Last week, the federal government unveiled its online health insurance marketplace, a milestone in the implementation of the Affordable Care Act. While much of the media coverage has focused on the technical glitches of the online portal—thousands of Americans reported long wait times, for example—a more fundamental policy question is at stake. The Obama Administration has argued that presenting insurance plans to consumers in one place and in a transparent fashion will spur competition among insurers, lower costs, and ultimately improve satisfaction. But what do we actually know about how consumers choose their health insurance plans? The stakes are high: If consumers choose wrongly, they may incur higher costs and premiums, and the federal government, which will subsidize these plans, may end up footing a larger bill. In this case, more competition may not lead to lower health care costs.

A new paper, published with the support of the Russell Sage Foundation, provides worrying evidence that consumers choosing health insurance are often prone to errors, over-confidence, and cost biases. The authors, Eric Johnson, Tom Baker, and Ran Hassin and their colleagues, used a model of the online health exchanges to test consumer decision-making. They asked subjects to pick a health insurance plan for a family of three, with a particular number of doctor visits and out-of-pocket health care costs over the next year. The results were dismal: When asked to choose among four plans, subjects selected the cost-effective option only 42 percent of the time, with the average error costing over $200. When presented eight options, subjects selected the correct option 21 percent of the time.

Why did these errors occur? Choosing health insurance is complicated. Consumers must make two calculations: first, they must estimate the amount of health care they will need (for example, the number of times they think they will visit a doctor in a year). Then, they must align their projection with a plan that has the right level of deductibles, premiums, and out-of-pocket costs, all daunting concepts. Moreover, past research suggests that consumers often over-weigh deductibles and out-of-pocket costs, skewing their preferences.

Still, the calculation is not impossible. The study asked MBA students at Columbia University to perform the same exercise on the model health portal, and they chose the correct option 73 percent of the time. Interestingly, the students that used Microsoft Excel during their calculations selected the correct option 85 percent of the time, making an error of only $47. "This suggests that having both the right mental model and the ability to execute these calculations may be a basic requirement to make good choices," the authors write.

Not all is lost. In a final set of experiments, the study tested interventions that could improve decision-making. When the model health care exchange included a cost calculator (along with an explanation of what the calculator did) and a "smart default" option that pre-selected the cost-effective insurance plan, performed improved measurably. Subjects were doing as well as the MBA students, and the average error fell from $533 to $77. The authors summarize their findings: "The bad news: Consumers left to their own devices seem to make large errors when choosing health insurance...The good news is that we have demonstrated that exchange designers can improve consumers’ performance markedly through the use of just-in-time education, smart defaults, and cost calculators. This list of potential design improvements is not exhaustive, and there are many other interventions that may improve choices."

The paper, "Can Consumers Make Affordable Care Affordable? The Value of Choice Architecture," can be downloaded here. It was funded by the Foundation's consumer finance working group, which explores new lines of behavioral economic research on consumer decision-making with the aim of improving the design of regulatory strategies in retail financial markets.

Rohan Mascarenhas is a student at the Harvard Kennedy School and a former communications manager at the Russell Sage Foundation.