Win McNamee / Getty

Senator Elizabeth Warren’s refusal to answer repeated questions at last night’s debate about how she would fund Medicare for All underscores the challenge she faces finding a politically acceptable means to meet the idea’s huge price tag—a challenge that only intensified today with the release of an eye-popping new study.

The Urban Institute, a center-left think tank highly respected among Democrats, is projecting that a plan similar to what Warren and Senator Bernie Sanders are pushing would require $34 trillion in additional federal spending over its first decade in operation. That’s more than the federal government’s total cost over the coming decade for Social Security, Medicare, and Medicaid combined, according to the most recent Congressional Budget Office projections.

In recent history, only during the height of World War II has the federal government tried to increase taxes, as a share of the economy, as fast as would be required to offset the cost of a single-payer plan, federal figures show. There are “no analogous peacetime tax increases,” says Leonard Burman, a public-administration professor at Syracuse University and a former top tax official in both the Bill Clinton administration and at the CBO. Raising that much more tax revenue “is plausible in the sense that it is theoretically possible,” Burman told me. “But the revolution that would come along with it would get in the way.”

At the debate, as throughout the campaign, Warren refused to provide any specifics about how she would fund a single-payer plan. Instead, whether questioned by moderators or challenged by other candidates, she recycled variants on the same talking points she has used in venues from campaign town halls to a recent appearance on The Late Show With Stephen Colbert. Rather than explaining what revenue she would raise to fund the plan, Warren insisted that under single payer, middle-income families would save more money with the elimination of health-care premiums, co-pays, and deductibles, regardless of any taxes imposed. “Costs will go up for the wealthy and for big corporations, and for hard-working middle-class families, costs will go down,” she said at the debate.

That calculation itself is disputed. And it begs the question: Even if families would eventually save under a single-payer system, a President Warren would still need to identify a politically plausible funding plan to pass such a program through Congress. By all indications, that looms as an extremely daunting project.

The new Urban Institute study helps define the magnitude of the task Warren (or Sanders) would face. The think tank modeled the costs of eight possible plans to expand health-care coverage that generally track ideas from the Democratic presidential candidates. By far, the most expensive was its version of the single-payer plan that Sanders introduced in the Senate and that Warren later endorsed: a blueprint that would eliminate private health insurance, require no co-pays or premiums from individuals, and provide everyone in the United States (including undocumented immigrants) an expansive benefits package including dental, vision, and home health care.

The 10-year cost of $34 trillion that the study forecasts nearly matches the CBO’s estimate of how much money the federal government will spend over that period not only on all entitlement programs, but also on all federal income support, such as the Supplemental Nutrition Assistance Program. Former Vice President Joe Biden said incorrectly at the debate that the single-payer plan would cost more annually than the total existing federal budget—it would cost less. (The CBO says Washington will spend about $4.6 trillion in 2020.) But over the next decade, the plan on its own would represent a nearly 60 percent increase in total expected federal spending, from national defense to interest on the national debt, according to CBO projections.

The Urban Institute estimates that a single-payer plan would require $32 trillion in new tax revenue over the coming decade. That’s slightly less revenue than its projected cost, because it would generate some offsetting savings by eliminating certain tax benefits the government now provides, such as the exclusion for employer-provided health care.

How big a lift is it to raise $32 trillion? It’s almost 50 percent more than the total revenue the CBO projects Washington will collect from the personal income tax over the next decade (about $23.3 trillion). It’s more than double the amount the CBO projects Washington will collect over the next decade from the payroll tax that funds Social Security and part of Medicare (about $15.4 trillion). A $32 trillion tax increase would represent just over two-thirds of the revenue the CBO projects the federal government will collect from all sources over the next decade (just over $46 trillion.)

Taxes that can fill that big of a hole are not easy to identify. Even by Warren’s own estimates, which some liberal economists consider too optimistic, her proposed wealth tax on personal fortunes exceeding $50 million would raise just $2.75 trillion over the next decade. That’s less than what would be required to fund a single-payer plan for one year. In any case, Warren has already targeted that revenue for other proposals she’s issued, such as providing universal preschool and child care, and canceling most college debt while funding tuition-free public higher education. Repealing the tax cuts for businesses and individuals that President Donald Trump and the GOP Congress passed in late 2017 would similarly raise about $2 trillion in federal revenue over the next decade.

Other tax options would likewise make a relatively minor dent. For instance, some Democrats have proposed for years to eliminate the current cap on the payroll tax—which stops taxing income above about $133,000—and instead impose the tax on all income above a higher threshold, such as $250,000. The CBO recently estimated that such a plan would raise about $1.2 trillion over the next decade, again a small share of single payer’s cost. Besides, Warren has already proposed such a tax hike and earmarked the money to increase Social Security benefits.

Burman told me that the broad-based income-tax increases that Sanders has discussed using to fund single payer—including raising the top income-tax rate past 50 percent and ending reduced taxation for capital gains—would likely cover about half of the proposal’s cost. If Warren or Sanders tried to cover the other half with a value-added tax—a sort of national sales tax that many European nations use to fund their social safety net—the rate would likely need to be set at about 25 percent, he estimated. “All of the things Sanders proposed plus a high VAT by European standards might get you there,” Burman said.

Alternatively, some advocates have discussed raising the payroll tax to fund a single-payer plan. Currently, the payroll tax is set at 15.3 percent of earnings, with the cost split between employees and employers. Former CBO Director Douglas Holtz-Eakin, now the president of the center-right think tank the American Action Forum, told me that level would need to substantially rise to fund a single-payer plan. He said, in a “ballpark” estimate, that Sanders’s plan “would require [a] payroll-tax hike of 20 to 25 percentage points.”

Whatever alternative Warren or Sanders select, a single-payer plan would require increasing federal revenue at a rate not seen in 70-odd years, both Burman and Holtz-Eakin told me. Measured as a share of the economy, total federal receipts tripled during World War II, rising from almost 7 percent to nearly 21 percent of the gross domestic product from 1940 to 1945, according to federal figures. Since then, federal revenue, compared with the broader economy, has generally oscillated within a fairly narrow range. The most it’s increased in a single decade is about 10 percent, during the 1950s, 1960s, 1990s, and in the past decade. (Revenue has increased despite the massive Trump tax cuts because the Great Recession vastly reduced federal revenues and created an unusually low starting point.)

Though federal revenue today still starts at a low level by historic standards (16.6 percent), providing more potential flexibility to raise taxes, the cost of a single-payer plan would swamp any such advantage. By 2029, with the added cost of single payer factored in, federal revenue would increase to close to 30 percent of the total economy. That would mean federal revenue would increase as a share of the economy by about three-fourths over a decade, far more rapidly than in any other 10-year period since World War II. It would also mean that federal revenue would considerably exceed the share of the economy it consumed even during World War II, when it reached 20.5 percent in 1944, a level unmatched since.

The central response from Warren and Sanders to concerns about their health plans’ cost has been to tout the overall savings for Americans, and the Urban Institute analysis suggests that for lower- and middle-income families, that’s possible. The study projects that households will save nearly $887 billion in annual costs and employers another $955 billion, some of which could revert to workers in the form of higher wages.

Though that’s much less than the new taxes required for the plan, the organization says that lower- and middle-income families might still come out ahead. “Higher-income people will likely face the greatest increases in taxes, meaning their new tax burdens would likely exceed their savings; the reverse is likely true for lower-income populations,” the report concludes. (Among other dissenters, that conclusion is disputed by Kenneth Thorpe, a leading health economist at Emory University and a former assistant secretary in Clinton’s Health and Human Services department. In a study Biden cited at last night’s debate, Thorpe calculated that, under single payer, 71 percent of households with private insurance today—almost 70 million households in all—“would pay more in new taxes than they would save through the elimination of premiums and cost sharing.”)

The debate over savings is impossible to resolve so long as Warren refuses to offer any indication of what revenue she would raise to fund her plan. On CNN this morning, South Bend, Indiana, Mayor Pete Buttigieg lashed her for dodging the question. “I have a lot of respect for Senator Warren, but last night she was more specific and forthcoming about the number of selfies she’s taken than about how this plan is going to be funded,” he said.

What is clear now is that the Sanders version of single payer—which Warren at the debate called the “gold standard” of health-care proposals—would cost far more than any other alternative. The new analysis found that plans similar to the one Biden, Buttigieg, and other candidates have proposed—centered on expanding a public option to compete with private insurance companies—would achieve nearly universal coverage at a cost of roughly $122 billion to $162 billion annually, depending on exactly how they are designed. Even what the analysts called a single-payer plan “lite”—requiring some co-pays and offering somewhat less generous benefits, without covering undocumented immigrants—would cost about $1.5 trillion annually, about half as much as the Sanders and Warren proposal.

Such comparisons are certain to compound the anxieties that many Democratic health-care experts feel about trying to defend in the general election a single-payer plan that would eliminate private health insurance and require such a large increase in federal spending.

“Many countries do not wrest the entire burden of every single person’s health care into the federal government,” says Neera Tanden, the president of the liberal think tank the Center for American Progress and a former health-policy adviser to Barack Obama and Hillary Clinton. “I think there are big questions about the United States moving from the most conservative health-care system to the most leftward government-run health-care system.

“There are positives and negatives to any of these options,” Tanden adds. “But one issue in a country that has more anxiety about the government’s role in people’s lives is whether it is feasible, or even sustainable over the long term, to have the federal government [grow so much] in size because the entire system of health care would be run through the government.”

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.