How Goes the Behavior-Change Revolution?

An all-star team of behavioral scientists discovers that humans are stubborn (and lazy, and sometimes dumber than dogs). We also hear about binge drinking, humblebragging, and regrets. Recorded live in Philadelphia with guests including Richard Thaler, Angela Duckworth, Katy Milkman, and Tom Gilovich.

About two years ago on Freakonomics Radio, we interviewed Angela Duckworth and Katy Milkman, researchers at the University of Pennsylvania, about an audacious new project they called Behavior Change for Good. It gathered together a “dream team” of behavioral scientists with the goal of advancing the science of behavior change and helping more people make good decisions about personal finance, health, and education.

On this week’s episode of Freakonomics Radio, we caught up with Duckworth and Milkman on the status of their project and the latest research in their circles. To that end, we also heard from four members of Duckworth and Milkman’s team about their research on pettiness; humblebragging; whether dogs are smarter than people; why people have such a hard time choosing the right health-insurance plan; and what people regret the most. Also joining us this week: Richard Thaler, a Nobel laureate who helped create the field of behavioral economics, who explained why psychologists may be better than economists at generating behavioral change.


When we first interviewed Milkman and Duckworth about their new project back in 2017, Milkman said they believed “the biggest problem in the world that needed solving was figuring out how to make behavior change stick.”

That’s the primary goal of Behavior Change for Good. Have they solved the problem yet? Short answer: hardly.

They’ve been running a big randomized, controlled trial involving 63,000 members of 24-Hour Fitness gyms. The participants signed up to take part in incentive programs designed by project scientists, with the goal of getting participants to build a lasting exercise habit.

Milkman describes the experiment as mostly a failure. The researchers tested 52 different programs that included nudges such as reminders and incentives such as cash (as well as one more program that consisted of no incentives and served as a control). In all 52 of the experimental programs, participants did go to the gym more during the 28 days of the experiment. The change, however, didn’t last.

“Most of the different versions of the program did create behavior change — so, like 50 to 75% created significant boosts in exercise for 28 days,” Milkman says. “It’s just that we didn’t do very well creating lasting change. So, after our 28-day program, pretty much we saw nothing in terms of behavior change. All 53 versions of the program, pretty much nothing sticks.”

Milkman and Duckworth aren’t giving up yet. They’re planning to continue the experiments with gym members. And in addition to experiments in education and personal finance, they’re also planning projects on medication adherence and childhood obesity.


When you go out to dinner with friends, and the check arrives, what do you do?

“Do you say, ‘Let’s just split it and all put in our credit cards?’” asks Mike Norton, a member of the Behavior Change for Good team and a professor at Harvard Business School. “Or are you the guy who takes the check and calculates everything and says, ‘Well, I only had six croutons so let me — I’m just gonna pay this much’?”

According to Norton, about 30% of people fall into the latter group. These people, Norton’s research indicates, are likely not very popular. In one experiment, Norton — with co-authors Tami Kim and Tina Zhang — presented participants with information from two different types of strangers’ Venmo accounts. (The payment app Venmo allows users to automatically and precisely split a bill and transfer funds to cover that bill.)

The first type of stranger is a penny-pincher — they’re the type that will go out to dinner with friends and, later, transfer $9.99 or $10.01 to a friend for dinner. The second type of stranger will split the bill evenly and just transfer $10. Norton found that research participants aren’t fond of the penny-pinchers, even those penny pinchers that transferred slightly more than $10.

“[O]ne thing that we tried to compare to is generosity,” Norton says. “So… who do you like better? Someone who pays you back $10, or someone who pays you back $10.03? So technically the $10.03 person is more generous. But they’re also really weird about money and really petty. And, in fact, that’s how much we dislike this behavior… we like the person more who paid us less. As long as they weren’t petty about it.”

Norton has also studied how this type of pettiness — which he defines as “attention to trivial details” — plays out in romantic relationships. He and his co-authors asked participants if their partners were petty with time or money.

“The answers to that question really, really predict not only dissatisfaction in your relationship, but we asked, ‘How upset would you be if your relationship ended?’” Norton says. “And people who are with a petty partner are less upset when they think about their relationship ending.”

Norton says that people engage in two different types of relationships with people and institutions: exchange relationships, such as the type people have with their banks, and communal relationships, such as the type people family have with family and friends. No one gets upset when their bank gets an interest payment correct, down to the cent. But people do get upset when their friends, or other people they’re in communal relationships with, treat them like a bank.

In that respect, Norton says the growth of apps like Venmo may have a downside.

“So what technology does actually — it’s more efficient, it’s better, it’s an improvement, but it actually is starting to default all of us into the dollars-and-cents world,” Norton says. “And there’s nothing wrong with that, but it does mean that it can be eroding social capital.”


Laurie Santos first started studying dogs because humans, via the domestication process, have engineered them to be just like us.

“So, if there’s anybody that’s going to be like us, any species that is likely to show our biases, dogs might really be one of those,” says Santos, a professor of psychology and head of Silliman College at Yale.

In one experiment, Santos and her colleagues focused on a phenomenon called overimitation, or humans’ penchant for imitating even silly, pointless behaviors. In an experiment conducted on four-year-old humans, Frank Keil and Derek Lyons at Yale showed children how to open a puzzle box, while also incorporating a bunch of unnecessary steps into the process. They found that the children copied all those

unnecessary steps, even when the puzzle box was very easy to open in one simple step. Researchers have documented similar behaviors among adults, albeit with more complicated puzzle boxes.

Santos and her colleagues repeated the experiment with dogs. They used a transparent box with a lid that could be easily flipped open to access a piece of food inside the box; the box also had an extraneous lever on its side. When the researchers showed the dogs how to use the box, they first moved the lever back and forth, over and over again.

“Now, in theory, if we did this with a human they would say, ‘I don’t really understand.’ Then lever, lever, lever, lever, lever, lever, open the box… ” Santos says. “That’s actually what humans four-year-olds do… what do the dogs do? Ran over, lifted the lid, and got the food.”

Dogs, in other words, are choosier about the social cues and behaviors they attend to and imitate. Santos says that an important implication of this work is that behavioral science researchers need to be mindful of humans’ inclinations to imitate other people.

“[B]ehavior change researchers have a phenomenon known as ‘social proof’… When you see other people doing it, you kind of think it’s a good idea,” she says. “I think most of the time we think of social proof, we think of good things. But there are all these domains in which it seems to go awry.”


Choosing health insurance is one of the more difficult and consequential financial decisions many consumers face. Justin Sydnor, an economist at the University of Wisconsin School of Business, has been studying how well people do with this decision. His findings are not encouraging.

In one experiment, Sydnor and co-authors Saurabh Bhargava and George Loewenstein partnered with a large company that decided to give its employees a lot of options when choosing an insurance plan. Employees were able to choose among 48 plans with different premiums, deductibles, copays, coinsurance, and maximum out-of-pocket amounts. Sydnor and his colleagues tracked the plans that employees chose and how much they subsequently spent on healthcare.

Part of what makes choosing health insurance plans difficult is that it’s often impossible to know how much healthcare you’ll need in the following year. But Sydnor and his colleagues found that most of the plans were a bad deal, regardless of how much health insurance a person might need.

“[I]t’s really the higher premium part that matters,” Sydnor explains. “So, what happens is the plans that had a lower deductible — say I wanted $500 instead of $1,000 — to get that plan I had to pay more than $600 extra in premium for the year. So best-case scenario, I might save $500. So, I get more insurance, but for sure I already paid over $600 for that.”

Sydnor outlined three possible explanations for these behaviors. People might genuinely be willing to pay (in the form of higher premiums) to avoid the shock of high deductibles. Consumers might be facing “choice overload.” Or people might really struggle to figure out how to calculate the benefits of each plan. Sydnor and his colleagues ran several online experiments to test the “choice overload” hypothesis; they found similar behavioral patterns when people only have two or four choices. People, Sydnor concludes, really just can’t make rational math calculations around health insurance decisions.

“[W]e’ve run some little experiments, and it looks like if you make it easier to compare the plans, you can really easily inform and improve these options,” he says. “But I think maybe the bigger implication is that we should just stop giving people choices about this.”


There are two types of mistakes people can make in life, says Tom Gilovich, a professor at Cornell University and one of the most prolific scholars in modern psychology. They make mistakes of action and mistakes of inaction. And there’s a particular type of mistake of inaction that people often cite as their greatest regret.

“[W]e’ve interviewed college students, prisoners in a state prison, a sample of geniuses,” Gilovich says. “In group after group after group, a very frequent regret is one of not doing something because of a fear of social consequences.”

Gilovich himself has one big regret of inaction. Five years after he got married, he came up with a solution to “the naming problem” — the question of how to decide what last name his new family should take. The assumption that women will take their husbands’ names, he notes, is sexist; combining names, however, only works for a generation.

So what was Gilovich’s solution to this conundrum?

“[A]t the very end [of the wedding ceremony], I say ‘Wait a minute. There’s one more thing. We’re gonna flip a coin to decide what the last name is… because it’s fair,” he says. “But what I like even more about it is that we don’t have any cultural institutions that celebrate chance, and chance is a huge part of our life. I think it’s my best idea. And unfortunately, it came five years too late.”


Richard Thaler, a recent recipient of the Nobel Prize in economics, is best known as a primary architect of what’s come to be called behavioral economics. He’s also the co-author of the wonderful book Nudge. One of the most successful nudges, and the greatest triumph to date of behavioral economics, has been Thaler’s work from a few decades ago, with fellow economist Shlomo Benartzi, in the realm of retirement savings. They argued that rather than relying on people to opt in to their 401(k), and fill out the burdensome paperwork and choose from countless investment options that confuse and intimidate people, it’s better to just automatically enroll them. This has resulted in millions of people saving billions of dollars for their retirement.

But what does it say about the field of behavioral economics, and behavior change generally, that this largest victory took place a couple of decades ago? Where are all the other victories? Thaler says that his retirement savings nudge had the advantage of being an easy fix.

“[P]eople are automatically enrolled… Then their rates are automatically escalated, and they’re given a default investment fund… ” Thaler says. “I think it’s no accident that that was a success because the fix was easy. Give me a problem where I can arrange things so that by doing nothing, people make the right choice, that’s an easy problem.”

Thaler is nonetheless glad to see Duckworth and Milkman, whom he describes as having “infinite energy,” tackling more complicated problems. And he approves of the large number of psychologists on the Behavior Change for Good team; he says they know a lot more than economists about how people form and maintain habits.

Thaler was recently reminded of just how hard behavior change is, when he was invited to a meeting in London that was focused on reducing binge drinking. There’s a long-standing tradition in England of buying rounds at the pub when out with friends — one friend buys the first round, the second friend buys the second round, and so on and so on. Obviously, if you’re out with eight friends, this tradition makes for a long night.

Thaler suggested that pubs institute a policy, for groups of three or more people, of running a tab. Thaler’s suggestion was leaked to the press; the proposal was not well received by the general public. People sent hate mail — letters saying they would never consider not buying a round for their friends.

“I can think up that change in the choice architecture, but how you would get that to change?” Thaler says. “I made no progress. We’re human. We have self-control problems. We’re absent-minded. We get distracted. And those things aren’t going to go away.”


You can find the full Freakonomics Radio episode, “How Goes the Behavior-Change Revolution?” at Freakonomics.com. You can also listen on Stitcher, Apple Podcasts, or any other podcast platform.

Stephen J. Dubner/ Freakonomics Radio

Written by

Stephen J. Dubner is co-author of the Freakonomics books and host of Freakonomics Radio.

Freakonomics Radio
Freakonomics Radio
Freakonomics Radio

About this PODCAST

Freakonomics Radio

Discover the hidden side of everything with Stephen J. Dubner, co-author of the Freakonomics books. Each week, Freakonomics Radio tells you things you always thought you knew (but didn’t) and things you never thought you wanted to know (but do) — from the economics of sleep to how to become great at just about anything. Dubner speaks with Nobel laureates and provocateurs, intellectuals and entrepreneurs.

Discover the hidden side of everything with Stephen J. Dubner, co-author of the Freakonomics books. Each week, Freakonomics Radio tells you things you always thought you knew (but didn’t) and things you never thought you wanted to know (but do) — from the economics of sleep to how to become great at just about anything. Dubner speaks with Nobel laureates and provocateurs, intellectuals and entrepreneurs.

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade