It is an illusion, really, that we alone determine what we choose. You walk into a restaurant and order a sandwich. Your spouse asks you what movie you want to see. Your doctor asks if you want to try a new drug to control your cholesterol. You search for a flight home for the holidays. You select an investment fund from your employer's retirement plan. You may not realize it, but in each choice, you have a hidden partner.
You choose something, but the restaurant, your spouse, the doctor, the airline's website, and your employer have all made decisions about how to pose those choices to you. Their design decisions will influence, intentionally or not, what you choose. The restaurant organized the sandwich options you saw, maybe with vegetarian on the right, meat on the left. Your spouse suggested some possible movies, but left others out. The doctor decided how to describe the consequences, side effects, and benefits of the drugs you could take to control your cholesterol. The website decided how to sort the flights and whether to present certain pieces of information, like the on-time performance of the flights and the baggage fees. And finally, your employer had a fund it would use for your retirement account if you did not make a choice.
We might vaguely agree that external factors influence our choices, but we don't appreciate that there are large and systematic ways in which presentation can change what we choose. Years of research have shown again and again that the way choices are posed affects our actions. We're much more likely to choose a retirement fund if the employer's website preselects that option for us. We make different choices if our doctor says that serious side effects occur in 1 percent of cases, as opposed to saying there are no side effects in 99 percent of cases. And while you might be convinced that you are picking the sandwich, many things about the menu, like the order of items and the headings of the categories, helped you make that choice.
All these details are part of something called choice architecture, the many aspects of how a choice is posed that can be manipulated, intentionally or inadvertently, to influence the decisions we make. The options may be the same, but the presentation can change your choice.
Before you make a decision, someone has molded many of the characteristics of that choice for you, and these design decisions will in some way affect what you choose. This book takes a close look at exactly how the way choices are presented can change the decision-making process. Whether they realize it or not, the people who present those choices-the restaurant owner, your spouse, your doctor, your employer-are choice architects. Stated more simply, choice architects are the designers of the decisions you make, just as you are the designer of countless choices for others and for yourself. To keep things simple, I will call the choice architects designers and the people making decisions choosers.
Choice architecture can be harnessed in surprising and surprisingly powerful ways, with positive results for both the designer and the chooser. But to do so, we must look closely at every detail of a choice's design, including elements we don't normally notice. After all, when we encounter a choice, we are too busy making the choice to contemplate how the presentation of options affects us. But get those details right, and better decisions can emerge.
I became interested in decision-making and decision design as I contemplated college. Coming from a working-class, Levittown-like suburb in New Jersey, it struck me that many of my friends from high school were considering very different options for their lives. In my high school, there were many academically talented students, but some were delighted to go to the local community college, and others, equally qualified, were applying to Ivy League schools. What made them consider different options? I saw, in the years that followed, that the choice of what to consider in large part determined their future.
Being lucky (and foolish) enough to talk my way into places where pathbreaking work in decision-making was done, I found myself in graduate school at Carnegie Mellon, and a postdoc at Stanford, places central to the beginning of a revolution in decision research and behavioral economics. This revolution was based on the idea that people use simplified rules of thumb, or heuristics, to make decisions. The classic demonstration involved people making inconsistent choices, caused by things that should not matter. Describing the same options as gains or losses can change choices. Telling people about irrelevant numbers can change how much they are willing to pay. People are much more willing to pay for insurance against a vivid risk, like cancer, than they are willing to pay for insurance that covers all diseases. There was soon a list of these effects, which was sometimes called "stupid human tricks," after David Letterman's comedy bits.
These results were usually cast as showing how people are irrational and generally bad at making decisions. I had always been interested in public policy and improving decisions, and I realized that we could flip that framing for positive ends: instead of exposing people as bad decision-makers by showing they were inconsistent, maybe we could make them better decision-makers by exploiting these inconsistencies. By changing how their choices were posed, we might help them make better decisions.
About this time, I was at the Wharton School of the University of Pennsylvania, organizing a team studying insurance decisions. We noticed that Pennsylvania and New Jersey were about to introduce new, less expensive types of auto insurance. Each state required insurance firms to use a different choice architecture to present the auto insurance choices. We predicted, correctly, that because of the way the choices were presented, the new policy would be more popular in New Jersey than in Pennsylvania. We wrote the governors and insurance commissioners of both states and even wrote an op-ed in a Philadelphia newspaper. Everyone ignored us, but the impact of the choice architecture was massive. The new policy was three times more popular in New Jersey. Because of their choices, Pennsylvanians paid billions of dollars more for auto insurance. Years later, my friend Dan Goldstein and I used the same idea to show that choice architecture made a big difference in whether people agreed to be organ donors.
A few years after that, I was working with a German auto manufacturer, consulting on the choice architecture of its website. On the site, people could pick engines, interiors, colors, and many other options when buying a car. But the automaker was presenting those options to buyers in a way that was detrimental to everyone involved: the automaker was losing potential revenue by guiding buyers to the cheaper options, and those options didn't really meet the buyers' needs or desires. With a few simple changes, we were able to increase the company's profit and improve customers' satisfaction with the vehicles they ultimately purchased.
More recently, I was a senior visiting scholar at the newly formed Consumer Financial Protection Bureau, consulting on many issues related to choice architecture. How do you write disclosures for loans and mortgages? How can you display information about a complex product, such as a prepaid credit card that has many different fees, given that the card has to fit in a wallet? How can you encourage people to comparison shop? And my favorite: How can you make sure people know what they're signing up for with overdraft protection, a banking service that pays your bill when you overdraw from your checking account? The bank will often charge something like $35 for each overdraft. Some people want this coverage, but others are incensed when a cup of coffee ends up costing $38 because of an overdraft. How can you make sure that everyone who has the product has made a good and informed decision?
Choice architecture is not just about how websites are designed or how policies are implemented. It is not just about professionals and researchers. We are all designers every day, posing choices to our friends, colleagues, and families. My friend is also a decision designer when, instead of asking his three-year-old whether she is ready to go to bed, he asks if she prefers to fly into bed or bounce into bed. He reported that bedtime became much less stressful for everyone involved when he started presenting options instead of a yes-or-no choice.
Choice architecture has a lot in common with actual architecture. Winston Churchill knew this, saying, "We shape our buildings and afterwards our buildings shape us." He was talking about the British House of Commons, which had recently been destroyed by German incendiary bombs during the last raids of the Blitz.
The building had been an oblong rectangle, modeled on the design of St. Stephen's Chapel, the site where the House of Commons previously met. Members were seated on opposing sides of the chamber. Some members of Parliament favored switching to a semicircular arrangement, like the amphitheaters used in both the U.S. Senate and House of Representatives. Others pointed out that the old chambers didn't actually have enough seats for everyone.
Churchill, Britain's prime minister at the time, would have none of it. In one of his better speeches, he argued to preserve "all the essential features of the old building." While politically conservative, he was not being sentimental or a traditionalist. He believed strongly that the shape of the building influenced its inhabitants, and that the nature of political discourse depended on two essential features of the old Commons.
The first was the oblong design, dividing the Commons into two halves, one for the ruling party, the other for the opposition, with the two parties directly opposite each other. Churchill argued that this was essential to the two-party system that dominated British politics. Looking directly at the opposition facilitated debate, concentrated attention on the other side, and took away attention from any shifting alliances within one's own party. Your side viewed the common enemy. The distance between the two halves was, by tradition, two sword-lengths apart. Churchill compared this to a semicircle where the speaker is at the front, addressing the entire chamber, looking as much at their supporters as at the opposition.
The second was the size of the building, which was too small to hold all 650 members of the House of Commons. This facilitated a conversational style of debate, and a packed house gave the proceedings a sense of importance and urgency.
Churchill prevailed, and anyone can witness the impact of architecture by watching Question Time, a famously confrontational affair that occurs every Wednesday at noon when Parliament is in session. The questions, often thinly disguised attacks, are posed to the prime minister and supported by quips and guffaws from the opposition. It is quintessential politics as theater.
Just as the Commons' oblong shape focuses attention on the opposition and causes members to think about their reactions and comebacks, choice architecture focuses our attention and thought on particular options while neglecting others. Physical architecture and choice architecture are always present and always have an influence, even when we are unaware of them. While a building may lack an architect, it does have windows and doors. Someone decided the placement of the doors, and that determines where we enter and exit. Similarly, intentionally or not, choice architects present choices that will make a difference in what information we examine and what we ignore. Not knowing about choice architecture can result in designs that steer choosers in ways neither we nor they imagine.
Doctors are busy decision-makers. In the examination room they complete, on average, ten different tasks per hour, including taking a patientÕs history, discussing symptoms, and performing a physical exam. But doctors have a new companion: an electronic health record system (EHR). Every encounter with a patient involves using the system to record facts like a patientÕs blood pressure, diagnoses, and treatment decisions. When researchers look at how doctors spend their time, they see that the EHR is the second most important thing in the room. Half a doctorÕs time is spent communicating with and examining the patient, but a whopping 37 percent is spent working with the EHR. The old physician's prescription pad and pen have been replaced by a screen, keyboard, and mouse. This may improve record keeping, but an EHR is a choice architecture that can affect how doctors treat their patients.
Generic medicines are chemically equivalent to brand-name medicines but much less expensive. A brand-name drug, such as the antihistamine Allegra, can cost five times more than its generic equivalent, fexofenadine hydrochloride. Generics are not just cheaper but can also result in healthier patients. When medicines cost less, patients take them more faithfully precisely because they are more affordable.
Hospitals have tried many ways to encourage doctors to prescribe generics, barraging them with emails, holding seminars, and making other interventions that prove equally ineffective. EHRs often have pop-ups encouraging doctors to prescribe generics. They are quickly dismissed and ignored. The overuse of pop-ups results in doctors quickly dismissing all alerts without reading them, something called alert fatigue. One study even tried paying physicians to prescribe generics, but that too failed. Changing physician behavior is so hard that some states simply allow pharmacists to substitute generics for the brand-name drugs doctors prescribe.
Paying doctors to prescribe generics does not work because it addresses the wrong problem. A simple change in the interface, developed by researchers at the Weill Cornell Medical College, has proven much more effective. It is based on how physicians remember drugs. Doctors simply have an easier time recalling brand-name medicines. After all, Allegra comes to mind much more readily than fexofenadine hydrochloride. Brand-name drugs are also heavily advertised. When pharmaceutical companies give doctors free samples as well as notepads and pens with the drug's name on them, it cements that name in their minds.
This leads busy physicians to get into a habit. When they need to prescribe an antihistamine, they start typing "All" into the EHR and the system helpfully auto-completes the field with "Allegra." The designers thought they were aiding the doctors with this function, and these keystrokes become habitual.
The Weill Cornell team changed the interface so that the EHR automatically substituted the name of the generic drug when a doctor started typing the name of the brand. For example, when someone typed "All" the system would immediately offer "fexofenadine hydrochloride." Doctors could go back to the brand-name drug by checking a box that said "Dispense as written." But they seldom did. Doctors left the generic option in place almost all of the time. The EHR substituted its perfect knowledge of the generic equivalent for the doctors' imperfect or nonexistent recall.
This more than doubled the proportion of prescriptions for generics, and since generics are, on average, 80 percent less expensive than brand-name drugs, this produced large cost savings both for the hospital and patients.