Sunny G. Yoder
The purpose of this paper is to examine how physicians in the United States are compensated. This is a matter of some interest since the method of compensation, that is, how the physician ultimately receives payment for his or her services, is an important element of the physician's economic environment. Another important element is the system of third-party payments: the policies and practices of government programs and private insurers defining the services for which they will pay, the conditions for payment, the form of payment, and its amount. Taken together, these two elements largely determine the relationship between the number and types of services a physician provides and the physician's gross income.*
For most solo practitioners, the payment method and the compensation method are identical, so that the relationship between payments for services provided and the physician's income is direct. For practitioners who are part of a group practice or HMO, however, the payment method generally differs from the compensation method. For example, although the physicians in a group practice may bill patients and third parties on a fee-for-service basis, individual group members may receive a salary. In an HMO, where payments are on a capitation basis, physicians also may receive compensation in the form of salary. In such instances the relationship between payments for the physician's services and the physician's income is more complex.
Students of the U.S. health care system express considerable concern about the economic incentives for physicians under different forms of compensation. While none of these observers believe that physicians conduct their medical practice solely for pecuniary gains, they do believe--and there is some, albeit limited, evidence to support this belief--that such motivations do influence physicians' behavior. To understand the nature of a physician's economic incentives it is necessary to understand how the physician is compensated as well as how payments for services are generated. This paper describes the different physician compensation methods, presents published data on levels of compensation, and estimates the distribution of U.S. physicians among the different compensation methods. The latter estimate is constructed in three steps. First, published data from the American Medical Association Masterfile are used to group physicians on the basis of their reported employment setting and professional activities. Second, for each of the groups, available evidence is used to estimate how the physicians in the group are distributed among compensation methods. Finally these estimates are combined to estimate the overall distribution.
Distribution of U.S. Physicians Among Employment/Activity Groups
According to the American Medical Association Masterfile (see Table 1), in 1980 there were approximately 415,000 active physicians who provided sufficient data to be categorized by employment setting and main professional activity. Four-fifths of all physicians were either residents enrolled in graduate medical education programs (15%) or were engaged in office-based practice (65%). Most of the remainder were federal employees, hospital-based practitioners, or were engaged primarily in teaching, administration, and research. A description of the methods and levels of compensation for each group in Table 1, based on available information, follows.
Table 1
Number and Percentage Distribution of U.S. Physicians by Employment/Activity Category, 1980.
Residents. In 1980, 62,000 physicians were enrolled in graduate medical education programs. Their compensation is in the form of salary, sometimes referred to as a stipend (Association of American Medical Colleges, 1980, page 169). Compensation levels for these physicians have grown significantly from the era when they literally resided in the hospital. Today residents' salaries reflect their apprenticeship role in providing medical services to hospitalized patients, ranging from a median of $18,900 for the first post-M.D. year to $23,200 for the 5th post-M.D. year in 1982-83 (Department of Teaching Hospitals, 1982).*
Federal physicians. In 1980, 15,360 physicians (excluding residents) were employees of the Federal Government. Roughly 7000 of these were in the military service, and another 6000 employed by the Veterans Administration. The remaining 2400 federal physicians were employed by the Public Health Service (including the Commissioned Corps) or other parts of the Federal Government (Kahn and Orris, 1982; Eiler, 1983). The basic form of compensation for these physicians is salary, although the details of compensation arrangements vary depending on where the physician is employed. In the military and the Commissioned Corps, the base salary is augmented with non-taxable housing and subsistence allowances, dependents allowances, and bonus pay for board certification and years of service. In addition, these physicians receive a lump-sum retention bonus for each year they remain in the service. Taking all these components of compensation into account, a physician entering the military or the Public Health Service Commissioned Corps earns approximately $40,000. Elsewhere in the Federal Government physicians are paid according to special GS pay levels, which at entry are $31,900 for GS-11, $32,000 for GS-12, and $45,400 for GS-13. An exception is the Veterans Administration which has its own pay schedule. Physicians may enter the VA at a senior grade, equivalent to a GS-14, at $41,277 per year, or at a chief grade at $48,553. In addition, the VA also has a system of special pay for board certification, tenure, responsibility level (e.g. service chief), and geographic location. This special pay can add as much as $22,500 annually to VA physicians' earnings.
Office-based practitioners. The nation's 271,000 office-based practitioners are compensated in several different ways, depending upon whether they are in solo or group practice and, if in a group, how the group's practice income is distributed. According to estimates from the AMA Periodic Survey of Physicians (a detailed survey of a 5% sample of office-based, non-federal physicians), 54 percent were in solo practice and 46 percent in group practice in 1980 (Kahn and Orris, 1982, page 286). The physicians in groups reported the following income distribution methods:
| Equal distribution | 11% |
| Straight salary | 8% |
| Salary plus profit share | 16% |
| Fee-for-service | 7% |
| Other or unknown | 4% |
| 46% |
(Kahn and Orris, 1981, page 286.) These percentages should be viewed as very rough estimates, subject to considerable error due to a relatively low (50%) response rate for the survey. Too, these categories are highly simplified characterizations of income distribution plans that often are quite complicated (see, for example, Medical Group Management Association, 1978).
Information on the net incomes of office-based practitioners in 1980, by specialty, are available from the Periodic Survey of Physicians. As published, these figures do not differentiate between physicians under different compensation methods, but are reported by specialty. As summarized in Table 2, physician net incomes in 1980 ranged from about $63,000 for pediatricians and general practitioners to about $99,000 for surgeons. Overall, average net incomes rose at a compound rate of 6.8 percent between 1970 and 1980. Average net income for all specialties, according to this survey, was $80,900 in 1980 (more recent AMA figures on net incomes of all practicing physicians are presented in a later section).
Table 2
Average Net Income from Medical Practice by Specialty, 1970, 1980, and Compound Percentage Growth 1970-1980.
Hospital-based practitioners. Roughly 31,000 non-federal medical practitioners are hospital-based. These physicians are concentrated in certain specialties, primarily radiology, anesthesiology, and pathology. Information on methods of compensation for these physicians was collected in a special 1979 hospital survey by the American Hospital Association. Steinwald (1983) summarized the compensation methods for these physicians as salary, percentage arrangements, and fee-for-service. Table 3 shows the distribution of hospital departments of anesthesiology, pathology, and radiology among these methods.
Table 3
Percentage Distribution of Hospital Departments By Compensation Method.
If one assumes that these distributions apply to the physicians in these specialties (an assumption that is partially supported by comparing Steinwald's results to those of a 1979 survey by the American College of Radiology) and weight these distributions by the number of physicians in each specialty, a rough estimate of how hospital-based physicians are compensated would be as follows:
| Salary | 20% |
| Percentage | 18% |
| Fee-for-service | 62% |
This estimate applies only to hospital-based physicians, that is, physicians who spend the majority of their time in hospital-related activities. Among physicians who have any financial arrangements with a hospital, almost 60 percent have a salary arrangement (AMA, 1982).
Hospital-based physicians are among the highest earning specialties in medicine. Net practice income for radiologists for 1982, as estimated by the AMA, were almost $137,000; anesthesiologists were estimated at $131,400. Pathologists' earnings were not separately reported (AMA, 1983). Steinwald analyzed the relationship between compensation methods and earnings levels for these physicians. Radiologists and anesthesiologists who received fee-for-service compensation had the highest gross, and net, incomes; salaried physicians in these specialties had the lowest incomes. The highest earning pathologists received a percentage of department revenues, while the lowest-earning pathologists were salaried (Steinwald, 1980, Table 4, page 72). Recent restrictions on reimbursem*nt levels for hospital-based physicians by the Health Care Financing Administration may lessen the differences in incomes between salaried hospital-based physicians and the others.
Incomes of all practicing physicians. In 1981 the AMA instituted a new series of surveys that include all practicing physicians, both office-based and hospital-based. Data on physician net incomes from this new Socioeconomic Monitoring System therefore are not strictly comparable with those from the former Periodic Survey of Physicians, and thus conclusions about trends in incomes must be made cautiously. In particular, since hospital-based physicians tend to be in the higher ranges, the new figures may overstate gains in income between 1980 and 1981. Table 4, based on the new survey, shows physicians' earnings for 1981 and 1982, and the percentage increase from 1981 to 1982.
Table 4
Average Net Incomes of Practicing Physicians by Specialty, 1981, 1982, and Percent Increase 1981-1982.
There is considerable variation around these averages, depending upon length of time in practice, practice mode and form of compensation, geographic location, and the extent to which a physician's services and patients are covered by health insurance. In his or her prime earning years, e.g., between the ages of 35 and 55, the average physician would be earning about $110,000, rather than the overall average of $99,500. These data indicate that 20 percent of physicians had net incomes of under $50,000 in 1982, while 25 percent earned over $125,000. Net incomes exceeded $200,000 for approximately 7 percent of physicians.
Academic physicians. Most of the non-federal physicians whose main professional activities are teaching, research, and administration are employed as faculty in the nation's medical schools. These physicians typically are compensated in one of two ways: (1) a fixed salary or (2) a base salary with the opportunity for supplemental earnings from medical practice. The degree of medical school control over practice earnings varies a great deal. Some medical faculty simply bill and collect their own fees independent of the school. However, in the majority of medical schools, practice earnings are channeled through an organized faculty group practice, or practice plan. These plans, not unlike private practice groups, have rules governing the collection and distribution of practice earnings among the group members and the medical school (Institute of Medicine, 1976; Hilles and fa*gan, 1977). Legally, they may be independent corporations, partnerships, or administrative units of medical schools (Jolly and Smith, 1981).
Data on medical faculty salaries are collected by the Association of American Medical Colleges and reported, by faculty rank, for the two forms of compensation. In 1983, 53 percent of faculty received a fixed salary, while 47 percent received a base salary plus supplemental compensation from practice. Their annual salaries, according to these data, were as follow:
| Salary | Base + supplement | |
|---|---|---|
| Instructor | $42,000 | $51,300 |
| Assist. professor | 62,700 | 66,000 |
| Assoc. professor | 64,100 | 79,900 |
| Professor | 71,600 | 94,100 |
| Chairman | 78,100 | 122,500 |
(Smith, 1983, Tables 5 and 6). For faculty who receive supplemental earnings from practice, these data include only those whose supplemental income is actually reported. Since many medical faculty have uncontrolled--and therefore unknown--outside earnings, the figures above probably understate average earnings for this group.
Other physicians. According to the AMA, “Other” physicians are those who work in insurance companies, corporations, pharmaceutical companies, voluntary organizations, medical societies, and other organizations. Presumably they work as salaried employees in these settings; no data are available on their earnings.
Overall payment methods. The data presented above, with some assumptions, can be used to construct a rough estimate of the numbers of physicians compensated by each of three basic methods: (1) salary, (2) incentive, and (3) fee-for-service. The salary and fee-for-service methods are self-explanatory. “Incentive” refers to physicians whose incomes are positively influenced by revenues they generate, including medical school faculty receiving salary supplements, hospital-based physicians under percentage arrangements, and group practice members receiving incentive compensation. Assuming that all solo office-based practitioners are fee-for-service, that all federal physicians are salaried (since their bonus arrangements are independent of the volume of services rendered), and distributing the rest of physicians among compensation categories according to the percentages presented above, I estimate that in 1980, excluding residents, 63,000 (18%) of physicians were salaried, 105,000 (30%) received incentive compensation, and 185,000 (52%) were fee-for-service.* If federal physicians also are excluded, then these percentages would be salary 14%, incentive 31%, and fee-for-service 55%.
Appendix
Estimated Distribution of Physicians by Work Setting and Type of Compensation, 1980.
Comparison with Other Estimates. Kahn and Orris (1982) employed a similar approach to estimating the distribution of U.S. physicians among compensation methods, but their results differ considerably from those reported here. They estimated that, including residents, approximately 53 percent of active physicians were salaried (my estimate, including residents, is 30%) and 47 percent were compensated by other methods (I estimate 70%). The estimates differ for several reasons. Kahn and Orris used two compensation categories, “salary” and “other”, rather than three, and included under salary a number of physicians such as medical school faculty who received salary plus incentive payments. Too, they assumed that 100 percent of hospital-based physicians, in comparison to my estimate of 20 percent, were salaried. Finally, their estimates are based on 1979 rather than 1980 Masterfile data.
Gabel and Redisch (1979) have estimated that, including residents, 71 percent of U.S. physicians are paid under fee-for-service, and 28 percent are on salary. They do not report their method of arriving at these figures, but the distribution is very close to mine if they counted physicians receiving incentive compensation with those on straight fee-for-service. A substantial number of physicians receive a mixed form of compensation; how they are counted makes a great deal of difference in estimates such as these.
The data presented here do not support any firm conclusions about the relationship between physicians' compensation methods and incomes. Physician earnings are affected by many variables, and a careful analysis is needed, taking into account specialty, years in practice, geographic location, hours worked, and other variables, in order to establish any systematic relationship between how physicians receive their income and their income levels. In any event we can be sure that the relationship is a subtle one, since a physician's earnings are not independent of his/her choices among specialties, locations, or practice modes, nor of decisions about medical practice. Even salaried physicians' incomes often bear some relation to the quantity of services they provide, even though the relationship is less direct than for fee-for-service practitioners. As one observer notes, to assess the economic incentives in salaried practice it is necessary to consider the incentives and reward structure of the institution paying the physician's salary (Reinhardt, 1983).
