Americans value “choice.” The word itself is attached to different sides of many major national debates, such as the one over abortion services (“pro-choice”) as well as the one over school privatization (“school choice”). In health care, choices are ubiquitous: people often have to decide which insurance company’s plan to select, whether to enroll in a plan with a high premium but a low deductible or vice versa, and which doctor, hospital, or pharmacy to use—to name a few. No health reform proposal, across the ideological spectrum, leaves Americans with no choices of providers, services, or sites of care. But even like-minded proposals differ on the extent of choice among health plans.

Looking across the range of potential public plan proposals for health reform in 2020, the ones with the most and the least choice of health plans are both amongst the biggest (measured in terms of eligibility). The Medicare-for-All proposal (S. 1804), for instance, would replace employer-based and individual market private plan coverage with a modified Medicare program. It would maintain a role for the Veterans Health Administration, Medicaid for long-term services and supports, and private supplemental coverage. These would wrap around, rather than be alternatives to, the Medicare-for-All system. However, the proposal would preserve Medicare Advantage, a program that allows enrollees to get their Medicare benefits through private plans. But compared to existing private markets, Medicare Advantage has stringent rules on benefits, networks of providers, payments, and other aspects of plan operations. Essentially, this proposal would put private plans into a public program.

In contrast, the proposal for Medicare Part E would put a public plan into private markets. Specifically, it would offer a Medicare Part E plan in the Health Insurance Marketplaces for individuals and small businesses; allow large businesses to purchase this plan or any other private plan; and would let employers that fund their own health benefits use the Part E plan as an administrator of their employee benefits. Only people eligible for Medicaid would not have this Part E option (current Medicare beneficiaries would be in the current Medicare program under both Medicare for All and Medicare Part E).

The Medicaid option proposal would offer Medicaid as a plan choice to the same people as would Medicare Part E—except that the option would only exist in states that authorize it. It is unlikely that states like Texas or Nebraska, both of which have yet to extend Medicaid eligibility to low-income adults, would make Medicaid available to all non-elderly residents.

The plan choices in the House public plan proposal from 2009 and Medicare X mirror those of the Part E proposal, except that they would not offer a public plan option to large businesses. The rationale is that large businesses have the leverage to secure affordable plans without access to a public plan. This approach also lessens the extent to which Americans’ health care providers would be paid Medicare versus private-plan payment rates, potentially increasing its palatability to health care providers and decreasing the expansiveness of the proposals’ impact.

Like Medicare for All, those proposals that open the doors of Medicare to specific populations—ending the waiting period for people with disabilities and Midlife Medicare—make it the default option rather than a competitive alternative in private markets. They too would offer private plan delivery of Medicare benefits through Medicare Advantage.

Lastly, proposals that inject public rates and programs into private plans would maintain—but not directly expand—existing choices. In fact, eligible individuals would simply benefit from these types of proposals without taking any action. To explain: a public reinsurance program would apply to plans that cover all privately insured people and would use its funding to leverage lower payment rates for high-cost claims within those plans. Similarly, the proposal to set a ceiling of Medicare provider payment rates for private individual market plans would affect insurers’ negotiations with providers but be invisible to those plans’ enrollees. As such, these proposals would not rely on people’s choice of a public plan (like Medicare for All) but they would not force people to change plans to get the benefit of lower prices from a public program (unlike Medicare for All).

These types of choices and their design matter because of their implications for costs and satisfaction. Choices structured to control for all but price and quality can reduce costs. And allowing choices can increase satisfaction. But choice poses challenges with insurance: people’s level of health care needs can fluctuate over time, and so they tend to pick different plans that match their needs. Pooling their premiums and spreading their costs across plans is critical for financing the system while maintaining plan options. This may not be possible if there is significant variation in plan design along with frequent switching of plans at the individual level as people’s health care needs change.

Moreover, the concept of “choice” has animated the Trump administration’s proposals to expand the sale of short-term and association health plans, among others. Some suspect this is subterfuge for driving plans with Affordable Care Act consumer protections out of the system by leaving them with fewer, sicker enrollees. The same criticism could be leveled at a public plan “choice” proposal that has different and preferential terms, which would thereby cause insurers to stop offering private plans. This parallel case illustrates why choice remains simultaneously attractive and dangerous in public policy and should be designed with care.