Part II: Healthcare Choice Architecture

doctor and nurse treat baby

By Brian Chiglinsky

It was the greatest scientific advance involving jelly in the pre-Uncrustable era. In 2000, students at Columbia University’s Business School and Stanford’s Psychology Department published a paper testing people’s commitment to jam. Specifically, they were looking at how a wealth of options affects a person’s buying habits.

The researchers set up two displays of jam – a limited choice option with only 6 flavors, and an extensive-choice option with 24 flavors. The display with 24 drew more onlookers, but ultimately fewer people actually purchased from that display.

Aside from proving that too many people can’t accept the clear perfection of grape jelly, the study also illustrated the psychological phenomenon of choice overload. Often, too many options exhaust and dissuade consumers from committing to a decision. Subsequent studies also found that too many options led to worse performance and less satisfaction with a decision.

This research fits into the crux of behavioral economics – the field that sits at the fascinating intersection of psychology and economics. Stephen Colbert  aside, none of us is perfect, and our decisions are not always perfect either. And these imperfections fall into familiar patterns, or heuristics, that can be identified and studied.  Behavioral economics is already transforming the implementation of public policy and the way we look at poverty.

It might also have some implications for health care. The national consensus around reforming our health care spending – specifically in reducing costs and improving outcomes – has focused on giving people a more active role in choosing how they get their health care. This was the consensus behind the Health Insurance Marketplace, and tools like Hospital Compare and Physician Compare for Medicare patients. Give people more choice, the thinking goes, and they’ll choose the options that lower their costs and fit their needs.

Behavioral research is starting to challenge this thinking. At a recent panel hosted by The Hamilton Project, Ben Handel and Jonathan Kolstad from UC Berkeley presented a paper called “Getting the Most from Marketplaces: Smart Policies on Health Insurance Growth”.  They point out that studies have shown that people who get health insurance from their employer often choose poor plans for their health needs and end up overspending by an estimated $2,000 every year, or spending 42 percent more on premiums than they need.

A lot of this is from things like inertia – merely not consciously shopping for a new plan every year. As a writer, I completely understand the tendency toward procrastination. But a lot of it is also that health insurance is complicated as hell. The monthly premium is often the most obvious point of information, but there are tons of complicating factors, like deductibles, cost-sharing, and the narrowness of provider networks, that can overwhelm people just trying to pick a plan to get their family through the next year.

But don’t worry, the behavioral econs have been here before. Their fingerprints are on the design of the insurance marketplaces created by the Affordable Care Act. The four different “metal levels” that distinguish the quality and price of plans help consumers broadly sort out what kind of health insurance they’re looking for.

Handel and Kolstad take that one step further. They propose two broad policies – one that provides a person with specific, individual-level information on their predicted costs and benefits under each health plan; and another, more libertarian-paternalist one that actively defaults individuals into plans that are estimated to have substantially lower costs with just as much protection, and the same network of doctors. The first proposal is their attack on choice overload; the second, their attack on inertia.

Their proposals are more fascinating than they are imminent – political challenges aside, it’s not clear that anyone has detailed enough data or that this data would be shared for these estimates. But they highlight the limits to the philosophy of flooding a consumer with information to improve health care.

They also bring up interesting dilemmas in how someone addresses choices in health care that you don’t find, say, in the jelly market. Choosing the wrong jelly doesn’t carry the risk of bankrupting you, it doesn’t affect your approach to your family’s care, and it hopefully isn’t related to unforeseen circumstances like illness (if it is, please for the love of God get different jelly).  

It’s also just not clear how much this new choice architecture could save. Although estimates pegged consumer losses in the $2,000 range in previous studies, these are studied after already knowing how consumer spending on health went for that year – and please don’t make me use that cliché about hindsight. It’s just not clear how often there are clear options for health insurance that are simultaneously cheaper and offer more comprehensive coverage.

But it might be worth exploring more by opening data and letting researchers in. Government could help alleviate the extra risk for consumers by paying when costs go beyond coverage obtained through a default, and by making it easy for consumers to opt-out of defaults or information-sharing entirely, they could make it a bit less liberty-encroaching. In the process, small savings for families could add up to substantial savings from the health care system as a whole.

In a place that’s been a dark, locked room for consumer choice and options like our nation’s health care, more sunlight is still the best broad strategy. But as our system heads toward one that empowers the individual patient even more, we should keep in mind the importance of how we present choices and how we help people get to outcomes that make them happier and more satisfied. And, in the process, maybe save them enough money to spend more on trying those 24 different types of jelly.