Given that physicians are some of the highest-skilled and highest-earning professionals in advanced economies, their incomes affect the allocation of top talent in the economy and are often cited as contributing to high and rising healthcare costs, especially in the United States. Regulatory policy (in part shaped by public sentiment) and market forces combine to affect physician incomes, and a better understanding of these forces is key to developing effective policies.
To study these forces, the authors analyze doctors’ income distributions in four developed countries, the United States, Canada, Sweden, and the Netherlands. They focus on physicians’ incomes relative to each country’s income distribution. To begin, they show that physician income is highly concentrated in each country’s top decile:
- In the United States (where 819,500 physicians average $458,100 in annual income), over 84% of physicians are in the top decile, with 26% in the top percentile alone.
- In the other three countries, roughly 60% of physicians are in the top decile. Sweden (55,097/$115,200) differs in having a lower share in the top percentile: only 5% of Swedish doctors fall into this bracket, compared with around 20% in Canada (100,660/$194,700) and the Netherlands (70,295/$185,700).
The authors’ other key findings are reflected in the accompanying figures. Figure 1 shows incomes relative to national income distributions. For example, the rightmost blue dot in Panel A means that 26 percent of US physicians have incomes at or above the 99th percentile of the U.S. income distribution. For the United States, Panel B, for example, reports both household-level AGI (adjusted gross income) and individual-level.
For both panels, the following patterns are illustrated:
- Across all four countries, physicians are a common high-income occupation. Few physicians are below the top two deciles of earners nationally.
- The United States has particularly high concentration of physicians in the top two percentiles. Physicians in Canada and the Netherlands are somewhat less concentrated in the top percentiles; their distributions show more mass from percentiles 70 to 95. The Swedish distribution appears different from the other three countries, as only about 5% of Swedish physicians are in the top percentile of the Swedish income distribution. More Swedish physicians fall in the 80th–94th percentiles than in the other countries.
- The US difference is most pronounced as regards physicians’ average income level rather than their relative position in the income distribution, as shown in Panel B. For example, US physicians in the top percentile earn nearly $1.2 million at the household level (AGI) and $1 million at the individual level (ACS), compared with $500,000 in Canada and under $400,000 in Sweden and the Netherlands.
The second figure reports the distribution of physicians’ location in the income distribution disaggregated by specialty.
- Panels B through D show that physicians’ concentration at the top of the income distribution is more pronounced for specialists than for physicians overall. For primary care physicians, in contrast, there is more mass in percentiles 60–94.
- In the United States, 42% of specialists are in the top percentile of the national income distribution. Combined with the 98th percentile, over half of American specialists are in the top 2%.
- Canadian and Dutch specialists are also prominent at the top, with over one-quarter of Canadian specialists and over one-fifth of Dutch specialists in the top percentile.
- While specialization does predict higher income for Swedish physicians, the absolute concentration at the top remains below the other countries, with only 7% of specialists in the top percentile.
The authors’ key finding that the overall income distribution within each country is closely related to physician incomes statistically explains the bulk of the US physician earnings premium, which may arise from physicians’ labor market alternatives and from high-income consumers’ willingness to pay for health. Institutional differences also matter, for example, as US physicians’ greater representation in the very top income percentiles may reflect greater business opportunities in the American healthcare system. In contrast, Swedish physicians’ lower presence near the very top of the income distribution likely reflects the predominance of salaried, public-sector healthcare employment, and perhaps the government’s power to reduce physician incomes.
What are the policy implications for the authors’ observations? Say, for example, that a policy aims to reduce US healthcare spending by lowering physician pay. If average individual physician incomes in the United States fell to the Swedish level, healthcare spending would fall by 6%, corresponding to $291 billion or 1% of GDP (from 17.6% of GDP to 16.6% of GDP). This modest effect on US healthcare spending would represent a $342,900, or 75%, reduction in average US physician incomes, which would likely have a substantial impact on workers’ employment choices.
However, if US physician incomes instead shifted to match Swedish doctors’ relative income distribution, this would imply a $230,400, or 50%, reduction in average US physician incomes. This would generate a correspondingly smaller reduction in spending: healthcare spending would fall by 3.8%, corresponding to $195 billion, or 0.7% of GDP.
Bottom line: Relative differences across specialties are largely consistent across countries, suggesting that similar economic fundamentals (e.g., different skill requirements or compensating differentials for job amenities), regulations (e.g., entry barriers and administrative price setting), or political forces are at play. Finally, although the broad patterns are similar, specialists commanding the very top incomes is more pronounced in the United States, suggesting that business opportunities or the higher return to skill interact with specialty fundamentals to create strong earning opportunities.