Despite seven years of demagoguery about the law’s effects, Republicans were unable to kill the Affordable Care Act. That’s given reformers on the left renewed confidence. The question of how best to keep the ACA’s unfulfilled promises will be a central issue in the 2020 election, and if Democrats regain power, the fight to further expand coverage and reduce health-care costs will likely be front and center.

Senator Bernie Sanders and Representative John Conyers have captured the energy of the progressive movement with similar proposals to make Medicare benefits significantly more generous than they are now—covering almost every medical expense, while eliminating co-pays and almost all other out-of-pocket costs—and then move almost the entire population into this more expansive program. Conyers would achieve this in one fell swoop, in under two years; Sanders would phase it in over four years.

But, as I wrote in August, these bills represent the most disruptive and difficult possible route to a comprehensive national health-care system. They’re too easy to misrepresent. Like any other deep reforms, they’d face concerted opposition from the right, centrist Democrats and health-care providers. But they’d face another obstacle, too: Because they compel so many people to give up their existing coverage for the promise of something better—all within a short period of time—they’re almost guaranteed to spark a popular backlash. Without strong popular support, these approaches are doomed to fail.

At the same time, Democratic lawmakers, think-tankers, and academics have offered a menu of incremental reforms that range from creating a public insurance program that would compete with private insurers in Obamacare’s exchanges to allowing people aged 55–64 to buy into Medicare to giving states the option of allowing anyone who lacks insurance to purchase coverage from their Medicaid programs. All of these ideas would bring about progress. And because they avoid a full-frontal assault on industry stakeholders, and leave the familiar employer-based insurance system in place, they’re more feasible politically. But they wouldn’t get us to universal coverage, and aren’t designed to aggressively control costs. More importantly, they’re unlikely to capture the imagination of young voters or energize the activist left. So the very practicality of these approaches is a major liability.

What we lack, and what we need, is a pathway to something like Medicare for All—or what Yale political scientist Jacob Hacker calls “Medicare for Most”—a scheme that establishes comprehensive health care as a right and is aggressive in challenging a status quo that continues to fail millions of Americans—but which would be merely very difficult to pass, rather than all but impossible.

So consider an alternative approach, one that combines the audacity of the Sanders and Conyers plans with an understanding of the difficulties involved in restructuring a sector of the economy that accounts for one out of every 11 American jobs.

We could roll Medicaid’s low-income coverage, the Child Health Insurance Program (CHIP), and other public spending into the Medicare system, creating a very large pool of people insured through that program. We can empower it to negotiate with providers, and then expand the program’s coverage, as a benefit, to older working people, children, and the poor. And we can then open up this new Medicare for the rest of the population—individuals and businesses—to buy into, voluntarily and at their own pace.

If those of us who extol the benefits of universal health care are right, then we can expect this system to be more efficient and better at controlling costs than the fractured private insurance market, and expect most companies to eventually take what would be a better deal. Over time, we’d get to something approaching an “all-payer” system, with a single rate-setter, which I’ll discuss in more detail below.

This would get us to universal coverage in a way that’s almost as ambitious as the Medicare for All proposals that are currently on the table, but the transition would be more gradual and less disruptive. More importantly, it would pit reformers against the health-care industry, rather than the public.

Advocates of the Sanders/Conyers approach to Medicare for All are caught in a rationality trap. They believe single payer’s simplicity and efficiency—and the lure of eliminating co-pays and out-of-pocket costs—will spur massive public support for a bill like Sanders’s, and that the outpouring of popular power will overcome the inevitable resistance from the health-care industry and its allies in Congress.

It’s true that several polls have found majority support for their approach, but that support is notably soft. Advocates cite a Kaiser Family Foundation survey from July that found that 53 percent of respondents supported an approach “in which all Americans would get their insurance from a single government plan.” (When asked specifically about “Medicare for All,” support rose to 57 percent, which is probably a reflection of Medicare’s popularity.) But the study also found that “the public’s attitudes on single-payer are quite malleable, and some people could be convinced to change their position after hearing typical pro and con arguments that might come up in a national debate.” When those who supported the plan were asked how they would feel if they heard that it gave the government too much control over the health-care system, 40 percent said they “would change their mind and would now oppose the plan.” There were similar shifts when people were asked how they’d respond to arguments about raising taxes to pay for it or were told that Medicare for All would eliminate Obamacare.

The central problem is that both Sanders and Conyers would compel almost six out of 10 people in the non-elderly population that get insurance through their employers (or their spouse’s employers) to move into a new program in a short period of time. This runs headlong into our well-established bias toward maintaining the status quo—we tend to fear change—and our tendency to loathe the idea of losing something that we already have, even if we might get something better in exchange, a concept known as “loss aversion.” (I discussed these problems in greater detail in August.) And that organic resistance to change will inevitably be amplified by the same dishonest messengers who turned the free-market-friendly Affordable Care Act (ACA) into a government “takeover” of the health care system.

In a perfect world, breaking the connection between employment and health insurance would be a laudable goal. Why should a business determine what kind of insurance its employees have access to? But that’s not the world we live in. The United States developed its employer-based system during World War II, when companies competing for workers in a booming wartime economy couldn’t offer higher pay because of government-imposed wage controls. So they turned to offering benefits like subsidized health insurance to attract those scarce workers. In 1943, the IRS ruled that employer-based insurance would be tax-free, and in the 1950s Congress codified the rule, creating a permanent exemption for health-care benefits. According to Miami University economic historian Melissa Thomasson, just 9 percent of the US population had voluntary, private insurance in 1940; by the 1960s, that number had grown to 70 percent. At that point, most Americans saw employer-based insurance as the natural order of things.

And while most other countries established their national systems back when health-care costs were only pennies on today’s dollar, and then used those systems to keep costs down over the intervening years, we face the unprecedented challenge of reforming a health-care sector that’s ballooned to 17 percent of our economy.

Our health-care system is anything but rational, but we live in an irrational world, and that’s the one in which we have to fight for universal coverage. The priority should be establishing that comprehensive health care is a human right guaranteed by the government—regardless of what kind of structure we use to do it.

Another irrational truth is that while the United States can certainly “afford” to pay for a single-payer system—we already spend far more per person on health care than any other country—a dollar in new taxes is, from a political standpoint, very different from a dollar that’s already being deducted from one’s paycheck for an employer-sponsored plan. “The basic problem is that the move to single payer involves a massive shift of resources,” wrote economist Dean Baker, himself a single-payer advocate, in Democracy last month. Baker worked with single-payer advocates in Congress to figure out how to transfer the 10 percent or so of the country’s economic output that’s currently spent on private-sector health care to the public sphere. In theory, it could work: If you’re spending a dollar more on taxes and a dollar less on private insurance premiums, you’re coming out even. But it might not feel that way. Employers pay part of that dollar now, and workers who end up paying higher taxes would have to trust that they’d get their share back in wages. It’s unlikely they would. Baker concluded that, “while in principle it should be a matter of indifference to people whether they pay money for health insurance to providers in the form of premiums and co-payments or to the government in the form of taxes, members of Congress who have to run for reelection don’t seem to think this is the case.”

Neither Bernie Sanders nor John Conyers propose specific funding mechanisms in their plans, enabling their fellow Democrats to sign on to the principle of universal coverage without facing a barrage of negative ads accusing them of favoring massive tax hikes. It’s savvy politics at this point. But when the time comes to get specific, and the opposition starts talking about increasing federal taxes by almost $1.5 trillion, it’s a safe bet that these ideas will face significant opposition, even if those new taxes are offset by reductions in private premiums.

It’s important to understand that the United States already finances half of its health care through the public sector, but we do so irrationally and inefficiently, with a fractured system that prohibits the government from using its market power to control prices and a huge discrepancy in resources from state to state. This is a serious problem on several levels. It undermines our public-health system’s buying power. And it enables states to sharply limit benefits.

Nineteen states have refused tons of federal dollars from the ACA to expand Medicaid by raising the cutoff for eligibility to 138 percent of the federal poverty line. (Under the law, the federal government picks up 95 percent of the costs now, and then 90 percent beginning in 2020.) This significantly undermined the Affordable Care Act’s promise of universality.

Under non-expanded Medicaid, states set their own eligibility criteria, and many of the same ones that refused the expansion have also made those criteria so onerous that the program covers only a fraction of low-income people. In 18 of the 19 states that refused the expansion, people without kids are ineligible for Medicaid regardless of how little they earn. For people in those states who do have children, the median cut-off for eligibility is just 44 percent of the federal poverty line; in Alabama and Texas, parents are ineligible for Medicaid if their household income exceeds just 18 percent of the federal poverty line.

At a time when there’s a lot of cynicism toward the government, and political disinformation is everywhere, Sanders and Conyers are asking people to take a huge leap of faith based only on single-payer’s promise.

But here we can take another lesson from the ACA. Whatever its shortcomings, expanding coverage to over 20 million people dramatically shifted the discourse around health care in the United States. It helped bring about majority support for the idea that the government should guarantee coverage, and it made the loss of insurance under the GOP’s various plans to repeal-and-replace politically toxic.

Health-care reformers could build support for single payer in a similar way—demonstrating its benefits by creating a model on the state level. But that comes with its own challenges. The most likely states to become models for single payer haven’t been able to deliver so far. Despite a high-profile campaign earlier this year, California lawmakers tabled a single-payer bill in June. Some activists blamed their fecklessness and fealty to big donors, but legislators said practicalities drove the decision; an arcane spending formula embedded in the California Constitution meant that the legislature couldn’t enact a single-payer system without simultaneously tripling education spending.

That formula doesn’t apply to measures passed by popular referendum, but if advocates put single payer on the ballot, it would be relatively easy for corporate stakeholders to flood the airwaves with misleading ads and kill the measure. That’s precisely what happened in California in 1995, when opponents “waged a well-funded, sophisticated media campaign that the proposition’s proponents could not counter,” according to the Kaiser Family Foundation. The referendum was defeated by a 73-27 margin. It happened again in Oregon in 2002, when a similar proposal was defeated 78-22, and then once more in Colorado last year, when a single-payer proposal went down 80-20. In a small, liberal state like Vermont, which passed a single-payer law in 2011, the political will was there but the state government couldn’t find a way to raise the revenues necessary to implement it.

There’s another way—one that’s national in scope but avoids the popular opposition that forcing a huge share of the population to give up their current insurance would elicit. It would also require significantly fewer tax dollars to work.

This missing path starts with the understanding that all of the goals of progressive health reform—universality, greater equity, an expansion of public health care, cost containment through efficiencies of scale and reducing the role of profit in health care—can be accomplished to a significant degree without dismantling the current system entirely.

First, we should nationalize Medicaid’s low-income coverage and CHIP, and fold them into Medicare. Red-state governments have proven that they can’t administer these programs properly. Nationalizing them would bring an end to miserly cut-offs for eligibility. And we should cover everyone earning up to 150 percent of the poverty line—the more generous cutoff that was proposed in early iterations of the ACA.

Automatically enrolling children in the program at birth and covering them until they reach adulthood would eventually lead to a cultural shift, with people from all walks of life expecting to have insurance coverage regardless of their family’s economic situation. Then, with our existing public-health dollars folded into a single large pool, we should empower the government to negotiate prices with providers and allow drugs to be imported from other countries with strong safety standards. And we should expand public health care further by lowering the eligibility age for Medicare—as a benefit subsidized by the federal government rather than as an option to buy into—to the age of, say, 55.

We would also eliminate the ACA’s exchanges and offer those who aren’t covered under a group insurance plan the option of buying into Medicare for a premium that reflects the cost of covering them adjusted for their current Obamacare subsidies. This would not only lead to a significant expansion of public health care and a dramatic reduction in the rate of uninsured, but it would also bring almost all public spending on insurance under one single umbrella.

If we stopped there and declared mission accomplished, we’d have a near-universal system. But it still wouldn’t approach the efficiencies of a single-payer scheme. Most of the working-age population would remain covered by a fractured web of private insurers, and would continue to pay for their limited bargaining power, higher administrative expenses and profits.

So we shouldn’t stop there. We should then allow employers to buy into the Medicare program—first small employers and ultimately every employer. If single-payer advocates are correct about its benefits—if it reduces administrative overhead and contains costs more effectively than private insurance—then, over time, most employers will buy into the program voluntarily because it would make good economic sense to do so.

That would push us not toward single payer but an “all-payer” scheme with a single rate-setter. In an all-payer system, the prices that health-care providers can charge are negotiated with the same massive market power that single-payer systems can bring to bear, but multiple payers actually foot the bills. Those rates can be set, or negotiated, by groups of insurers and providers, as in Germany—a market-based all-payer system—or by a governmental authority, as is the case with Maryland’s all-payer system for hospital costs.

In Maryland, private insurers and government programs both pay hospitals rates set by an independent board of health-care experts. No more than three of the seven commissioners can be affiliated with health-care providers, which prevents the commission from being captured by the industry.

Established in 1971, the system has had a measurable impact. According to a 2009 study published by Health Affairs, the cost of a Maryland hospital admission went from 26 percent above the national average in 1976 to 2 percent below the national average in 2007. It’s also resulted in greater access to care and more equitable care, because even those on Medicaid can visit the state’s top facilities and its academic medical centers. It makes no difference to the government because they charge the same amount.

Austin Frakt, a health-care economist at the Department of Veterans Affairs and an associate professor at Harvard and Boston University, says, “This is a way to have multiple payers, and have the bargaining power of a single-payer.”

He points out that Medicare is currently a de facto all-payer system. While around one-third of those covered by the program have a private Medicare Advantage plan, a regulation sharply limits the incentive that private insurers have to pay providers more than traditional Medicare does, and no incentive for providers to accept less for the same services. With some slight variations, public and private Medicare plans pay the same for similar services.

By phasing out an exemption in the ACA that allows smaller companies to not cover their employees, over time this new, expanded Medicare program would effectively become a single-rate-setter in a similar way. With an employer buy-in, paired with an expansion of public health care to cover the uninsured, we’d capture many of the benefits of a single-payer system without dismantling the employer-based system that, for better or worse, most Americans have grown accustomed to for the past 70 years.

Rather than figuring out how to transfer close to 10 percent of our economic output from the private sector to the government in a relatively short period of time, we’d have to figure out how to finance only the fraction of those costs required to cover children and people aged 55–64.

The proposals offered by Bernie Sanders and John Conyers both contain a political hand grenade that’s just waiting to explode: In the name of improving Medicare, they would do away with Medicare Advantage plans—the private insurance plans that around one-third of Medicare enrollees choose over traditional Medicare—as part of their ban on all private insurance that offered the same coverage as their new entitlement. That would mean stepping on a proverbial “third rail” of American politics by making a huge number of current retirees give up the insurance plans that they chose. Just imagine what Fox News would do with that one.

If we want to get rid of the Medicare Advantage plans, it would be better to grandfather them in for current enrollees and stop offering them to new beneficiaries. But it would also be smart to allow private insurers to continue to sell the kind of heavily regulated policies that they currently do under the Medicare Advantage program to companies and prime working-age individuals who buy into Medicare. Make the public benefit truly public, and give private insurers a role in covering those who get their insurance through their employer.

Bernie Sanders is right that you should begin a negotiation by asking for more than you hope to get. Progressives shouldn’t negotiate with ourselves by starting in the center. But if you go into a salary negotiation hoping to get $60,000 per year, and you start by asking for $600,000, they’ll either think you’re in a completely different league, or nuts, and show you the door. It’s important to articulate a viable transition away from our deeply entrenched system.

Our health-care costs are so high that major reform will doubtless generate major opposition, but this approach avoids the kind of immediate shock to the existing system that would likely doom the approach embraced by Bernie Sanders and John Conyers. It doesn’t fit neatly on a bumper sticker, but we could still capitalize on Medicare’s popularity and familiarity by calling it an alternative pathway to Medicare for All.

It may be incremental, but it’s not timid. It would represent a massive expansion of public-health coverage that would inevitably face concerted opposition, and be difficult to enact into law, but by building on demonstrable progress over time rather than pushing everyone into a new insurance scheme based on a promise that it will be superior, it would make a popular backlash less likely.