The literature examining the recent trends in specialization is, at best, conflicting about why young physicians are choosing careers outside of generalism. Compensation, debt, lifestyle, prestige, personality, and demographic factors have all been identified as likely issues but their relative importance is not clear.

In this issue, McDonald and colleagues (5) have attempted to better demonstrate the relationship between student debt and specialization among internal medicine residents. Using the Internal Medicine In-Training Examination Residents Questionnaire, the authors related the career plans of internal medicine residents in their final year of training to their educational debt. With 74% of all potential respondents participating, the authors found that, as debt level increased, residents were less likely to choose subspecialty careers.

This surprising result is consistent with another empirical study done on a similar population (6), but it is in stark contrast to the opinions of many authors who assert that money, whether in the form of higher debt or lower income, is the primary reason for a lack of interest in primary care (7). McDonald and colleagues' data are consistent with years of graduation questionnaire data from the Association of American Medical Colleges, which suggest that debt is far less important than other factors, including mentorship, lifestyle, options for fellowships, salary expectations, family expectations, length of training, and the competitiveness of a specialty (8).

Of most interest, international medical graduates (IMGs)—who now account for half of all trainees in internal medicine and family medicine—were also more likely to initially choose generalist careers (including hospital medicine) if they had higher debt loads, although many may be restricted to particular specialties because of service obligations. It is also unclear whether those with high debt levels are U.S. IMGs who may behave more like domestic graduates and whether U.S. graduates and IMGs may eventually choose different career paths from the ones they decide upon at the end of residency.

Among many explanations for the counterintuitive result of this survery, several stand out. One is self-selection. McDonald and colleagues have examined internists in training, a unique subset of physicians, many of whom may have already self-selected out of medicine's most lucrative specialties (after already opting out of more profitable nonmedical careers). Although profit maximization is the most rational economic theory for physician behavior, some physicians may be satisfied with achieving a minimum "target income" while balancing their own interests and preferences for professional life (9, 10).

These preferences may include "lifestyle" factors that relate to flexibility of scheduling, on-call responsibilities, total work hours, and others. The Association of American Medical Colleges' recent survey of physicians younger than age 50 years suggests that younger physicians' practice plans are highly motivated by time for personal and family activities (11). Other research has shown primary care physicians are perceived to be better able to control work hours, setting of practice, and call schedules (12). However, young physicians differ in the importance they place on lifestyle and practice income. For instance, internal medicine residents who choose cardiology, gastroenterology, or nephrology seem to be more likely to report higher income as a reason for specialty selection than those who choose other specialties (13). Women, who now make up nearly half of all graduating medical students, seem to be less driven by income compared with lifestyle or other factors (13).

The educational and training system has not influenced interest in generalist careers. Without substantive change in mentorship or curriculum, generalism saw an upsurge in interest among U.S. graduates during the 1990s when most were told, implicitly or explicitly, that managed care would create a future health care delivery system oriented toward primary care. Conversely, anesthesia, among predictions of future surplus, experienced a precipitous drop in applications to residency programs. Multivariate analysis of medical school graduates in 1995 suggested that demographic factors were more important to choosing a career in generalism than were generalist missions, curricula, or admissions preferences for students with a stated interest in generalist careers (14). The surge in interest in primary care was short-lived, and young physicians quickly altered their career paths, again with no antecedent, sweeping changes in education and training.

Physician specialty choice may be less influenced by curricular change than by students' experience with the work of a particular specialty and how well it matches their interests and goals. Young physicians have a realistic picture of clinical practice but not a clear picture of how the system may evolve. Generalists are faced with the challenge of maintaining expertise in an ever-expanding knowledge base, whereas specialists can focus on refining their knowledge of a narrower scope of illness while using the most advanced technology available. More specialties are also available to choose from: Before 1980, only 20 subspecialty certifications existed—today, physicians can be certified in more than 145 specialties and subspecialties (15).

Young physicians are also faced with uncertainty about the future of primary care and how their work will be distinguished from the work of other health professionals who may have fewer years of formal education and training. The growth of the elderly patient population, which will bring with it the challenges of treating a future patient population with chronic diseases not easily treated in a 15-minute visit, may be less attractive to many young physicians who opt for a specialty where they can "successfully" treat a discrete problem in a short period of time.

Given these many influences on specialty choice, young physicians have a very complex decision task. They must somehow factor these influences into the desire to help others that propelled them into the field of medicine—a process that each young physician will carry out in his or her own unique way.

None of this suggests that the influence of financial incentives should be ignored, nor should policymakers shirk their responsibility to influence and reflect societal values in payment rates, because more than half of care in the United States is publicly funded (16). However, the field of internal medicine, and workforce researchers, may need to look beyond financial issues to better understand what attracts physicians to more specialized areas of practice and work to address not only income but also how we as general internists and specialists practice medicine—and what this means for patients. Those who look only to blame policymakers and payers may do future patients and physicians a great disservice.