance appraisal resembles that of a typical non-student/employee. At the end of a rotation, residents are evaluated by the attending physician. Residents are also evaluated annually by the department head. Residents are evaluated based on certain criteria such as clinical judgment, basic medical knowledge, technical skills, interpersonal skills, ascertaining and documenting health status and health risk factors, diagnosis skills, communicating effectively with patients and staff, providing appropriate supervision of patients and

23
The Social Security Administration reported to the Government Accounting Office that if residents are determined to not be subject to FICA tax, the expected loss of revenue to the OASDI trust funds will be $3.9 billion for the years 2001 through 2010. In addition, the SSA estimates that 270,000 medical residents will lose some coverage over the next ten years if medical residents’ services are excluded from coverage under the FICA. Social Security Coverage for Medical Residents (GAO/HREHS/GGD-00-184R, August 31, 2000).

22

support personnel, administering effectively, charting, history taking and continued professional growth and self-discipline. The narrative comments are typically directed toward the medical resident’s work habits, interpersonal skills, teaching skills, punctuality, attitude and fund of knowledge.

Dismissal. The causes for dismissal and the dismissal process are also similar to what might be expected in a non-student/employee relationship. A resident may be dismissed from a residency program, but that is very rare. Any attrition is generally due to residents quitting the residency program. If a resident is dismissed from a residency program, it is generally because of the quality of the resident’s clinical work, not the resident’s academic performance. In addition, grievance procedures also typically exist in the event an institution wishes to take an adverse action against a resident.

F. Contrast with Medical Students.

Because modern undergraduate medical training has a strong clinical component, it could appear that a residency merely continues a student status that we have acknowledged for M.D. candidates. However, the facts developed by the residents demonstrates that there are striking differences.

Medical students are enrolled for credit at a medical school and pay tuition. Medical students participate in clinical clerkships at a teaching hospital for which they receive grades and credits toward a M.D. degree. During the typical 2 years in clinical rotations, medical students learn to identify diseases and possible treatment plans and obtain basic knowledge of how hospital services function. They generally observe and perform only very simple procedures. They do not get paid for the services they perform; receive employee benefits; develop and initiate specific patient treatment plans; order tests and write prescriptions; make any independent medical decisions–such as changing medicine dosages; or take charge of a ward or other service areas, especially during the night. They are generally not even allowed to write notes for the medical record. In addition, the agents found that medical students at affiliated universities are not permitted to work more than 19 hours per week and do not receive benefits. Residents, unlike medical students, generally have temporary or permanent licenses. Absent this license, their responsibilities may be limited to those of a medical student. One house staff manual stated that residents who have not obtained their medical license must act as a medical student until the license is obtained.

G. Moonlighting.

Given the number of hours worked per week, it is difficult in many cases for medical residents to moonlight. However, in many residency programs residents with an unrestricted license to practice medicine may moonlight. As noted, a resident can generally receive an unrestricted medical license after completing one year as a resident in an accredited residency program and passing part three of the U.S. Medical Licensing Exam. Residents stated that while moonlighting they performed many of the same tasks that they performed within the residency program, such as patient physicals

23

and histories, writing patient care orders, ordering tests and medications, and performing routine procedures.

Fellows and residents nearing the end of their residencies are more likely to moonlight. For one OB/GYN program, medical residents indicated that moonlighting was reserved for residents in their third or fourth year. In one internal medicine program, a work study program was available during a four month block during which the resident was paid at the normal staff physician rate of pay of more than $100,000 per year. In another case, high performing residents were allowed to moonlight within the department and receive the pay of a regular staff physician. In some cases, agents noted that medical residents had W-2s reporting wages in excess of $100,000. This could be because of moonlighting, work study, or a resident who had completed a residency during the year and become a regular staff physician.

The FICA taxes with respect to moonlighting compensation in some cases were included in the refund claims. It is correct that if an employee of a S/C/U has the status of student at the S/C/U, then all of the compensation for services for the S/C/U are excepted from FICA. However, by the same token, all of the services must be considered in determining whether the services are incident to and for the purpose of pursuing a course of study. To use our extreme example again, if a resident earns
$100,000 (e.g., $50,000 from the residency and $50,000 for moonlighting at the institution), it would be difficult, if not impossible, to conclude that the resident’s services were incident to and for the purpose of pursuing a course of study.

H. Economic Aspects

Initially, we observe that medical residents provide patient care–the activity which is a hospital’s core business. We believe this fact is relevant in determining the nature of the relationship.

The agents commented on the economic relationship between medical residents and a teaching hospital, feeling that the economics color the nature of the relationship. The agents observed that a teaching hospital would require the services of other health care professionals, such as nurse practitioners, physicians’ assistants and additional staff physicians to replace the services of medical residents. Moreover, the agents noted that teaching hospitals receive approximately $100,000 per year or more for each medical resident in the form of Medicare and Medicaid payments.24

Medicare. The Federal Government makes payments to teaching hospitals in connection with resident services under Medicare part A and part B. Under part A, teaching hospitals receive reimbursement for a portion of the cost of GME. Medicare part A payments comprise two elements. First, Medicare makes “direct” payments to

24
Medicare payments for GME totaled $6.2 billion in 1999. In addition, states subsidize GME through Medicaid payments. Assuming there were 90,000 residents in 1999, that equates to approximately $69,000 for each resident. As discussed below, the amount received by a particular institution will vary in part based upon the number of Medicare patients at the hospital.

24 reimburse teaching hospitals for the stipends paid to residents and other program costs such as the salaries of supervising faculty.25 These direct GME payments reflect the product of three components. The first component is the teaching hospital’s direct GME costs for 1984, the “base-year,” as adjusted for inflation.26 The second component is the hospital’s current number of residents. The third component is Medicare’s share of the hospital’s inpatient days. Second, Medicare makes “indirect” payments in the form of a percentage add-on to the teaching hospital’s basic diagnostic related group (DRG) operating payments. The add-on percentage is based upon the hospital’s ratio of residents to its beds. Indirect payments are intended to reflect the higher costs that teaching hospitals incur per case because it is assumed they provide care that is generally more complex and technologically sophisticated.27

Teaching hospitals also are indirectly compensated for resident services under Medicare part B. The attending physician must be the physician of record for Medicare part B payments, but the resident often times provides the services or performs the procedure subject to the attending physician’s supervision.28 The agents also learned that in some cases the supervising physician could be a resident or fellow who is “credentialed” to perform the procedure. In addition, fellows in some cases have “attending level privileges” and thus act as the attending physician.

Taxpayers typically assert that although residents provide services, their services could more efficiently be performed by other health care professionals, such as nurse practitioners or physicians assistants. Leaving aside whether these workers can perform services such as reading film or performing surgery, anecdotal and empirical evidence suggests that use of residents may be beneficial to teaching hospitals financially.

Anecdotal Evidence. By way of anecdotal evidence, an agent noted that a GME committee in its recorded minutes expressed fear that reducing residents’ hours would create a financial burden for the institution. As further anecdotal evidence, In Boston Medical Center and House Officers’ Association/Committee of Interns and Residents, 330 N.L.R.B. No. 30 (1999), 1999 NLRB Lexis821, *202, the petitioner cited a 1994

25
42 CFR § 413.86.
26
The Centers for Medicare & Medicaid Services (CMS, formerly the Health Care Financing Administration) advised the Service that the base-year costs upon which reimbursement rates are computed presumably included FICA taxes on resident salaries, because FICA taxes would have been considered an allowable cost under the Medicare program. Thus, it appears that institutions are seeking refunds of some amounts that have already been reimbursed by the Federal Government.
27
42 CFR § 412.105. See Report to Congress by the Medicare Payment Advisory Commission, Rethinking Medicare’s Payment Policies for Graduate Medical Education and Teaching Hospitals, page xi (August, 1999).
28
42 CFR §§ 415.170, 415.172 and 415.174 provides the Medicare rules on the supervision required to bill for services under Medicare part B. In order to bill under Medicare part B, the attending physician be “present” during the “key portion” of a procedure. 42 CFR §§ 415.172. For certain services in a clinic setting, however, the attending physician’s presence is not required. 42 CFR § 415.174.

25 recommendation by a committee at the former Boston City Hospital stating that “[u]nder the current reimbursement system, the cost of house staff and attending physicians is virtually free. The cost of providing services without a teaching program would be significantly higher.” The budget report stated that all GME costs at Boston City Hospital were reimbursed, as well as a portion of indirect overhead costs. Overall, more than 82 percent of total GME costs were reimbursed. The report also noted that the cost of alternate providers such as physicians assistants and nurse practitioners was higher than the cost of residents, and the availability of alternative providers was problematic. The report concluded that a service delivery model without a teaching relationship did not appear to be a viable option.29

Empirical Evidence. Although some taxpayers assert that the use of medical residents is economically inefficient, there is empirical evidence that residents do allow teaching hospitals to carry out their mission more cheaply than if they did not use residents. First, we note that the Balanced Budget Act of 1997 changed the Medicare reimbursement system because the existing system allegedly encouraged teaching hospitals to train more physicians than the market could bear.30 In addition, we note three studies that have concluded that teaching hospitals benefit economically by employing residents to provide patient care instead of hiring other higher paid health care professionals. One study was conducted in New York in the late 1980's to determine the economic effect on teaching hospitals as a result of the reforms to combat the problem of resident impairment that arose in the wake of the Libby Zion incident. The study concluded that the services of up to 4,000 additional health care professionals, at a cost of $220 to $270 million annually, would be required to replace the lost patient care services by medical residents in New York due to the reforms.31 Another study concluded that the net cost of replacing resident services at D.C. General Hospital would be in excess of $17 million annually.32 Moreover, a recent study published by the American College of Chest Physicians concluded that “[p]atients
[greater than or equal to] 65 years old cared for by a faculty hospitalist service with the active participation of medical residents appear to have a 1-day [length of stay]

29
In addition, there are many cases considering whether residents’ stipends are excludable under § 117 in which the courts have nearly uniformly concluded that medical residents are compensated for the valuable services they provide to teaching hospitals. See, for example, Meek v. Commissioner, 608 F.2d 368, 373 (9th Cir. 1979) (“Although the hospital evidently could continue to provide medical care without the services of interns, the interns did perform valuable services which, if the interns were excused from performance, would have to be performed by others.”).
30
P.L. No. 105-33, §§ 4621 through 4630. Congress also enacted a program of incentive payments to encourage residency programs to reduce the number of residents.
31
Kenneth E. Thorpe, Director, Program on Health Care Financing, Harvard University School of Public Health, A Revolution in Graduate Medical Education: The Implications of Regulatory Reform in New York State (February, 1989).
32
Alan Sager, Ph.D., Professor of Health Services, Boston University School of Public Health, D.C. General Hospital Should be Renewed, Not Closed or Converted (Sept. 18, 2000).

26 reduction, significant total cost reductions, and significantly lower subspecialty consultation rates than comparable control subjects receiving routine private care.”33

Thus, the agents’ general impression of the economic relationship between residents and teaching hospitals is supported both by empirical and anecdotal evidence. We are not economics experts, and we recognize that some may dispute whether teaching hospitals save money from using medical residents. However, the evidence suggests that teaching hospitals do, in fact, benefit economically from the services that medical residents provide. Some teaching hospitals may have financial difficulties, but perhaps this is not because of using medical residents, rather it may be that some teaching hospitals have come upon hard times despite using medical residents.

I. On-the-Job Training.

Finally, we wish to emphasize our view that on-the-job training should not fall within the student FICA exception. The Greenbook states that “GME focuses on the development of clinical skills and professional competencies and on the acquisition of detailed factual knowledge in a medical specialty.”34 We are certain that residents gain in confidence, judgment, independence and experience as they go through their residencies. However, this describes what occurs generally in on-the-job training, and, indeed, in all work experience, especially for professionals and other highly-skilled workers. When asked to identify educational activities, residents routinely responded that every time he or she sees a patient, learning occurs. Indeed, a physician who is a radiation oncologist with many years of experience proclaimed: “I’m still learning.” Hopefully, we all are learning as we carry on our day-to-day activities. That, however, does not mean we are all students for purposes of the student FICA exception. We do not believe Congress could have intended that result.

J. Conclusion.

On balance, we have concluded that the services that medical residents provide are not incident to and for the purpose of pursuing a course of study. Instead, we believe that medical residents are engaged in a structured form of on-the-job training. We conclude therefore that it is appropriate to deny the examined refund claims.

33
Dani Hackner, The Value of Hospitalist Service, American College of Chest Physicians, Chest No. 2, Vol. 119, page 580 (February 1, 2001).
34
2000-2001 Greenbook, page 31.

27
Appendix

Finally, we wish to supplement our discussion in the April 19, 2000, memorandum of the significance of the repeal of the medical intern exception in 1965. Specifically, we believe it is important to consider the St. Lukes decision, which we believe provides context in which to view Congress’ actions in 1965.

The Medical Intern Exception. The SSA of 1939 amended the Code with respect to medical interns. Section 1426(b)(13) of the Code excepted:

Service performed as a student nurse in the employ of a hospital or a nurses' training school by an individual who is enrolled and is regularly attending classes in a nurses' training school chartered or approved pursuant to State law; and service performed as an intern in the employ of a hospital by an individual who has completed a 4 years' course in a medical school chartered or approved pursuant to State law.

The House Report provides:

Paragraph 13 excepts service performed as a student nurse in the employ of a hospital or a nurses’ training school by an individual who is enrolled and is regularly attending classes . . . ; and services performed as an intern (as distinguished from a resident doctor) in the employ of a hospital by an individual who has completed a four years’ course in a medical school chartered or approved pursuant to State law.

H. R. Rep. No. 728, 76th Cong. 1st. Sess. 49 (1939), 1939-2 C.B. 538, 550-51 (emphasis added); see also S. Rep. No. 734, 76th Cong. 1st Sess. 58, 1939-2 C.B. 565,

578. Thus, Congress distinguished interns from residents. This suggests that
     Congress intended to exclude interns from coverage, but not residents.

St. Lukes Hospital v. United States, 333 F.2d 157 (6th Cir. 1964). St. Lukes claimed a refund of FICA taxes based upon the student intern exception under § 3121(b)(13) of the Code. The years at issue were before 1965. The refund claims were computed based upon the remuneration paid to medical school graduates in their second or subsequent year of clinical training.

At trial, the plaintiff’s witnesses stated that in 1939 the ordinary and accepted meaning of the word “intern” was the same as the ordinary and accepted meaning of the word “resident.” The plaintiff asserted that although members of the medical profession used these terms differently, the public used these terms interchangeably. The Government countered with testimony that these terms had different meanings in 1939. It asserted that the term “intern” meant a medical school graduate in his or her first year of training, whereas the term “resident” meant an individual who had completed one year of training and endeavored to gain additional training to become a specialist. The trial judge sided with the plaintiff finding that it was appropriate to ascribe to Congress the public’s understanding of these terms, which was not to distinguish between the terms “resident” and “intern.”

28
On appeal, the Sixth Circuit found that the terms “intern” and “resident” had different meanings in 1939. The court noted, however, that between 1939 and date of the hearing some significant changes had taken place affecting the use of these terms. The court noted that the lines between interns and residents had blurred since 1939, but from 1939 to 1961 an “intern” had been regarded as an individual receiving hospital training during the first year following medical school, and the main qualification of a resident is the completion of an internship. The court noted that Congress in 1939 explicitly distinguished residents from interns by using the language “(as distinguished from a resident doctor).” Id. at 163. See H. R. Rep. No. 728, 76th Cong. 1st. Sess. 49 (1939), 1939-2 C.B. 538, 550-51.

The plaintiff persuaded the trial court that imposing FICA tax would lead to an absurd result because it would required residents to pay for something which would never be of any benefit to them. However, the appellate court received information from the Social Security Administration that a resident could benefit in the form of disability benefits and death benefits to the resident’s family. Id. at 163-64.

Finally, the court noted that exceptions from social security coverage are to be narrowly construed. Id. at 164

The court concludes the opinion by stating:

In all of the above we do not ignore the fact that distinctions between interns and residents-in-training have been substantially reduced in the years since 1939. The resident training program has been greatly expanded and its educational aspects have been greatly enhanced. No doubt these developments lend some weight to the argument for expansion of the intern exemption to cover residents-in-training. It seems clear to us, however, that meeting these changed conditions, if indeed there is warrant for dong so at all, is the function of legislation and not that of judicial interpretation.

Id. (emphasis added).

As discussed below, the Social Security Amendments of 1965 (SSA of 1965) repealed the medical intern exemption, covered medical doctors under SECA, and covered medical residents and interns working in federal hospitals under the FICA. These changes affecting medical doctors are arguably Congress’ response to St. Lukes. That response arguably was: “Not only do we think that medical residents should be covered under the FICA, as the court in St. Lukes held, but we believe that interns should be covered as well.”

The Social Security Amendments of 1965. The legislative history underlying the SSA of 1965, Pub. L. No. 89-97, suggests that Congress intended that medical residents be covered under the FICA. Section 311(b)(5) of the SSA of 1965 amended
§ 3121(b)(13) by striking the medical intern exception.

With respect to the repeal of the medical intern exclusion, the Senate Report states:

29

Section 3121(b)(13) of the Internal Revenue Code of 1954 excludes from the term “employment,” and thus from coverage under the [FICA], services performed as an intern in the employ of a hospital by an individual who has completed a 4-year course in a medical school . . . .
Section 311(b)(5) amended section 3121(b)(13) so as to remove this exclusion. The effect of this amendment is to extend coverage under the [FICA] to such interns unless their services are excluded under provisions other than section 3121(b)(13). Thus, the services of an intern are covered if he is employed by a hospital which is not exempt from income tax as an organization described in section 501(c)(3) of the Code.

S. Rep. No. 404, 89th Cong. 1st Sess. 237-38 (1965). The last sentence makes indirect reference to the exclusion from FICA for services performed for exempt organizations under § 3121(b)(8)(B) of the 1954 Code. That exclusion was repealed by the Social Security Amendments of 1983 (Pub. L. No. 98-21). Nothing in the legislative history indicates that Congress believed interns (or residents, who were even further along in their medical careers than interns) were eligible for the student FICA exception.

In addition to revoking the medical intern exception, § 311 of the SSA of 1965, entitled, “Coverage for Doctors of Medicine,” changed the law in two other ways which affected medical doctors. First, § 1402(c)(5) of the 1954 Code was amended to eliminate the exception from the definition of “trade or business” for physician services, thus making these services subject to self-employment tax. Second, § 3121(b)(6)(C)(iv) of the 1954 Code, which provided an exclusion from the definition of employment for “service performed in the employ of the United States if the service is performed by any individual as an employee included under § 5351(2) of title 5, [U.S.C.], (relating to certain interns, student nurses, and other student employees of hospitals of the Federal Government),” was amended to add, “other than as a medical or dental intern or a medical or dental resident in training.”

These provisions, taken together, appear to indicate Congress’ intent to create a scheme under which all medical doctors are covered under the social security system, whether or not they are still in training, whether or not they are self-employed, or whether or not they work for the federal government.