The generally accepted target is a physician work force that is not much larger than the present one but that comprises at least 50% generalists [5]. The number of physicians in the United States—there is now 1 physician for every 420 persons—will increase if the number of medical schools, the number of medical students, and the rate at which physicians from other countries enter U.S. practices remain the same. To control the total number of physicians, some expert groups have recommended placing a cap on the number of residencies offered—for example, 110% of U.S. medical graduates [5].

If the number of physicians remains about the same and the proportion of generalists increases, then the number of specialists must decrease. The internal medicine community has voiced its commitment to general internal medicine [6, 7] but has been less attentive to the question of how the specialist work force might be reduced without diminishing the remarkable strength of subspecialty medicine in research, teaching, and the care of some patients.

The United States can reach its physician work force goal in two ways. One is to change medical education so that more medical students either choose to become generalists or are unable to become specialists. The other is to rely on the marketplace to change the incentives in medical practice so that some physicians who are now subspecialists will choose to become generalists.

The article by Christakis and colleagues in this issue [8] is about this second option. The authors describe changes from specialist to generalist and from generalist to specialist practice among the 335 438 physicians in active practice between 1982 and 1986, before much national attention was given to health care system reform. Movement occurred in both directions but, as expected, with a net gain in specialists. Physicians of all kinds switched from specialist to generalist fields, even at a time when there were far fewer incentives for doing so than there are now. The change was concentrated in one group: About 15% of physicians who were not generalists had a secondary interest in generalist practice, and this group constituted 65% of the new generalists.

Producing more generalists has received most of the attention so far [5]. The advantages of this approach are that it is evolutionary, depends on individual choice, and fosters training to match careers. Also, this approach seems feasible because it can rely on central control over the production of specialists. Medical schools and residencies respond to accreditation requirements and the not-so-gentle persuasion of the federal government, which provides funding for medical education through research grants and payment for patient care. The government can, if it chooses, institute both financial incentives (such as new grant programs and new ways of allocating payment for cognitive and procedural services) and disincentives (withdrawal of funding) to shape medical schools to its tastes.

However, reshaping the work force through training alone is a slow process. The training pipeline is long, and physicians entering the work force are just a small proportion of those already in practice. If, beginning today, 50% of medical school graduates enter generalist practice, it would take until the year 2040 before 50% of physicians were generalists [9]. This is too slow for an urgent national problem.

The other way of increasing the proportion of generalists is by inducing subspecialists now in practice to become generalists. This is already happening. Managed care organizations, which command a growing proportion of jobs for physicians, have sought to meet their needs for primary care physicians by offering better pay, fringe benefits, and quality of life [10]. They are also making fewer places for some specialists than were available in their regions in a less structured, fee-for-service environment. These policies are affecting the career choices of physicians in practice and perhaps also those of medical students.

It is not that individual managed care organizations care so much about the nation's physician work force; it is simply that they need more and better generalists and fewer specialists to compete effectively in the marketplace. The sum of incentives and disincentives they offer to generalists and specialists in a largely private system of care is, in effect, national policy.

Changes in the marketplace have been so recent and fast-moving that there are few studies of their magnitude and location. But informal reports are numerous enough to make it clear that fundamental changes are taking place. With so little hard information to go on, the article by Christakis and colleagues is especially welcome.

If the marketplace accelerates the rate at which subspecialists of internal medicine become generalists, will all be well? Not necessarily. There is no reason to believe the marketplace will control the total number of physicians, at least in the short run [11]. Also, subspecialists are socialized by training and experience to value their role in health care [12]. If they are forced into generalist roles, their hearts may not be in their work. They may not be prepared to give high-quality general care, for example, for contraception, incontinence, depression, dermatitis, and prevention of atherosclerosis or colorectal cancer. Some subspecialists have continued to see patients whose main problems are outside their subspecialties but many have not, and have limited their contact with colleagues, journals, continuing medical education courses, and recertification outside their subspecialties since completion of their residencies, which may have taken place decades ago.

What are the lessons for internists in practice? Physicians must now be willing to change practice styles or even jobs, just as workers on the assembly line, in the executive office, or on the farm have been. More physicians will work in systems of care rather than alone or in small independent groups, and they must appreciate the basic goals of these organizations—prevention, efficiency, and fairness to all subscribers. Understanding among the various disciplines of medicine is needed more than ever—by the generalists, who are still relatively weak in numbers, prestige, and earning power; by the specialists, who are threatened with a decline in options and control; and by the administrators, who may be tempted to use their clout to force changes in the practice of medicine with insufficient understanding of how those changes may affect the care of patients.

In the midst of all the trouble and uncertainty, physicians should pause to reflect on what is special about their work. They are respected and are well-paid to help people. They have the intellectual stimulation of integrating their knowledge of the biology of disease, the human condition, and the environment of medical care. They retain great latitude in clinical decision making. For both specialists and generalists, these benefits are likely to endure. Few workers are more privileged.