The “moderates” in the Democratic presidential primary hope to build on the Affordable Care Act, rather than make the more dramatic shift to Medicare-for-all — and even Sen. Elizabeth Warren (D-Mass.) has changed course to embrace a plan that would move to Medicare-for-all more slowly. But if they want to build on the ACA, they need to confront a serious perversity of the private insurance system underlying it: The system’s financial health depends on people choosing the wrong plans.

A central goal of health-care restructuring ought to be radically simplifying insurance choices, so people don’t mistakenly choose plans that cost them thousands of dollars more than necessary each year.

Multiple studies demonstrate that despite the high stakes — financial and medical — many people make objectively poor health insurance decisions. In one study from 2015, three behavioral economists analyzed insurance choices of 50,000 workers at a Fortune 500 company that offered its employees a menu of health plan options: The workers could mix and match such features as deductibles, co-pays and out-of-pocket maximum payments, with the result that 48 distinct plans were possible. More than half of these employees ended up choosing plans that were objectively worse than alternatives. They chose to pay $500 more in premiums, say, to reduce their deductible by $250. As a result of misguided choices, the average person utterly unnecessarily spent the equivalent of 2 percent of their salary — and low-income, female, older and chronically ill patients spent even more.

Similarly, in 2015, a pair of MIT economists analyzed the insurance choices of millions of people deciding on Medicare plans to cover medication expenses. Medicare’s Part D offers basic drug coverage through private plans that offer a range of premiums, co-pays and so on. Again, there are sometimes objectively correct decisions in these cases: People who take specific medications should choose plans that generously cover those treatments. In this case, however, only 12 percent of people chose the best plans for their circumstances, and the average person spent 24 percent more on medicines than they would have spent under an alternative available plan.

To understand why the financial viability of the ACA’s insurance plans rely on bad choices, imagine what would happen if everyone made good ones. Consider two plans that an insurer might offer on an exchange. One is a low monthly premium plan with high deductibles, the other a higher-premium plan with low deductibles. All else being equal, people with expensive chronic illnesses ought to choose the low-deductible plan, paying more each month on premiums to avoid high out-of-pocket expenses for medical care they know they’ll need. By contrast, relatively healthy people ought to gravitate toward the first plan, saving money each month while taking a calculated risk that they never get sick enough to be responsible for the full cost of their deductible.

However, if a high-premium, low-deductible plan attracts an increasing number of people with expensive chronic conditions, the insurance company will be forced to raise monthly premiums so that the plan will survive financially. That will cause even more relatively healthy people to shift toward the high-deductible plan. The result will be another round of premium hikes for the low-deductible plan and further flight of healthy consumers.

This is the dreaded death spiral, which ends in plan insolvency. And it would happen much more often if people made the most efficient insurance choices for themselves and their families.

Why do people make insurance mistakes? One problem is opaque terminology. In a normal consumer context, the word “deduct” suggests a bargain (wouldn’t you like your car dealer to “deduct” $2,000 from the list price?). But a $2,000 insurance deductible means your insurance won’t cover your expenses until you’ve spent $2,000 out-of-pocket on medical care. In a nationally representative survey of privately insured Americans, more than 1 in 5 couldn’t define “deductible” (though 97 percent were confident they knew the answer). In that same study, almost half were confused about the meaning of the term “maximum out-of-pocket cost” — the total paid out in deductibles and co-pays before an insurer takes over full coverage of a person’s expenses for the remainder of the year.

There’s also sheer information overload, which aggravates the problem. Monthly premiums, co-pays, in-network vs. out-of-network deductibles, drug plans — who other than an actuary is going to be able to make a wise decision when faced with a dozen plans that vary across these dimensions?

The Medicare-for-all systems proposed by Sen. Bernie Sanders (I-Vt.) and (on a slower timetable) Warren solve the issue in one way: People would no longer make bad insurance choices because they would no longer have insurance choices. Everyone would receive coverage from a single plan regardless of how healthy they are; no one with chronic illnesses would be priced out of the market.

But other candidates and voters are opposed to that approach, either because they think it is politically unworkable or just bad policy. If they stick with the ACA and private plans, they ought to make it much easier for people to select plans that make the most sense for them.

Radical simplification of plans might mean Congress would mandate that every plan on the ACA exchange have the same deductible. Based on research suggesting that deductibles discourage people from receiving necessary care, I would prefer this uniform deductible to be either small or nonexistent.

That doesn’t mean that ACA revisions should eliminate all out-of-pocket expenses. Well-designed co-pays can make it less likely that people seek unnecessary medical interventions. Ideally, Congress would push for coverage built on the idea of reducing co-pays for necessary, high-value services (such as statins, for people with a history of heart disease), while increasing co-pays for medical care of dubious value (such as MRIs for simple low-back pain) — an approach known as “value-based insurance design.”

Insurance choices would still exist. For starters, plans would differ in cost, with some insurance companies achieving lower premiums by negotiating lower prices with providers and others trimming expenses by reducing their operating margins. Plans would also differ according to the services they offer enrollees (such as broader dental or vision coverage), the network of providers they offer, or even the programs they develop to promote their customers’ health.

Such revisions wouldn’t eliminate all confusion, and educating consumers would remain important. But crucially, the survival of the system — the avoidance of a death spiral — would no longer depend on self-defeating choices by befuddled consumers. That would reduce the incentive to confuse people.

The point of insurance shouldn’t be to trick healthy people into buying expensive plans they don’t need. A good insurance system should provide affordable coverage for everyone, whether they are sick or healthy, without confusing the living daylights out of them.