A question that comes up in discussions of healthcare reform—and the transition to single-payer—is the potential savings via reductions in compensation for physicians’ labor.  Some assert that the transition to single-payer would, or should, involve historic reductions in pay.  To have an informed discussion, however, we have to first know how much we currently spend on physician salaries.

There are a few different ways to estimate this.  According to CMS’s National Health Expenditures (NHE) accounts, spending on “physician and clinical services” was $664.9 billion in 2016, or about 20% of our health spending.  20% is also the figure that health economist Uwe Reinhardt used a while back, presumably using CMS’ figure.   According to CMS, that figure includes spending on physician offices and outpatient care centers, but also on independent medical and diagnostic laboratories, which muddies the waters.  At the same time, it excludes salaried doctors at hospitals or nursing homes.  But importantly, according to Reinhardt, it reflects expenditures on offices and clinics (including practice overhead), not take-home income.  Reinhardt states that physician income is about half of this spending, which would be 10% of NHE.  For 2016, that would be roughly $330 billion.

Clearly there are several assumptions in that estimate that might make it either too high or too low.  An alternative approach is to multiply the number of practicing doctors by their pay.  According to the Bureau of Labor Statistics, there were 713,800 physician jobs in 2016, while median pay was “equal to or greater than $208,000 per year.”  Hence, this pay figure is presumably an underestimate.  Additionally, mean pay could be higher than median pay depending on the distribution of salaries.  Putting that aside for a moment, multiplying 713,800 by $208,000 gives us $148 billion in physician salaries, or roughly around 5% of 2016 NHE, half that of Reinhardt’s figure.  For the reasons stated, however, we should consider this an underestimate.

Dean Baker, in a piece in Politico, takes this same approach, using somewhat higher numbers.  He notes that there are more than 900,000 doctors, linking to a figure that puts the number of doctors at 1,045,910 in 2013.  However, that sources also notes that there were 148,000 inactive physicians, which would put the number of active physicians slightly shy of 900,000.  Although higher than the BLS figure, that’s similar to a figure from the Kaiser Family Foundation, which reports that there were 968,743 “professionally active” physicians in 2018.  Baker assumes that doctors are paid around $250,000 a year, on average, based on a figure from Medscape’s 2016 Physician Compensation Report.  So, if there are 900,000 doctors paid at $250,000 a year, we have a figure of $225 billion, or a bit less than 7% of 2016 NHE.  That figure is about halfway between the NHE/Reinhardt estimate and the estimate using BLS figures.  (Note: This is based on Baker’s figures, which for income is from 2016.  However, if we use the 2018 KFF figure, 2018 Medscape salary figure, and 2018 NHE, the story doesn’t change much – salaries come to a bit less than 8% of NHE).  

In sum, then, it would appear that physician salaries account for somewhere between 5 – 10 % of healthcare spending.  Let’s assume it is Reinhardt’s 10%, so as to err on the side of greater potential savings via reduced pay.  As he notes, that would mean that if we were to dramatically slash physician pay in a historically unprecedented fashion — say by 20% — US healthcare spending would fall by a grand total of 2%.  Reinhardt argued that in return for that 2% reduction in healthcare spending we would have a “wholly demoralized medical profession,” and so considered this approach a “poor strategy.”

Others may disagree.  Yet regardless of one’s opinion on what physician pay should be, these numbers demonstrate that it is by no means necessary to reduce physician reimbursement, much less slash it, to make single-payer affordable.  After all, even a historic reduction in physician salaries would produce only a 2% reduction in health spending.  Such a pay cut would presumably coincide with a modest increase in utilization under single-payer, which would mean more work for less pay.  It’s hard to see how this wouldn’t alienate the medical profession and turn them against single-payer, possibly dooming its prospects.

It is probably for this reason that major transitions to public universal healthcare systems in history—i.e. US Medicare, the Canadian universal healthcare system, and the United Kingdom’s National Health Service—did not reduce compensation for healthcare workers’ labor.¹  We should consider the wisdom of that approach today as we get into the nitty gritty of designing the American single-payer system.^2,3,4

1. As an article in the Independent notes, Nye Bevan “was pragmatic enough to know that he could not run the NHS without consultants, doctors and nurses.”  He made concessions to the doctors and increased the nurses’ pay.

2. As a side-note, reducing economic inequality absolutely must be a priority. To accomplish that, however, I would argue that we should pursue progressive taxes—including on unearned income from wealth—as opposed to specifically targeting the wages of well-paid healthcare workers, particularly given that pursuing the latter could potentially scuttle the passage of single-payer healthcare, a reform that itself would do a great deal to reduce inequality.

3. Keep in mind I am referring to take home reimbursement, not overall reimbursement. By reducing provider-side administrative costs, which a single-payer system will accomplish, physician fees can fall somewhat without reducing payment for labor.  After all, some fee reduction will be the way that provider-side administrative savings are realized.  Matt Bruenig argues that the provider rate cuts assumed in the Mercatus Center single-payer report could probably be wholly accounted for by administrative savings.  I’m also not referring to for-profit physician-owned hospitals or other for-profit healthcare delivery businesses, which should not be allowed in the system—I’m speaking about compensation for labor.  Finally, there should certainly be attenuation in pay disparities, but basically everybody already agrees on that.  After all, simply moving to a single rates system will mean equalizing pay for providers who care for richer and poorer patient populations.

4. Obviously, reimbursement will not (and practically cannot) stay exactly the same—I’m speaking about avoiding large changes. Making healthcare professionals education tuition-free can help ease the transition for doctors who do less well.