TLDR: Senator Warren’s Medicare-for-All plan would impose hidden rationing costs of $8.9 trillion in its first 10 years (equivalent in magnitude to 15.9% of the plan’s total cost over that period). This includes time lost waiting in line as well as the cost of premature deaths arising from restrictions on the availability of technology.
In part 1, I showed that the new federal taxes required by Senator Warren’s Medicare-for-All plan would impose a hidden cost of $3.5 trillion to $14.3 trillion in deadweight losses in its first ten years. This is the #1 reason her single-payer plan is a singularly bad idea.
In part 2, I showed that her single-payer health care plan would add anywhere from $4.6 to $7.7 trillion dollars in waste (inclusive of fraud and abuse) to an American health system that already wastes about $1 trillion a year, making waste the #2 reason this plan is a bad idea.
The #3 reason the version of Medicare-for-All conceived by Senator Warren is a bad idea is because of the inevitable rationing it will produce. In other well-known single-payer systems, this rationing takes several forms, including restrictions on the availability of treatments or, more commonly—especially in Canada—rationing by waiting.
Rationing Availability of Services
Rationing of services can occur in two ways. The first is through deliberate administrative decisions not to cover certain expensive medical technologies.
The second is through the inevitable shortages that arise when monopoly public payers impose price controls that underpay providers. Venezuela has become a classic illustration of this: “free” care there has resulted in an 85% shortage of medicines along with and a 90% deficit of other medical supplies used to treat severe conditions like cancer and hemophilia. Two years ago, USA Today reports that “more than 13,000 doctors — about 20% of medical personnel — have left the South American country in the past four years to find better opportunities elsewhere.” Other estimates put the figure at 15,000 doctors who have left shortages of drugs and equipment and poor pay. Consequently, the situation has gotten to a point where thousands are dying early as the medical system implodes there.
While we might think or hope Venezuela is a worst-case outlier, the reality is that we already see signs of this in the U.S. where physician underpayment and red tape associated with Medicaid has resulted in physician non-participation rates as high as 32% in states such as NJ. Because it pays more generously than Medicaid, Medicare is not quite as bad, but Medicare non-participation exceeds 20% in states such as Nevada and Virginia. And the Medicare actuary has been warning for years that under Obamacare, this situation is likely to get far, far worse after 2030 as the law’s payment restrictions ultimately force Medicare physician payment rates below those of Medicaid! Here’s a grim picture from the actuary’s latest report last April.
Source: Shatto, John D., and M Kent Clemens. “Projected Medicare Expenditures under an Illustrative Scenario with Alternative Payment Updates to Medicare Providers.” Baltimore, MD, April 22, 2019. www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/ReportsTrustFunds/Downloads/2019TRAlternativeScenario.pdf.
And there is no question that rationing is such an enduring feature of Britain’s National Health Service that the Nuffield Trust has written an entire report on the matter: Rationing in the NHS. As I explained recently, advisory committees to NICE (the National Institute for Health and Care Excellence) use a threshold for recommending treatments of between £20,000 and £30,000 per quality adjusted life year (QALY). Since 1 British pound currently equals $1.29 in American currency, that translates into roughly $26,000 to $39,000 per QALY.
This is not a hard and fast ceiling: as explained in the Nuffield Trust report: “NICE recommends when people should and should not receive treatment, but its general guidance (in contrast with its technology appraisals) is not binding. 15 Clinical Commissioning Groups (CCGs) are responsible for planning and commissioning of health care services for their local geographic area. So there ends up being differences across CCGs in terms of the circumstances under which caesarean births are covered, for example. That said, there are unquestionably instances in which medications are denied altogether on grounds they are too expensive.
Medicare currently spends approximately $88,000 a year on kidney dialysis for each patient who has end-stage renal disease. Without it, such patients would die. So Americans evidently are willing to pay $88,000 for one added year of life. But taking into account the average quality of life of kidney dialysis patients implies a cost/QALY of $185,000 .
This amount is substantially larger than the NICE cost-effectiveness threshold. In fairness, U.S. GDP per capita is 34% higher than in the UK, but even if we adjust the threshold to account for this (lifting it to between $35,000 and $52,000), the cost/QALY for kidney dialysis still would fall quite far beyond its upper limit. What is worrisome is that one of Senator Warren’s chief health policy advisor’s—Dr. Donald Berwick—is a well-known huge fan of NICE! Senator Sanders likewise is a enthusiast of the NHS. To my knowledge, Senator Warren has not weighed in on the merits of NICE but if she is going to rely on Dr. Berwick as an advisor, he probably should explain whether he favors following their stringent cost-effectiveness guidelines by denying coverage for kidney dialysis or whether instead he believes the NHS instead has made a terrible mistake in setting its threshold too low.
The rationing of technology under government-run health systems unavoidably plays out in terms of adverse health outcomes. Ho and Preston (2010) have shown that the unusually vigorous deployment of life-saving technologies by the US health care system at very old ages has culminated in better diagnosis and treatment of diseases of older people and greater longevity for older patients in the U.S. compared with other developed countries.
As one example, Philipson et al. (2012) have shown that after adjusting for age, gender and cancer site, the average American cancer patient survives 1.8 years longer than a counterpart living in the European Union (see CEA 2019, p. 215 for a chart showing how this survival advantage has grown over time).
Rationing by Waiting
By both eliminating prices for patients and reducing provider payment rates, Medicare-for-All would create a double whammy that inevitably would erode the quality of care. Mulligan and Tsui (2016) have shown that quality degradation is the typical result in markets where prices are prohibited from doing so. In health care, such quality degradation would likely manifest in the form of shorter appointment times, longer patient travel times, or longer waiting times to receive care.
Indeed, rationing by waiting is a persistent feature of both the NHS and Canada’s single-payer health system. According to the most recent annual report from the Fraser Institute “the median wait time for medically necessary treatment in Canada this year was 19.8 weeks.” This is more than double the 9.3 weeks Canadians waited in 1993, when the Fraser Institute began tracking wait times for medically necessary elective treatments. In short, this is a problem that is non-trivial in magnitude and which also appears to be systematically growing over time.
The latest report from the Council of Economic Advisers codifies how poorly single-payer countries such as Canada and Britain fare when it comes to waiting times:
· 61 percent of Americans waited less than 1 month after being advised that they needed surgery. The comparable figures for Canada and the United Kingdom, two countries frequently cited as models by M4A advocates, were 34.8 percent and 43.4 percent, respectively.
· Only two countries (Germany, at 71.2 percent; and Switzerland, at 73.2 percent) had a slightly higher percentage of patients able to see a specialist within 4 weeks of referral than the U.S. (69.9 percent), and neither of these countries has a single-payer system (Mossialos et al. 2017). The figure for Canada was 38.0 percent, and that for the U.K. was 48.6 percent.
· The reports includes extensive tables and graphs on pp. 216-218 further documenting the dismal record of single-payer countries when it comes to keeping patients waiting in line for care.
We also have solid evidence from Canada that under single-payer health systems, physicians deliberately reduce the supply of their services. That is, the waiting time problem is a combination of the excess demand that arises under “free” care (discussed in part 2) and a shrinkage of supply, which simply aggravates the situation . As summarized by Wharton School chaired professor Patricia Danzon a quarter century ago: “In Quebec, in the two years immediately after the introduction of universal health insurance, home visits dropped by 63 percent, telephone consultations fell by 41 percent, physician time spent per office visit declined by 16 percent, and office visits rose by 32 percent.” I cannot think of any reason to believe that American physicians would respond any differently were a single-payer system adopted here.
Single-payer enthusiasts claim that single-payer systems provide more physician care than in the U.S. In Japan, for example, the latest OECD data show that annual physician visits per capita in Japan (12.8), for example, is three times as high as in the U.S. (4.0). But as Prof. Danzon astutely points out, the average Japanese patient sees the doctor for a mere 5 minutes per visit, whereas the average American’s doctor visit is 15-20 minutes. So the total amount of MD face-time appears to be comparable in both countries, but in Japan it is inefficiently allocated across 3 times as many visits, each of which has a hidden time cost. A recent Harvard study shows that a typical American doctor appointment (average length 20 minutes) actually took 121 minutes inclusive of travel, waiting, paying, and completing paperwork.
Even if this figure is only half as large in Japan due to lower administrative costs (an assumption, not a proven fact), it implies that the average Japanese patient incurs 50 minutes of time costs per visit, so the “excess” 8.5 visits they receive each year translate into about 7 hours of wasted time every year. Multiply this times 250 million adults in the U.S. and the result would be 1.75 billion wasted hours. At the most recent U.S. average hourly wage of $26.52, this would translate into $46 billion a year of wasted time just for physician care alone.
But time spent waiting is not limited to physician services. In Canada, hospitals are paid a fixed budget per year giving the perverse incentive to fill their beds with low-cost “bed-blockers” (the mere fact this problem has a name illustrates its pervasiveness). Canada has the same supply of hospital beds per capita as the U.S. (2.7/1,000 population). But the average length of stay in Canada is 36% higher (7.5 vs. 5.5 days) in part because, as Prof. Danzon explains, “more hospital beds in Canada are occupied by elderly patients with average length-of-stay of over sixty days, despite waiting lists for acute care admissions.” The consequence is avoidable suffering as patients wait months rather than weeks for various types of surgery.
How Much Rationing Would There Be Under Medicare-for-All?
Social Losses Due to Restrictions on Medical Technology. The Philipson results show aggregate net social gains from America’s higher spending on cancer treatment relative to other European countries that averaged between $47 billion annually (measured one way) and $171 billion measured a different way (I have converted their 2010 dollar estimates into 2019 dollars [S3]). The authors report a 59% difference in spending per cancer patient in 1999 (the last year analyzed): $70,000 in U.S. vs. $44,000 in European Union. However, the U.S. a GDP per capita that is about 25% higher than any of the 10 countries included in the comparison group. Moreover, the Medicare Trustees have shown that the income-elasticity of technology is 1.4, implying that 58% of the cancer spending difference might be attributable to income alone and have nothing to do with rationing practices in other government-run health systems [S4].
Accordingly, I have assumed that if the U.S. were to adopt the rationing practices observed in other European countries (regrettably, Canada was not included in the comparison), it might lose 42% of the annual gains reported by Philipson, which when inflated at the medical inflation rate used by Medicare actuaries (3.4% annually), yields an aggregate loss that would range from $229 billion to $846 billion over the first 10 years. This represents the value of life years lost were U.S. survival times following cancer diagnosis to shrink to the levels observed in other European nations (implicitly, given how I did my calculations, it implies that survival times would shrink 58% less than estimated by the Philipson team (these losses are monetized using $150,000 per lost year of life, for reasons explained in Philipson).
I view this as an extremely conservative estimate of losses due to technology rationing. Cancer admittedly accounts for a lot a health spending, but there are plenty of other conditions that are expensive to treat and for which technology influences outcomes—heart disease being the best example—for which comparable social losses might well arise.
Patient Time Losses. Prof. Danzon’s study is a quarter-century old, but it is the best stab I’ve seen at trying to quantify the hidden burden of waiting in line in the Canadian health system.
On the physician side, she concluded that patient time costs under a single-payer health system likely amounted to anywhere from 10 to 110 percent of spending on physician services. This is a conservative calculation insofar as it assumes that care is rationed to its highest-valued uses. If this is incorrect, then there would be an additional hidden cost associated with failures or delays in treating the most serious illnesses. Government actuaries project that between 2020-2027, 19..9% of health spending will be for physician services. So when applied to the 10-year estimates of expected spending under Medicare-for-All, this implies there could be anywhere from $1.1 to $12.3 trillion in hidden time costs not included in the official estimates of the cost of the Warren health plan.
On the hospital side, she synthesized findings from several Canadian studies to conclude that the foregone health benefits of excessive waiting times likely amounted to at least 7 percent of hospital spending but possibly could be as high as 11 percent. In light of projected hospital expenditures of $18.4 trillion in the first 10 years of Senator Warren’s version of Medicare-for-All, this implies an additional hidden waiting time cost of between $1.2 and $2.0 trillion were we to bring Canadian-style health care to the U.S.
So the grand total hidden cost associated with the kinds of rationing typical in the Canadian-style single-payer approach proposed by Senator Warren would range from $2.6 to $15.2 trillion in its first decade.
Compiled by Christopher J. Conover, Duke University, using sources and methods described in text.
These are extremely conservative figures, as they only take a very rough stab at the adverse effects on health and longevity from the exclusion of certain high-cost medications that would be likely under single-payer health care.
Things are looking bleak: now we know that Senator Warren’s misguided single-payer plan would produce in its first 10 years:
- $8.7 ($3.5 to $14.3) trillion in deadweight losses due to higher federal taxes.
- An additional $6.1 ($4.6 to $7.7) trillion worth of waste in a U.S. health system that already has $1 trillion in waste every year.
- An additional $8.9 ($2.6 to $15.2) trillion in costs related to restrictions on the availability of technology and time losses related to rationing of physician and hospital services.
So we’re now up to additional hidden costs expected to be $23.7 trillion in its first but possibly as high as $47.2 trillion associated with Senator Warren’s single-payer plan (all this for a plan her advisors claim would cost “only” $52 trillion over that period.
I hope readers better understand why I believe this plan is misguided (and raises the question of what it would take to convince single-payer enthusiasts to abandon the idea). With that said, there’s two additional reason to oppose this plan that I will address in the final two installments.