When it comes to picking health insurance, many consumers make choices that are detrimental to their financial health and that can cost them hundreds if not thousands of dollars a year.

This appears to be true whether they are shopping for coverage on a government exchange or evaluating offerings in an employer-sponsored benefit program, research shows.

Part of the problem is what you would expect: People struggle with questions involving numeracy, uncertainty, risk and loss. What’s more, few decisions are as emotionally wrought as those involving one’s health or financial well-being (or that of one’s family), which makes mistakes all the more likely.

But research also has found that the common errors people make—such as overpaying for plans that aren’t a good fit—can be exacerbated by the way insurance options are presented to them.

The good news is that by understanding the biases that lead to bad decisions, people can learn to make better choices. These insights also can be used by consumers who are in the market for other types of insurance coverage, such as homeowners, since the studies suggest that people make a set of consistent errors when shopping for insurance products.

Deductible bias

The first error is one we call deductible bias. A study by Saurabh Bhargava, George Loewenstein and Justin Sydnor, published in the Quarterly Journal of Economics last year, asked people to select from a small menu of health-insurance plans that differed in terms of premiums and deductibles, but offered the exact same medical network.

To illustrate, imagine you were asked to select a health plan from the following two options:

1) Plan A, which has a premium of $1,400 and a deductible of $500

2) Plan B, which has a premium of $800 and a deductible of $1,000

In the study, the two plans were equally popular. Plan A probably seemed like the more prudent option to many of the people, since it comes with a much lower deductible.

Alas, choosing a low-deductible plan can be a big financial mistake.

Here’s the simple math to explain why. If you never got sick, then Plan B is obviously the better choice, since you wouldn’t have to worry about your higher deductible. (You would only pay $800 in premiums a year rather than $1,400.) But even if you did get sick and used up your entire deductible, Plan B is still the better choice. After all, the maximum annual cost of Plan B is $1,800. ($800 in annual premiums, plus a $1,000 deductible.) That’s still $100 less than the worst-case scenario of Plan A, which is $1,900. ($1,400 in premiums, plus a $500 deductible.) In this example, the lower premium and higher deductible plan is always the cheaper option.

This example might seem far-fetched but it was adapted from an actual health-plan menu offered to employees at a Fortune 100 company. In the same paper, the researchers showed that the majority of employees at that company exhibited this deductible bias, which resulted in them spending 24% more on health insurance than they needed to, or the equivalent of about 2% of their annual salary.

The authors concluded that these costly mistakes in plan choice were driven by a fundamental misunderstanding of how basic insurance features, such as a deductible, relate to the overall price of health care. One might even view this excess spending as an annual tax levied on those who lack a sophisticated understanding of how insurance works.

The problem with labels

Recognizing the complexity of health-insurance decisions, many policy makers and plan sponsors designed health-plan exchanges and offerings with an eye toward simplifying consumers’ choices.

For example, the online exchange from which consumers enroll in Affordable Care Act, or ACA, plans organizes health plans into metal tiers defined by the generosity of their financial coverage—ranging from Bronze plans (covering a lower share of costs, at a lower premium) to Platinum plans (covering a higher share of costs, at a high premium).

However, our own new research with Dr. Loewenstein, a professor of economics at Carnegie Mellon University, shows that these labels, however well-intentioned, can backfire. We demonstrated this by simulating choice from a simple menu of three plans designed to mimic the choice of plan tier faced by a typical ACA enrollee. We then approximated the total spending associated with each plan, using information about average medical costs and personalized estimates of participants’ medical use.

To see how the metal labels influenced the quality of consumer decisions, we compared them against three other kinds of labels: generic labels such as Plan A, B and C; generic labels accompanied by a personalized plan recommendation; and labels that explicitly linked plans with levels of anticipated medical use (high-use, medium-use, low-use).

In short, we found that while people made insurance mistakes regardless of how plans were labeled, the metal labels led to the worst decisions. Only one-third of individuals choose the plan best suited to their medical needs (the same rate as random-chance), resulting in overspending equivalent to about one-quarter of average plan premiums—an amount bigger than most of the year-to-year changes in premiums that seem to dominate political discourse.

Why do the metallic labels backfire? One possibility is that people treat metal labels as signaling differences in the quality or reliability of health care rather than differences in financial coverage. (For example, one might assume the Gold plan provides superior quality of care to the Bronze plan, even though the latter may be more appropriate for someone in good health). In a study, published in the New England Journal of Medicine, researchers Peter Ubel, David Comerford and Eric Johnson switched the labels of the Bronze and Gold plans. They found that many people still said they preferred the Gold plans, and that the majority of people who scored below average in mathematical ability preferred the Gold plan, even when it was just a Bronze plan in disguise.

Solutions

How can this research help people make better insurance decisions?

First, consumers need to think carefully about how much loss they can afford. The optimal deductible for insurance involves a loss that is somewhat painful but not devastating. Instead of seeking out plans with the smallest deductibles, and thus incurring substantially higher premiums, they should be willing to trade a manageable deductible for lower premiums.

Consider home insurance. While many consumers focus on getting a low deductible, what’s often more important is ensuring that they have enough coverage to rebuild their home in the event of a disaster. (It’s easy to forget how expensive rebuilding can be.) In fact, many people might want to consider trading a higher deductible for more overall coverage, just to avoid a big increase in their premiums.

Our second recommendation is for people to focus on their expected need for a given insurance product, rather than some arbitrary label. Are you young and in good health? You may not consume enough health care to benefit from the increased financial coverage offered in a platinum plan, so you’re paying for something you don’t need. On the other hand, a primary earner with three children might need more life-insurance coverage than he or she thinks. A policy for $250,000 may sound like a lot, but the level of coverage is far below what most experts would recommend.

By understanding the kinds of biases that lead to errors, we can help people avoid buying coverage they don’t need, while investing in the protection they do.