A tracking poll from the Kaiser Family Foundation finds that 56 percent of Americans favor a fully government-run Medicare for All insurance plan, but that an even larger 68 percent favor a mixed public-private approach with a public option. The most common reason given by those who support a public option but oppose Medicare for All is a desire for choice. They do not oppose the idea of public health insurance, but they do not want to be forced onto it.

Given those public attitudes, it is not surprising that many Democratic presidential aspirants have shied away from Medicare for All in favor of plans based on a public option. Pete Buttigieg emphasizes the element of choice in the very name of his plan, which he calls Medicare for All Who Want It. Other candidates backing one or another form of public option include Joe Biden, Amy Klobuchar, and Michael Bloomberg.

There is a dilemma at the heart of the public option approach, however. Is it possible to offer choice in health care coverage and still achieve universal coverage, another cherished goal of reformers? Or does the attempt to achieve universality inevitably require making enrollment compulsory?

In my view, it should be possible to preserve meaningful choice in health care while ensuring universal access to coverage. But doing so will require attention to some details of program design that the candidates’ plans have not fully addressed.

The issue

Although the universal-vs.-voluntary dilemma is inherent in all public option plans, the version advanced by the Buttigieg campaign has drawn the most attention to the issue. At least part of the reason is that the Buttigieg plan comes closest to addressing the dilemma directly. Writers for The Washington Post, The New Republic, the Wall Street Journal, and Slate have all argued that his public option is designed in such a way that it would inevitably become compulsory.

The central question in all of this is how to ensure that the risk pool covered by the public option includes an adequate number of healthy subscribers along with those who are ill. If people can easily buy into coverage only after they become sick and drop out at will when they recover, the average cost of claims made by those left in the pool rises and premiums become unaffordable. The result is the notorious “death spiral,” known to economists as adverse selection.

One way to control adverse selection would be to make coverage mandatory. The Affordable Care Act included a step in that direction in the form of a tax penalty for healthy people who did not obtain coverage. Unfortunately, that “individual mandate” proved to be one of the least politically popular aspects of the ACA. Shaken by court challenges and rendered meaningless by Congress, the individual mandate is no longer in force.

Another way to deal with adverse selection would be to refuse care to people who do not have insurance. However, no one wants to see people dying on the streets because doctors and hospitals have turned them away. Already, the law requires that anyone who makes it to an emergency room must be at least stabilized regardless of ability to pay.

Still another approach would be to impose a late enrollment penalty for people who do not sign up for coverage when first eligible, waiting to apply for coverage only after they get sick. That is the strategy used by Medicare Part D prescription drug coverage. Medicare calculates the penalty by multiplying 1 percent of the “national base beneficiary premium” ($32.74 a month in 2020) times the number of full, uncovered months a person did not have Part D or creditable coverage. For example, suppose someone was not enrolled for 30 months before finally enrolling in February 2020. The monthly premium penalty would be 30 X 0.01 X $32.74, which comes to $9.80 a month (Medicare rounds the penalty to the nearest ten cents) or $117.60 per year, paid for the rest of the person’s life. As of 2020, the theoretical maximum annual penalty can be over $600 a year.

These approaches could be combined in any number of ways. For example, although it does not fully spell out the details, the Buttigieg plan appears to call for a combination of mandatory enrollment and late enrollment penalties. In this case, however, the late enrollment penalty takes the form of retroactive monthly premium payments for a full year of coverage rather than an increase in premiums going forward, as under Medicare Part D. Since premiums under the Buttigieg plan could be as high as 8.5 percent of income, retroactive enrollment could mean an unexpected bill of thousands of dollars. The Buttigieg team has also considered requiring people to show proof of insurance when paying taxes. Those who could not do so would be automatically enrolled in the public option, with premiums collected through the IRS.

Critics say that such measures would make the program “Medicare Whether You Want It Or Not,” producing a political backlash far greater than the Obamacare individual mandate. Fortunately, however, there are better ways to deal with the universal-vs.-voluntary dilemma.

Guidelines for a balanced solution

1. Automatic enrollment for the poor. All candidates’ versions of a public option include premium subsidies for people with low incomes and waive premiums altogether for people below a low-income cutoff, most likely set at the federal poverty level. Automatic enrollment would be a reasonable way to achieve universal coverage for people whose incomes were low enough to qualify for a zero premium.

For example, Biden’s version of the public option envisages automatic enrollment for anyone below the low-income cutoff at whatever point they come in contact with government services, such as enrolling children in school, applying for food stamps, or renewing a driver’s license. Anyone who fell through the cracks would be enrolled automatically the first time they sought health care services. Since low-income people are not subject to premiums, they would face no surprise bills or retroactive penalties.

2. Eligible income, not total income. Once people are enrolled, administrators would have to monitor tax returns for increases in income. It would be desirable to avoid sharp premium increases for people who move from just below the poverty line to just above it. Such variations in income are not unusual for low-skill workers who have sporadic connections to the job market or who have little control over hours worked. One way to avoid undue premium shocks would be to peg premium subsidies to “eligible income,” rather than total income. Eligible income, in this context, means total income minus the low-income cutoff below which premiums are fully subsidized.

For example, suppose the low-income cutoff is equal to the federal poverty rate (about $25,000 for a family of four) and the maximum premium is set at 8.5 percent of eligible income. If the premium were based on total income, a family earning exactly $25,000 would owe no premium, but one with income of $25,100 would face a nontrivial premium of $2,133 a year, or $178 a month. Abrupt “benefit cliffs” like these distort work incentives and encourage concealment of income. If premiums were instead based on eligible income, the extra $100 of income above the low-income cutoff would mean just $8.50 a year in premium payments. Avoiding abrupt premium increases at the poverty threshold would encourage people to earn more and could reduce total spending for the insurance scheme in the long run.

3. Self-insurance as a legitimate option. Although automatic enrollment in the public option would be appropriate for people at or below the poverty level, I do not see it as either necessary or desirable for those with higher incomes. Self-insurance should be a legitimate option for people who voluntarily choose to go without insurance, provided they are willing and able to pay their own medical expenses. Devices such as tax-free health savings accounts make that a reasonable choice for many young, healthy people with good incomes. Proactively hounding them to enroll is only going to make a lot of voters angry, as did the ACA’s individual mandate.

In one sense, legitimizing self-insurance would appear to mean giving up on the goal of universal coverage. However, universality in the sense of universal access to insurance coverage would be maintained so long as late enrollment in the public option (with appropriate penalties) remained possible. Guidelines 4 and 5 explain how that could be done without opening the door to uncontrolled adverse selection.

4. The role of a backstop fund. An innovative provision of the Buttigieg health care plan is the use of a “backstop fund” as a way of dealing with people who neither voluntarily sign up for insurance nor meet the income cutoff for fully subsidized premiums. However, the role of the backstop fund should be somewhat different than envisioned by the Buttigieg campaign.

In the Buttigieg version, the backstop fund appears to cover the expenses of uninsured persons who seek care, and is then replenished by late-enrollment penalties (retroactive premiums) paid by those same people. In my view, the principal role of the backstop fund should instead be to protect providers, especially emergency rooms, against uncollectible bills. Under current law, hospitals end up recouping uncollected emergency room bills through higher charges on other patients, including those whose own insurance is fully in order. With a proper backstop fund, that would be unnecessary.

The backstop fund should not, however, protect the voluntarily uninsured people with adequate incomes from the need to pay for services. If such people seek non-urgent services and cannot present an insurance card, the provider should have the right to ask for payment up front. If they go to an emergency room and are not able to pay on the spot, the backstop fund should reimburse the provider but should then have the right to enroll the patient in the public option, with a substantial late-enrollment fee, possibly in the form of retroactive premiums. If the late enrollment fee is less than the charges for care, the backstop fund should have the right to seek payment of the excess amount from the patient.

5. Late enrollment. A somewhat different problem would be posed by people who opt for self-insurance but apply for coverage before they actually seek services. Those would include both healthy people who simply change their mind about insurance and people who learn that they have developed conditions such as, say, diabetes, that will require ongoing care.

Private insurers, including employer-sponsored plans, should be allowed to refuse such applicants or to impose actuarially fair premiums based on preexisting conditions. The public option should be required to enroll them at standard rates, but should apply substantial late-enrollment penalties, perhaps modeled on that used for Medicare Part D or perhaps in the form of retroactive premiums. (However, brief gaps in coverage should not be penalized. For example, people who lose employer-sponsored insurance or who lose coverage due to a change in family status should be allowed to enroll in the public option without penalty, provided they do so promptly. For example, Medicare Part D allows a 63-day grace period.)

Conclusions

It is understandable that many voters prefer the idea of choice that is offered by a public option over the one-size-fits-all approach of Medicare for All. However, as this commentary explains, a plan based on a public option needs to be designed carefully if it is not to fall victim to adverse selection.

A public “option” that turned out to be compulsory would be a political non-starter. A public option that had no mechanism for controlling adverse selection would not be financially viable. And a public option that provided no backup enrollment mechanism for people who initially opt for self-insurance but later face changed circumstances would not meet ethical norms of fairness. The guidelines offered above should make it possible to steer a path between these unacceptable outcomes and arrive at a program that is both universal and voluntary.

Finally, the fact that it would be possible to design a public option in a way that resolves the universal-vs.-voluntary dilemma does not mean that is the best way to go. As I have written elsewhere, a health care system based on the principle of universal catastrophic coverage could resolve the dilemma in a much more straightforward fashion. Under such a plan, everyone could automatically receive coverage for financially catastrophic health care needs without the payment of any premium. “Financially catastrophic” would be defined in relation to income, with full first-dollar coverage for the poor and full financial responsibility for the affluent. Alternatively, the Fair Care Act offers a Republican approach to health care reform that would at least come close to universal protection without public options or mandates of any kind.

Whatever approach is taken to health care reform, it is critical that the universal-vs.-voluntary dilemma not simply be ignored. Any workable plan must be designed to deal with it from the outset, or face a public backlash after the window for a suitable legislative fix has closed.