Bernie Sanders's health care plan is underfunded by almost $1.1 trillion a year, a new analysis by Emory University health care expert Kenneth Thorpe finds.
Thorpe isn't some right-wing critic skeptical of all single-payer proposals. Indeed, in 2006 he laid out a single-payer proposal for Vermont after being hired by the legislature, and was retained by progressive Vermont lawmakers again in 2014 as the state seriously considered single-payer, authoring a memo laying out alternative ways to expand coverage. A 2005 report he wrote estimated that a single-payer system would save $1.1 trillion in health spending from 2006 to 2015.
But he nonetheless concludes that single-payer at a national level would be significantly more expensive than the Sanders campaign believes, and would require workers to pay an additional 20 percent of their compensation in taxes. He also argues it would leave 71 percent of households with private insurance worse off once you take both tax increases and reduced health care expenditures into account.
Sanders's camp is, naturally, skeptical. Sanders's policy director Warren Gunnels told me Thorpe's analysis is a "total hatchet job." The disagreement ultimately comes down to a question of how optimistic you are about single-payer's ability to reduce health care spending.
Sanders's plan is very optimistic, assuming huge reductions in per-person health care spending that bring the US much closer to existing countries with single-payer like Canada (which spends nearly 48 percent less per person) or Australia (more than 56 percent less). "If you look at every other country that has adopted a universal single-payer health care system, their costs per capita are far lower than they are in the United States," Gunnels told me.
Thorpe is less rosy. He assumes the US can reduce the cost of prescription drugs by a fifth, that it can pay providers much less than private insurers currently do, that it can significantly slow down how fast health care spending grows, and that it can gain some substantial savings from simpler administration. But that still isn't enough to make the plan affordable without a massive tax increase.
You can read Thorpe's full analysis below. Keep reading after to see why Thorpe got such different results from Sanders's team, and the arguments each makes for their view.
A $1.1 trillion shortfall
Thorpe finds that Sanders's plan would cost an average of $2.47 trillion per year from 2017 to 2026, on top of existing federal spending on health care. Sanders's campaign's analysis, by contrast, estimated an average cost of $1.377 trillion.
In a phone call, Gunnels explained the $1.1 trillion gap. It comes down to five factors:
- Sanders assumes $438 billion more per year in administrative savings than Thorpe; Thorpe assumes that total health spending will fall by 4.7 percent because single-payer is simpler to administer, while the campaign has anticipated a reduction of 16 percent (changed in a later email to 13 percent).
- Sanders assumes $324 billion more per year in prescription drug savings than Thorpe does. Thorpe argues that this is wildly implausible. "In 2014 private health plans paid a TOTAL of $132 billion on prescription drugs and nationally we spent $305 billion," he writes in an email. "With their savings drug spending nationally would be negative." (Emphasis mine.) The Sanders camp revised the number down to $241 billion when I pointed this out.
- Sanders assumes $216 billion more per year in savings because Thorpe thinks eliminating copayments and deductibles will lead to people using a lot more health care (10 percent more, to be exact), and Sanders's camp is more skeptical (they assume 6 percent more).
- Sanders assumes $160 billion per year in savings relative to Thorpe because, they argue, he includes elective procedures like plastic surgery, which single-payer wouldn't cover. Thorpe disputes this: "Cosmetic surgery, really? That's $12 billion a year and in the second decimal of rounding." In other words: There's no way excluding plastic surgery can give you $160 billion of savings.
- Sanders assumes that states will pay $100 billion more per year in Medicaid and SCHIP spending than Thorpe does, because they think states will keep paying in the exact same amount they currently pay into those programs. Thorpe is skeptical, noting that in the Supreme Court's 2012 Obamacare ruling, it "said in essence you cannot force states to make spending on a expansion of Medicaid —how is the world can you expect states to contribute toward the costs of programs that are eliminated?"
Which is all to say that the gap between Thorpe and Sanders is the gap between an economist who is optimistic that single-payer can save some money and a campaign optimistic that it can save a huge amount of money. "Their savings numbers are — well, politely said — simply wrong," Thorpe writes in an email.
They're also apparently malleable. When I pointed out that the yearly savings numbers they were presenting on prescription drugs were literally impossible, the Sanders camp revised the number to $241 billion — huge and arguably implausible but not larger than total annual spending on prescription drugs. A follow-up email also revised down the assumed administrative savings from 16 percent to 13 percent and the savings on utilization up from $216 billion to a whopping $660 billion.
"They are just throwing things against the wall to see what sticks," Thorpe says.
A 20 percent tax hike?
Thorpe's more skeptical estimates in many cases derive from the attempt at implementing single-payer in Sanders's home state of Vermont. That effort assumed a 4.7 percent reduction in administrative costs — even before taking into account that the transition effort itself would cost a huge amount. That's the number Thorpe borrowed, and it's a big part of why he finds the plan to be more expensive than the Sanders campaign does. The Sanders camp argues this is too small because Vermont would still maintain Medicare and other federal health programs, adding complexity and reducing administrative savings.
Vermont also estimated that single-payer would require a 11.5 percent payroll tax on all businesses and a progressive income-based premium ranging from 0 percent to 9.5 percent, with the top rate kicking in for those at four times the poverty line ($102,000 for a family of four in 2017).
Thorpe, similarly, estimates that you'd need a 14.3 percent payroll tax on employers for a national single-payer plan, and a 5.7 percent income-based premium, for a combined 20 percent tax — about what Vermont estimated.
That's much higher than Sanders's campaign is suggesting; they want a 6.2 percent payroll tax and a 2.2 percent income-based premium, along with a large number of other tax increases on the wealthy.
Gunnels disputes the 20 percent tax increase figure in fiery terms. "That is absolutely absurd, it's absurd, it's outrageous," he said in a phone call. "It's coming from a gentleman that worked for Blue Cross Blue Shield. It's exactly what you would expect somebody who worked for BCBS to come up with. It's not even worthy of any type of serious reporting, because it would not happen."
Side note: It's true, Thorpe did once do a consulting assignment for BCBS, conducting "a study a decade or so ago looking at the racial and income characteristics of who enrolls in Medicare + Choice [now called Medicare Advantage] plans," per his recollection. But he's an academic who's broadly respected across the spectrum, and who's been sympathetic to single-payer in the past. The Sanders campaign's characterization here seems unfair, even ad hominem.
There's also a dispute as to who would pay the employer payroll tax. Like most economists, Thorpe treats the employer payroll tax as paid entirely by workers. In an email, he explained that he's following standard practice used by the Congressional Budget Office and the Joint Committee on Taxation in calculating their estimates of legislation. "CBO and JCT assume that total employer compensation is fixed so this is fully borne by workers," he writes.
But Gunnels argues that the cost is not passed along to employees, because employers are also saving billions in premium costs. He especially argues that poor employees wouldn't pay more, because, "Many of these people were making minimum wage, so employers can't really do anything to cut their salaries further. They're already making as little as possible." Of course, these businesses could simply fire minimum wage workers instead.
As to the larger point, economists generally think of employer health benefits as part of employees' total compensation, so if the payroll tax digs into that pot of money, as Gunnels says, employees still ultimately pay the bill. That said, in fairness, that may not mean that their cash wages fall, just that they won't get a big raise to make up for employers reducing their compensation by no longer buying them health insurance.
Thorpe: Most households would be left worse off
"There would be substantial distributional impacts (large number of households and businesses that pay substantially more and less) of any plan that has to raise a total of 20 percent of total compensation relative to current law," Thorpe writes.
He's not kidding. Thorpe estimates, taking into account taxes he thinks that plan needs to finance itself, that many groups would pay more:
- 71 percent of total working households with private insurance would pay more.
- 57 percent of households of workers in businesses with fewer than 50 employees would pay more.
- 65 percent of working young adult (18 to 26) households would pay more.
- 85 percent of working households on Medicaid would pay more.
- 66 percent of working households on Medicare (a minority of Medicare recipients, most of whom are retired) would pay more.
That is to say, those shares of those groups would pay more when you take into account both the taxes and the fact that they wouldn't have to pay for health expenditures like premiums, deductibles, copays, etc. anymore. In each case, though, a substantial minority, including many with large health bills and/or hugely private expensive health care plans, comes out ahead.
The Sanders campaign disputes these numbers, as they're based on the tax numbers Thorpe derived, which Sanders's staff thinks are much too high. So Thorpe also analyzes the impact assuming the 6.2 percent and 2.2 percent taxes included in Sanders's official plan, noting that they would leave the program underfunded.
He concludes, just as UMass Amherst economist Gerald Friedman did in his analysis of Sanders's plan, that most households would benefit. Only 27 percent of total working households with private insurance would pay more according to Thorpe, with 72 percent paying less; young adult workers, workers in small businesses, and workers on Medicare would gain too.
The big exception is Medicaid. Thorpe finds that 72 percent of Medicaid workers would pay more under single-payer even under Sanders's lower tax rates. Medicaid currently has very limited cost sharing for families in poverty, mostly limited to prescription drug copays, and so single-payer would offer such families little in the way of health savings while making them pay an additional 6.2 percent in payroll taxes, even if most don't have taxable income that'd be hit by the 2.2 percent income-based premium.
VIDEO: What is single-payer healthcare?
Be sure to subscribe to Vox on YouTube for more explainer videos