Skip Navigation

The Physician Gap: How we can encourage US physicians-in-training to choose primary care

The US healthcare system is a terrible value proposition. Compared to peer nations like France, Germany, and South Korea in the Organization for Economic Co-operation and Development (OECD), the US spends nearly twice as much on health care, but with few measurable differences in health outcomes. A 2015 analysis by the OECD shows that US life expectancy is now one year below the OECD average, while in 1970 the US was one year above average; the report attributed the discrepancy to “the highly fragmented nature of the US health system, with relatively few resources devoted to public health and primary care.” Furthermore, the life expectancy gap between the US and leading countries is increasing: US men can expect to live 4.3 years fewer than their Swiss counterparts, and US women 5.4 years fewer than Japanese women. The US also does poorly in terms of equitable distribution of health outcomes: socially disadvantaged populations bear disproportionate burdens of disease and injury.  We find a clear example of this when looking at infant mortality: in 2012, black infant mortality was more than double that of white infant mortality, a figure that propelled the US into the bottom half of OECD nations for general infant mortality. The discrepancy is stunning: for every six dollars of GDP, the US spends approximately one dollar on health care. Why do we still fare just as badly – or even worse – compared to nations whose health expenditures are only a single-digit percentage of their GDP?

One reason for the US’s sub-par health care system is that the US has a much greater proportion of specialists – dermatologists, surgeons, radiologists – than its peers. In 2016, the OECD reported that in the US there were more than two specialists for every general practitioner; by comparison, in countries like Australia and Canada, generalists make up around half of the physician workforce. (Generalists are much less expensive than specialists: the estimated annual income of a primary care doctor in 2017 was $217,000 while the typical specialist earned $316,000.) From a policy perspective, this price discrepancy implies that an increasing supply of specialists contributes to corresponding increases in health care expenditures.

What, then, is behind this trend of few physicians-in-training choosing to become generalists? A significant factor is the large and growing financial incentive to specialize. To fix the physician workforce imbalance, we need to close the widening financial gap between specialists and primary care physicians. The combination of relatively low reimbursement and high workload (due to the shortage of primary care doctors) make generalists among the unhappiest physicians in the profession. Primary care physicians’ annual incomes are roughly equal to their medical school debt; they work more than 50 hours per week, and less than a fifth would choose primary care again if they had the opportunity to redo medical school.

As mentioned previously, Canada has been successful in encouraging more equal numbers of aspiring physicians to become generalists and specialists. Much of this can be attributed to the smaller disparities in expected incomes in the two fields. In 2011, a US orthopedic surgeon made 2.4 times more than a primary care physician while a Canadian orthopedic surgeon made 1.6 times more than a Canadian primary care physician. In the US, the estimated lifetime earnings for surgery subspecialties was 1.6 million dollars more than primary care doctors. This discrepancy discourages aspiring physicians to become generalists — they face similar debt and training times but for significantly less pay.

According to an article in Family Medicine, “Primary care career choice plummeted to current low levels as the income ratio fell from 0.78 in 1985 to 0.50 in 2017. The strong historic association between income ratio and specialty choice, when combined with anticipated professional return on investment for extra years of training, suggests that a ratio of about 0.80 will likely result in the 40% target of primary care career choice.” The correlation between a growing income ratio and a decreased number of primary care physicians suggests that increasing generalist compensation (or decreasing specialist compensation) in the US will aid in efforts to re-balance the current ratio.

Canada’s 1:1 ratio of generalists to specialists is not an accident. In 2004, Canada faced its own physician workforce pipeline problem when only roughly a quarter of medical students expressed interest in primary care. Reform in care coordination payments led to increases in family physician income from 40 to 60 percent, and the current ratio of generalist to specialist income is approximately 0.83. When the U.K. faced a similar problem in 1996, it introduced reform in payments for quality outcomes, which increased the ratio of family physician to specialty income to close to 1.0 in 2004. At the same time, the percentage of medical students opting for family medicine increased from 15 to 35 percent. In the US, there is a similar opportunity to reform medical reimbursements: Medicare payments are currently calculated based on the relative value unit (RVU) of various procedures, and RVUs are updated by a committee of physicians whose 31-member board currently consists of three non-subspecialty primary care doctors.

One popular argument against increasing physician reimbursement for primary care physicians is that although income disparities are great, differences in debt do not seem to affect specialty choice. Although many medical studies have attempted to draw a correlation between medical school debt and the choice to specialize (or the choice of subspecialty), this correlation is still not well-documented. One possible explanation for this is that many aspiring physicians are already financially well-off – medical school students are a self-selecting group – and thus wealthy students with no debt may still choose high-earning specialties because they expect high lifetime incomes. This theory is supported by the fact that large amounts of medical school debt is shown to deter public medical school graduates and students from lower-income families from entering primary care, but not private medical school graduates. At the same time, students from lower-income families are also more likely to be interested in and choose primary care or family medicine than students from higher-income families.

Even if we cannot draw a definite connection between medical school debt and specialty choice, a clear relationship between students’ subjective income expectations and specialty choice does exist. A 2001 study by the Wharton School demonstrated that a $10,000 increase in expected specialist income relative to expected generalist income was linked to a six percent increase in the probability of entering a specialist profession. Furthermore, a report by the Council of Graduate Medical Education to Congress in 2008 showed a direct correlation between higher salary and higher fill rates by US graduates. Accordingly, in 2007, less than half of family residents were US medical graduates, while more than 90 percent of orthopedic surgery residents were US medical graduates. Again, by decreasing the compensation gap, we would reduce the financial incentive to specialize.

This fact, coupled with the fact that the gap between the incomes of specialists and generalists is already wide and continuing to grow, does not bode well for the future of the US physician workforce. In fact, primary care income is decreasing: from 1996 to 2003, generalists’ income decreased by 10 percent (adjusted for inflation).

Unfortunately, the US has not implemented serious policies to encourage medical school students to enter primary care. Without action, the primary care crisis will worsen.  The number of US medical graduates choosing internal medicine residencies at all is alarmingly low: around 3,900 students matched internal medicine in 1985, compared to 200 in 2014. Even among those who do choose to pursue internal medicine residencies, many – more than half – eventually join subspecialties such as rheumatology. And, although graduates of internal medicine residencies were twice as likely to report interest in becoming generalists, many do not follow through on that “early promise.”

This issue is especially politically relevant because if the US does not control medical costs, even continuing to fund Medicaid and Medicare may soon become unsustainable. In 1950, health care expenditures accounted for fewer than five percent of gross domestic product (GDP); in 2015, that figure was 17.8 percent or roughly $9,900 per person. These costs have steadily increased every year with the exception of a short period in the 1990s when the introduction of aggressive managed care — a system to reduce healthcare costs through a variety of “management techniques” like restricting access to certain doctors or hospitals and requiring authorization from a primary care doctor to see a specialist — temporarily curbed growing medical expenditures. Furthermore, more and more of that burden is not borne by individuals themselves: higher-income individuals typically have employer-sponsored (ESI) health care insurance which shields them from increasing insurance premiums. Those with Medicaid and Medicare, however, are insulated from costs because the government pays for the vast majority of these expenditures. The government, however, cannot reasonably continue to foot a bill which is both increasing steadily in cost per person and being asked to support an ever-growing number of people as the US population sickens and ages.

Of course, a shortage of primary care doctors cannot by itself explain the poor performance of the US healthcare system relative to its peers. Income inequality, weak environmental regulations, and a lack of emphasis on public health all contribute to our comparatively bad health. In the same way, the current disinterest in primary care is linked to a number of factors, including a lack of prestige and fewer opportunities to practice independently. But from a political perspective, much of this is impossible to change: for instance, medical students’ perception of the relative prestige of these professions.

What we do have control over is how much our different kinds of doctors are paid. (For the most part, private insurance companies follow the pattern of payment set by Medicare and Medicaid reimbursement rates.) From that perspective, there is only one answer. We currently face soaring healthcare costs that consume a sixth of our GDP, a higher percentage than any other nation. We suffer from inefficient care that over-emphasizes specialists completing procedures (e.g. Mohs surgery) at the expense of valuing primary care physicians, who build long-term relationships with their patients and can champion adoption of healthy, disease-preventative lifestyle changes. And if primary care doctors are so critical to the viability of our healthcare system, we need to compensate them accordingly — not pay then half of a specialists’ salary.


About the Author

Ashley Chen '20 is the President of the Brown Political Review and a Senior Staff Writer for the US Section. Ashley can be reached at