156 - Katy Milkman: The Art and Science of Lasting Behavior Change

4 Jul 2025 · 51 min · 18 chapters

Ask about this episode

Ask anything about it. ChatGPT or Claude reads this page and answers with the times it was said.

Connect VO and ask about every podcast you hear, including the moments you saved. Add to ChatGPT · Add to Claude

In short

Behavioral economics and psychology applied to lasting behavior change, focusing on Katy Milkman’s “Behavior Change for Good” mega-studies that run large randomized “tournament” tests of many interventions at once, aiming to identify scalable, low-cost programs that improve real-world outcomes.

Guest backgrounds

Katy Milkman is a Wharton (UPenn) professor with appointments in Penn’s Perelman School of Medicine and School of Arts and Sciences. She earned summa cum laude from Princeton, then a Harvard PhD in computer science and business. She co-founded and co-directs Penn’s Behavioral Change for Good Initiative; former president of the Society for Judgment and Decision-Making; fellow of the Association for Psychological Science. Public communicator (advised White House, Google, DoD; wrote for NYT/Economist; hosts Choiceology; author of How to Change).

Key claims

People are not perfectly rational; behavior change can be engineered using psychological insights. Large-scale, apples-to-apples mega-studies outperform one-off experiments for policy-relevant guidance. Mentorship accelerates academic growth.

Notable examples

24-Hour Fitness mega-study (53 programs; best: prevent “two misses in a row” with small incentive changes). COVID-19 vaccine messaging with Walmart Pharmacy/Penn Medicine/Geisinger (ownership language like “reserved for you” + multiple reminders).

Written by AI. May contain mistakes. Listen to the episode to check what was said.

Chapters

Tap a time to open that second in VO

Understanding Behavioral Economics

1:37 to 2:52

Katy Milkman explains the concept of behavioral economics and its relation to psychology.

“Welcome to the podcast, Professor Mokvin.”

The Evolution of Behavioral Economics

2:53 to 6:28

Discussion on the history and key figures in behavioral economics, including Kahneman and Tversky.

“I'm a little bit of, to some degree, an outsider even from that community.”

Katy's Perspective as an Outsider

6:29 to 7:59

Katy shares her unique position as a researcher straddling psychology and economics.

“I'm going to introduce a perturbation to your model that's based on psychological insight, and then I'm going to see if I can test my model empirically.”

Katy's Academic Journey

8:00 to 9:12

Katy discusses her winding path through academia and how she came to her current niche.

“I do do some lab research and I do try to look at the why, but it's not the driving feature of my work.”

Exploring Undergraduate Experiences

9:13 to 11:15

Katy recounts her undergraduate experience at Princeton and her struggles with economics.

“Yeah, because a lot of people are probably in similar ways.”

Bridging Disciplines in Academia

11:16 to 14:00

Katy explains her senior thesis project that connected operations research with American studies.

“And it was, I thought, really interesting.”

Exploring Thesis Connections in American Studies

14:00 to 17:44

Learn how the guest bridged statistics and American literature in their thesis.

“And Princeton requires that if you write a senior thesis, it has to connect the discipline of your major with the discipline of any.”

The Impact of Thesis and Early Career

17:44 to 19:55

Discover the surprising attention the guest's thesis received and its career implications.

“My PhD was in computer science and business at Harvard.”

Transitioning to Behavioral Economics

19:55 to 22:46

Understand how the guest's interest shifted towards behavioral economics and decision making.

“The New Yorker story is so cool, and I'm definitely going to check it out right after this.”

Researching Procrastination and Decision-Making

22:46 to 28:05

Examine the guest's dissertation work on procrastination in movie watching.

“They would be like, actually, I'm going to watch the summer blockbuster and I'll hold off on the documentary.”
Show all 18 chapters

Collaboration with Angela Duckworth

28:05 to 28:58

Learn about the partnership between Katy Milkman and Angela Duckworth focused on behavior change.

“She's done a lot of work on grit and understanding what predicts success, but she was just getting interested in, hey, I don't want to just understand traits that predict success.”

The MacArthur Foundation Competition

28:59 to 35:15

Discover the story of how a competition led to a significant initiative in behavior change science.

“So this is a funny story in a sense because it's the story of how a failure created a success.”

Launching Mega Studies

35:16 to 37:16

Explore the concept of mega studies and their role in advancing behavior change research.

“And we thought that was funny because we were like, yeah, that is the ultimate question.”

Insights from Exercise Interventions

37:17 to 41:34

Learn about impactful strategies from mega studies aimed at improving exercise habits.

“And there's one that's being launched shortly that Angela and I are both advising, but we're not running, related to trying to figure out how do we improve happiness.”

Further Research During COVID-19

41:35 to 42:05

Find out about the projects conducted during the COVID-19 pandemic focusing on health behaviors.

“We don't want people to feel like they've totally fallen off the wagon when they have a blip.”

Behavior Change Interventions and COVID-19

42:05 to 45:32

Learn about effective communication strategies for behavior change during crises.

“And he developed an intervention that focused on just giving people the information that there's a growing number of Americans who are interested in exercising and are exercising more.”

The Importance of Mentorship in Academia

45:32 to 48:22

Explore the vital role of mentorship in academic and personal growth.

“Is there anything that kind of speaks to you?”

Reflections on a Rewarding Conversation

48:22 to 49:28

Hear insights on the value of meaningful conversations and connections in academia.

“that can go poorly and that's its own mess.”
Hear the part that matters, and keep it.Open this episode in VO. Double tap your headphones to save a moment as you listen.
Get VO free

Transcript

Automatic transcript. May contain errors.

0:00Misha:Welcome back to the Stanford Psychology Podcast. I'm Misha, and today I'm so happy to be talking with Professor Katy Milkman. Katy is the James G. Dinan Professor at the Wharton School of the University of Pennsylvania, where she also holds appointments in the Perelman School of Medicine and the School of Arts and Sciences. After graduating summa cum laude from Princeton, she earned her PhD from Harvard University, where she studied both computer science and business. Her research explores how insights from economics and psychology can be harnessed to change consequential behaviors for good. To this end, she co-founded and co-directs the Behavioral Change for Good Initiative at Penn.

0:47Misha:Katie is a fellow of the Association for Psychological Science and the former president of the Society for Judgment and Decision-Making. Beyond her academic success, Katie is also a celebrated public communicator. She's advised numerous organizations, including the White House, Google, and the Department of Defense, and her writing has appeared in the New York Times and The Economist. She also hosts the popular behavioral economics podcast, Choiceology, and she's also the author of the international bestseller, How to Change. In today's episode, we'll talk about her work on large-scale mega-studies, the science of building better habits, and the important role that mentorship plays in an academic career.

1:36Misha:So, without further ado, here's our conversation.

2:06Misha:Welcome to the podcast, Professor Mokvin. It's such an amazing pleasure to have you on.

2:10Katy Milkman:It's great to be here. Thank you so much for having me.

2:12Misha:Yeah, of course, of course. So I want to start out, I guess, pretty broadly. So this is, of course, the Stanford Psychology Podcast, and a lot of our guests do psychology and psychological work and whatnot. And I invited you on as a behavioral economist. And I, of course, have my reasons for inviting you on. But I guess I just want to start with what is behavioral economics, maybe for the listeners who don't know, one. And two, what do you think the link is between psychology and behavioral economics? How did the two fit together? How can the two help each other? And how do you kind of see that?

2:49Misha:Yeah, those are great questions. So I'm going to do my

2:52Katy Milkman:best to define behavioral econ, but I have to tell you.

2:55Misha:I'm sure you'll do well. I'll try.

2:56Katy Milkman:I'm a little bit of, to some degree, an outsider even from that community. So behavioral economics, I think is a term used by economists to refer to a subfield in economics that tries to bring psychological insight to the original models that economists posited of people as perfectly rational actors. So as soon as an economist starts adding to a model of a perfectly rational actor, some behavioral traits like limited memory or impatience or inability to parse signals come from different correlated sources. You know, if you treat those correlated signals not as correlated but as independent, now that would be a behavioral model.

3:40Katy Milkman:And it really started with the incredible work of Kahneman and Tversky in the late 1970s, mid-1970s onward, that brought the idea that people are biased, that we have heuristics and biases, that we are loss-averse, that when we face risky prospects, We don't behave in the ways that a traditional rational actor model would posit. That started the field. And then folks like Richard Thaler, George Lowenstein, and, you know, of course, many others since then, Matthew Rabin, have taken it to another level, Colin Kammerer. There's sort of a giant number of people who ran with that ball and have made it into a serious subdiscipline in economics.

4:23Katy Milkman:And I actually would say that I am an outsider in some ways. So when I hang out with psychologists, they call me an economist. When I hang out with economists, they call me a psychologist. So I fit in neither bin. I do go to behavioral economics conferences, but I am one of the more psychologically focused researchers in those conversations. In fact, I taught last year at a really fun Russell Sage Behavioral Economics summer camp. It's a two-week camp for graduate students that I actually attended back in 2006 when I was a graduate student. It's an amazing rite of passage. About 20 people get handpicked from normally over 100 applicants to have this experience.

5:05Katy Milkman:And all the greats show up, and it's in a retreat. And last year, I got to teach in it, which was this amazing honor. and Richard Thaler, who's one of the founders of the field and won the Nobel Prize in 2017 in economics, criticized the folks who'd organized. He said, you know, one problem with behavioral economics right now is it's not engaging seriously with psychologists. Look around. You didn't even invite any psychologists to present. And they said, well, we invited Katie. And I was like, well, my PhD is in computer science and psychologists think I'm an economist. So you kind of missed the mark.

5:39Katy Milkman:Anyway, this is a very long answer, but it's a funny field, and the boundaries are a little bit blurry, and it's not clear. I think that's a good thing, but we need more folks who have a psych background talking to economists, and economists also probably need to come over and figure out what a real psychologist looks like.

5:58Misha:Yeah, yeah, that's interesting. Why do you feel like you're an outsider? Is it just the nature the work that you do? Or is it because you integrate more with psych?

6:07Katy Milkman:I mostly don't write papers that have formal models of human behavior in the introduction that build on a classic economic model and then change a parameter. And I'd say that expectation is that there's a theoretical model at the front end of any manuscript you write that leads into maybe an empirical exercise or maybe not, but that's really engaging seriously with the economics literature by saying, you know, we're going to take your standard models that you use to figure out various predictions, you know, whether they be game theoretic or whatever prediction you're trying to make, welfare predictions.

6:46Katy Milkman:I'm going to introduce a perturbation to your model that's based on psychological insight, and then I'm going to see if I can test my model empirically. Or maybe I don't. Maybe I just posit a model. That's sort of behavioral economics. And my work tends not to start with a model. It tends to start with a description of the world that I think is more accurate, perhaps, than the one we have. It might build on the idea that people are not perfectly rational. Then I go out and test what I think is true. But I'm not spending a lot of time writing a new theoretical model of human nature that's mathematical and including proofs in my papers.

7:23Katy Milkman:And that means essentially I'm not holding a calling card for behavioral economics. I'm also not a psychologist, right? A psychologist has a different calling card that they bring to the table of sort of like I'm deeply focused on mechanisms and understanding the origins of behavior and maybe engaging, you know, with even what's happening in the brain. I'm more interested in phenomena that deviate in surprising ways from our expectations of human nature. And so that phenomenological focus and policy focus make me a bit of an outsider when it comes to psychologists. So I'm more likely to run straight to doing a field experiment with the Department of Education, looking at 13 million borrowers and how we can change their behavior than I am, perhaps if I have an idea, to go into a lab environment and really prod what's the underlying reason for this.

8:17Katy Milkman:I do do some lab research and I do try to look at the why, but it's not the driving feature of my work. So that makes me a misfit in both places. But there is this middle ground. So hopefully to the listener who's like, yeah, I don't want to be writing mathematical models or in the lab. Where do I fit? The good news is you can be a misfit like me in the middle. There is a place for us.

8:39Misha:That's a really positive takeaway, actually. So I guess for that listener, maybe another thing they're wondering is like, how do you come to occupy this niche, this middle ground? And you said that you got your PhD in computer science, but maybe you can walk through like, you know, your undergrad and then on to grad school or really your trajectory to where you are now and your research interests.

9:03Katy Milkman:Yeah, sure. I'm happy to do that. And I will say I'm a bad role model if you're looking for a straight and narrow path to a clear final destination.

9:12Misha:I think that's the best role model, actually. Yeah, because a lot of people are probably in similar ways.

9:16Katy Milkman:Yeah, no, I actually agree. I think it's good to know that there are winding paths and there are ways that confusion can lead to good outcomes. But of course, if you know exactly what you want to do and just go directly there, that can be more efficient. But yeah, my path was not straight and narrow.

9:33Misha:Not linear.

9:34Katy Milkman:Not linear. So yeah, my undergraduate degree is from Princeton University, which is an amazing place to be an undergrad, much like Stanford is an amazing place to be an undergrad. I got there without a clear idea what I wanted to be when I grew up, like most 18-year-olds. And I knew that I really liked math. And I wanted to just study something applied because I liked real-world problems. And so that led me to think maybe I'll be an econ major. So I took econ 101 as a freshman and I just have to say it literally is the worst class I've ever taken. I hated it. I won't name the person who taught it because I don't want to shame anyone.

10:07Misha:What made it so bad?

10:09Katy Milkman:The person who taught has never been allowed to teach again. They were not invested in the undergraduate experience. They were writing a book at the same time and sort of not paying attention. And we'd get, we didn't have a textbook for the class. It was just really disorganized. This person literally came and read to us during class. Like instead of speaking to us, it was reading. And I would sit in the front row in the middle with coffee because I was like, I know this is really boring and I'm going to need every situational factor working for me staying awake and like paying attention. And still a couple of times I dozed off.

10:43Katy Milkman:And I mean, admittedly, like I should have been sleeping more overnight if I'm falling asleep in my classes. On the other hand, I didn't fall asleep in any other classes. So it was just very poorly taught.

10:52Misha:But against all odds, you you persevered. Well, I didn't major in economics.

10:57Katy Milkman:So the short yeah, the short story is like I was like, huh, cross that off the list. I'm not going to be an economist. This is not for me. I hate it. So then I had to figure out what to do. And I ended up having a roommate who was an engineer who was studying something called operations research and financial engineering, which is a unique department, a small department at Princeton that had been formed just a few years before I arrived. I didn't know that it was brand new. And it really focused on applied math, applying math and computer science to real world problems related to either financial markets or like transportation systems or, you know, logistics of any kind, learning things like optimization and coding and statistics and probability theory.

11:42Katy Milkman:And it was, I thought, really interesting. And I thought, OK, I didn't buy the assumption. One, by the way, it wasn't just that the class was poorly taught in econ. I also just didn't buy econ. Like the econ that was fed to me had no behavioral insights in it. It just said, look, people are perfectly rational. They never make mistakes. They optimize on everything. And based on those assumptions, we can figure out all these things about the world. And I was like, this is absurd. There is no human I've ever met who follows any of these assumptions. What a waste of time to sort of like mathematically model people in this ridiculous way and then trust the outputs.

12:16Katy Milkman:So it seemed like a total waste of quantitative effort to me. I obviously have a different perspective on it now, but I thought econ seemed like garbage for many reasons. And so operations research sounded pretty cool because I was going to get to use math tools and souped up skills of other kinds that were quantitative. But I wasn't going to have to make faulty assumptions, right? It would be like, oh, what's the distance between factory A and factory B? What's the cost of gas per mile? You know, like now we're going to optimize. So all of the assumptions seemed valid and I was going to get to apply my toolkit to interesting problems.

12:51Katy Milkman:So I was like, I'm in. Sign me up. I went to summer school in order to catch up on the required physics and chemistry. I should have been taking my freshman year if I'd known I was an engineer. And I transferred to the engineering school where I had to take a whole bunch of extra requirements. And I majored in operations research. And it was wonderful. It was a small new department. The faculty were super welcoming. There was a guy there who I literally went to his office hours who was the department chair, a wonderful statistician named Erhan Shinlar. We would talk about Turkish architecture and Taiki, the goddess of chance.

13:23Katy Milkman:And I just showed up every week for his office hours just to chat and learn. It was amazing. I started doing some independent research projects on things that have nothing to do with what I study now. But I was getting a taste for what it meant to be a professor by interacting with all these folks. And I just I was having a blast. So I started getting interested because of that immersion and that amazing program, which I'm still now I'm sort of on an advisory board and I've stayed very close with folks, even though I'm in a very different field. So I was interested in academia from that. The other thing that happened as an undergrad that was sort of amazing for me is I had to write a senior thesis.

14:00Katy Milkman:And Princeton requires that if you write a senior thesis, it has to connect the discipline of your major with the discipline of any. They're not called minors there. They'd be called minors elsewhere. They're called concentrate. Oh, wait, what are they? No, they're not. They're called certificate programs. Sorry, I'm getting all my lingo wrong 20 years out. So basically, you're minoring your major. And I was majoring in this sort of stats type operations research field, but minoring in American studies because I loved American literature and history and wanted to immerse myself in that during the time I wasn't spending learning math and computer science.

14:35Katy Milkman:So I had to figure out some way to bridge these things. So I came up with it. I spent a lot of time thinking, what am I going to do with this thesis? This feels like a big thing. And I ended up coming to the plan to study New Yorker fiction, which is like a major piece of American studies. You know, fiction about America, published in America by one of those prestigious magazines, to me was like classic American studies fodder. But then I was going to do a statistical analysis of it. So I read a decade of New Yorker fiction. I literally had to get the physical copies because nothing was digitized at that point.

15:09Katy Milkman:So I spent a ton of time with like bound copies of old magazines in the bottom of a library reading. I coded all these features of the literature. You know, now you just throw this in an AI engine. But I, as a human, read hundreds of stories and classified their features. Who was the author? What were their demographics? Was it written first person or third person? Where was the story set? Who was the protagonist and what were their demographics? What were the themes? So I did all this and then I did this big analysis of a couple of questions that I thought were really interesting. One primary question was, to what degree is fiction autobiographical?

15:43Katy Milkman:I'm going to analyze this statistically because we assume in our American Studies classes that we should really understand the life of the author before and when analyzing a work of fiction. But to what extent is that informative? I wanted to know how often do authors write about people who resemble them demographically versus people who differ from them? Also, did changes in the editorial staff really shift the fiction? Like, how much does it matter who's in a leadership role at a magazine that's this prestigious in terms of shaping the kind of work that then gets out there? And the big findings were editorial shifts at the very top, like the editor-in-chief mattered very little.

16:20Katy Milkman:Editorial shifts at the top of the fiction department mattered massively in terms of the types of authors and stories that are getting published thematically, demographically, etc. etc. And also authors wrote about characters who resembled them demographically at an extremely high rate. But women and minorities were more likely to write about characters who differed from them than white men, which I thought was also super interesting. So sort of dominant group didn't step out of their skin as often as non-dominant group members. So those were the kind of key findings, like maybe not rocket science, but actually people got really excited about this thesis.

16:51Katy Milkman:No one had really done this kind of analysis of fiction before. I didn't know I sort of like invented something this new, but I had. And now people do this kind of thing all the time, but it really didn't exist that field to much degree. There's actually one professor at Stanford, Franco Moretti. I remember him who did do this kind of work and was like, want to come be my grad student? Anyway, the work got all this attention the day I graduated. The New York Times published a story on the cover of the arts section about my thesis, like with a huge picture of me holding a bunch of New Yorker magazines.

Read the full transcript

17:21Katy Milkman:And I got all this feedback, all these people wanting to talk to me. I'm like so excited about this. And it was this bizarre experience of being like, wait a minute, when I'm deeply intellectually curious about a research question, like the world might care too. And they might be really interested in what I have to say. Like, wow, this is really cool. Maybe I should do this with the rest of my life. So some combination of all those experiences led me to go to grad school. My PhD was in computer science and business at Harvard. The program there seemed neat to me because I was graduating from college at the beginning, like during the internet boom of the early 2000s.

17:57Katy Milkman:And I was like, wait, technology is going to change everything. We're going to have all this interesting data on consumer behavior and like digital signatures of people's decisions. I didn't know I was interested in decision making, but fundamentally that was at the heart of this. And I was like, digital business, it's going to be really neat. All this data, I should analyze it. So I got to that program. I had never heard of behavioral economics. I had to take a microeconomic sequence at graduate school in order to just progress in my program. And that's when I discovered this whole field, everything that Kahneman and Tversky had done, Thaler had done, and so on.

18:30Katy Milkman:And I was like, wow, this is so interesting. And I convinced the mentors who were sort of in charge of my program, this is interesting. It's related enough to information technology developments and can be applied. You know, you can use it to analyze internet data and consumer behavior. So go ahead. They were very flexible. I was very lucky to be in a very flexible program. And that's sort of when I found my way. I started wandering into faculty offices who studied this stuff. And they were like, we've never heard of your PhD program. You're kind of weird, but sure, we'll let you hang out and you can do what you want to do.

19:05Katy Milkman:And so it's not the most circuitous route that anyone's ever taken, I think, to a field. But I was definitely an outsider and somehow magically was lucky to have these people be welcoming to me. And then I've been doing this kind of research at the intersection between psychology and economics ever since and had amazing mentors in grad school. Most importantly, Max Bazerman, who is my PhD dissertation committee chair, but also some other amazing folks like David Leibson. Well, Nathan was very pivotal as well. Kathleen McGinn and a computer scientist named David Parks, who was a big supporter of interdisciplinary research.

19:42Katy Milkman:So I got lucky. And then somehow I got this job at Wharton and the rest has been just really great fun with a lot of constraints removed from my path by funding sources and support.

19:55Misha:The New Yorker story is so cool, and I'm definitely going to check it out right after this. I could totally see how it got you interested in research.

20:02Katy Milkman:experience.

20:04Misha:Yeah, yeah. Actually, do you want to walk a little bit through like post-graduation and then how you got the job at Wharton and then how you kept going down that road if you did or changed it if you changed? And now there's, of course, behavioral change for good, which we can get into. Yeah, sure. But yeah, maybe to continue with the story a bit.

20:26Katy Milkman:Yeah, my happy ending doesn't just end with getting an assistant professor job. It's true. This career is long. There are many twists and turns. Yeah, it's really the beginning. And it really was. It does feel still like such a stroke of luck that my job market was challenging because no one knew where to fit me. And I did a lot of interviews without getting jobs. Lots of rejection before finally somehow magically this amazing place, the Wharton School, made me an offer as an assistant professor to join a very peculiar interdisciplinary group. I'm in a group called Operations, Information, and Decisions, which you might note actually is exactly the right name for someone who did their undergraduate degree in operations research, their Ph.D.

21:05Katy Milkman:in a program that was actually called Information Technology and Management, and then studies decision making. So I found my, like, unique special place, and somehow they picked me. I showed up, and I will say I still didn't really know what I was doing. I knew I loved behavioral economics. I knew I was fascinated by how people make decisions. I knew I cared about studying it in the field, like in the wild, which was also what my New Yorker project had been about. My dissertation work used data, well, actually initially used data from Netflix, and then they took it back. So I learned a lot about working with corporate partners and how they can be cruel.

21:40Katy Milkman:And then I had to go do my dissertation basically again. I had to go find a Netflix knockoff company called QuickFlix in Australia that had the same business model. And they gave me a much smaller data set to answer the same research question. And I thankfully replicated my original findings from Netflix and Quick Flicks and that I was able to publish. The study was looking at whether or not people this is such dinosaur stuff because it's like, oh, the business model literally doesn't exist. That used to exist. But they used to send you physical DVDs in the mail. You'd make a list. You'd order a list and you'd say, like, here are the next 20 movies I want you to send me.

22:15Katy Milkman:They'd send you the top three. When you finished one, you popped it in a prepaid mailer and they sent you the next one on your list. And what I was interested in was procrastination around watching the movies that we know we should watch that are good for us versus the ones that we want to watch because they're instantly gratifying. So think like the distinction between documentaries and summer blockbusters, right? Highbrow versus lowbrow. And my prediction was that people would say when thinking about the future and making these lists, oh, I'm going to watch lots of documentaries, lots of highbrow movies.

22:45Katy Milkman:But when then the movie came and arrived in their home and they had to decide what to pop in their TV that night, they would procrastinate. They would be like, actually, I'm going to watch the summer blockbuster and I'll hold off on the documentary. That's like future me. And we found this pattern very clearly that people procrastinated on watching highbrow movies. So they would order them. They might order the highbrow movie before the lowbrow, but they're much more likely to return them out of rental order when there's a highbrow preceding a lowbrow than vice versa and holding times on average are longer for these highbrow movies.

23:15Katy Milkman:So it was just a way of studying present bias, procrastination, tendencies, the fact that people have preference reversals between now and later and doing it with this digital data set. So that's what I'd done. You know, that was the job talk I gave. It was the centerpiece of my dissertation. And it was fun. It scratched an itch. It was like, oh, this is cool. Sort of like the New Yorker work. But one thing that was really missing and that has become a very important part of my career over the last decade and was an evolution was to realize that what I really wanted to do with my career was actually research that was not just curious, interesting, fun to talk about at cocktail parties, but that might actually make the world a better place.

23:52Katy Milkman:And I sort of didn't figure out how to do that and that that was something important to me until a few years into my faculty job at Penn, when I was still really, you know, in explore mode. So computer scientists talk a lot about explore exploit, right? There's this tradeoff, like when you're trying to figure out a path. And by the way, as an undergrad or, you know, even as a grad student, you should really be in explore mode, like try a lot of things, take the econ class. And by the way, even if it's poorly taught, maybe take a second one and, you know, take the American literature class and check out computer science.

24:24Katy Milkman:Go take a class in mechanical engineering and something about, you know, linguistics. Really explore widely. And then eventually you convert to exploit. Once you find what you're really good at, what you're really passionate about, you exploit. And you just do that. I was still very much in explorer mode as an assistant professor. So one of the things I explored was we have an amazing medical school at the University of Pennsylvania. There's a lot of amazing medical schools. We have, you know, one of the top five in the world by pretty much any measure. And not only that, but we have a very compact campus.

24:53Katy Milkman:So it's not far away. When I was at Harvard, we had an amazing medical school, but you had to basically drive to get from where I was studying on main campus to any of the people at the medical school. But at Penn, it's a 10-minute walk to the medical school. And there was a group of people there studying decision-making and really world-class people. Kevin Volpe is a really huge name in studying behavioral economics and health decision making, and he was leading this up-and-coming growing group of doctors, many with PhDs, some without, who were trying to figure out how do we apply the insights from behavioral economics to extend life expectancy and to improve people's health outcomes.

25:30Katy Milkman:So I started showing up at research group meetings and seminars that he was organizing just to explore. I still remember one day a woman giving this presentation about health and healthy eating and health habits, and she put up a pie chart that totally changed my life. The pie chart broke down the percentage of premature deaths that are estimated to be due to various causes in the United States. And it had, you know, wedges for things that I think of as obviously going to be big causes of death, like, you know, genetic factors, environmental factors, accidents, all the stuff you'd think like, why do people die?

26:07Katy Milkman:Well, these are the reasons. It also had a wedge for decisions people make that could be changed. So these are decisions about things like buckling your seatbelt, taking your medications, getting preventative screening, eating healthy, exercising, do you drink, do you smoke? That wedge was the biggest of all, the decisions in terms of predicting premature death or causing premature death. It was about 50 percent, 40 to 50 percent of our deaths in this country are due to behaviors we could change. And why that changed my life is because I just didn't understand until that moment what the massive opportunity was that existed if we could figure out how to harness the insights we're generating from behavioral science, behavioral economics about human behavior to change people's trajectories, that the health impact could be so massive.

26:58Katy Milkman:And of course, it wasn't just that I was really interested in health. What that also made clear to me was almost certainly there are other behavioral patterns that are similarly snowballing and matter more than I appreciated. These are not sort of like 2 % effects the way I was thinking of them. But almost certainly if we could change people's daily decisions about, you know, how much they focus in school, whether they show up for their test, whether they study, that would have massive implications. If we could help people think differently about financial decision making, that could have massive implications, probably much bigger than I'd appreciated.

27:31Katy Milkman:So that realization and that moment really changed what I focused on. And I decided, OK, I'm going to use this skill set I've built as a researcher, not just to scratch intellectual itches about what movies do people run and how fast do they return them and what New Yorker fiction articles focus on. But let's try to save some lives and make lives better. Like, can we let's make the world a lot better with behavioral science. And that's when I really started shifting the direction of my work towards trying to figure out, can we do behavior change work for good? And around that time, I also met Angela Duckworth, who is a brilliant psychologist at the University of Pennsylvania, who is also super interested.

28:14Katy Milkman:She's done a lot of work on grit and understanding what predicts success, but she was just getting interested in, hey, I don't want to just understand traits that predict success. I want to know how to help people develop success and how do we create long-term positive change in their lives. Like it's not enough to just know what predicts good outcomes. Let's create good outcomes. And so that very much gelled with where my motivations lay and we work really well together and started teaming up on a variety of things. And eventually our partnership blossomed. I'm happy to tell the story of how we built an initiative called Behavior Change for Good Initiative.

28:49Misha:Yeah, yeah, that'd be great if you want. Okay.

28:52Katy Milkman:Yeah, no, I'm like, I don't know if your listeners are interested in this level of detail.

28:56Misha:No, I think that it's just super interesting. All right.

28:58Katy Milkman:I will keep telling you the narrative of a professor's adventures. So this is a funny story in a sense because it's the story of how a failure created a success. In 2016, for the first time in the history of the MacArthur Foundation, which you're probably familiar with if you've heard of like MacArthur Genius Awards, which, by the way, Angela has won. Lots of other amazing people we know have them, too. But anyway, it's an amazing foundation that does all sorts of great research support of work, including the Genius Grants that they give out each year. And for the first time in their history, they said, we're going to dig into our endowment and we're going to make$100 million investment in a team that's poised to change the world in a positive way using whatever methods they propose.

29:46Katy Milkman:And we're going to have a massive global competition for who gets our$100 million to change the world for better. They hoped as many people would apply as possible, but they wouldn't accept more than one applicant per institution, meaning Harvard gets to put forward one team. Penn gets to put forward one team. University of Mississippi gets to put forward one team. Sesame Street gets to put forward one team. So the University of Pennsylvania, where Angela and I both worked, thought this sounds really cool and we'd like to be it's going to get a lot of media attention. It's a really cool opportunity to talk about big science and big ideas.

30:23And we want to be really in there in the competition. So our president, Amy Gutman, at the time announced an internal competition for the one team that would advance and said, we really want to see great things.

30:36Katy Milkman:And we know that it's a lot of work for basically a lottery ticket that has very low probability of payoff. So we're going to support the teams that do best in our internal competition to some degree to make their visions come to fruition, even if they don't win$100 million. We'll give you some startup funds. So it's not just a lottery ticket. There's a little bit more behind it. So Angela and I literally were sitting together working on something at the time this email from the president came through. And we looked at each other and we said, let's put our name in that. Why don't we propose doing something really ambitious toward the goal we're already interested in working towards?

31:09Katy Milkman:Like, what would we spend$100 million on if we wanted to make massive radical advances in the science of positive long-term behavior change? What would that look like? How do you spend$100 million productively to accelerate this field that we think should be getting a lot of attention and energy? And the idea we came up with was we need to do massive team science, just the way that in sort of the hard sciences, if you want to build a particle accelerator or a space telescope, you bring in huge teams to really put all their heads together and energy together and work collaboratively. We need massive team science, too, if we're going to solve big social problems.

31:52Katy Milkman:What the formulation we came to look like was we're going to bring together a team of a bunch of amazing minds from across disciplines. So think economics, psychology, sociology, computer science, law, medicine. Who haven't I named?

32:08Misha:I'm sure there's other fields that we were bringing to the table, marketing, management, et cetera.

32:13Katy Milkman:bring all the people who are interested in behavior change from all these fields together and have them team up. And then the second thing was, but on what? And we cooked up this idea, which we now call the mega study of let's have them run, let's run tournament based science. So we partner with large organizations, say, you know, CVS Pharmacy, right, which is like a massive,

32:36Misha:one of the top, I think, 10 Fortune 500 companies and has a major health mission.

32:41Katy Milkman:and we'd figure out some target behavior, say getting a vaccine or, you know, taking medications regularly for your diabetes, whatever it might be. Let's have all the scientists come up with their best idea for how to target that behavior change outcome and then run a tournament where all of CVS's customers would be randomly assigned to different programming to target this outcome based on what different teams of scientists thought would work. One team says we're going to get medication adherence up by paying people in this in the following way. Another team says we got to change mindsets and we're going to show them programming of this type, you know, videos and messaging that really changes their mindset about medication adherence, right?

33:21Katy Milkman:All these teams come up with their best ideas and we test it. It's a tournament, everything head to head. We randomly assign people to get these different versions and we see what works, what increases medication adherence most or whichever outcome of interest. We say we do this on lots of different outcomes. We should study this in a lot of different areas is to start having insights gel. So not only would we want to target medication adherence and say vaccine adherence, but let's try to figure out what promotes exercise and what promotes savings and how do we keep kids in school and how do we get them to do more math homework?

33:52Katy Milkman:And hopefully some insights would bubble out of these things that could cross-pollinate. And so that was a proposal we put together. We put together a team. We put together this idea that we should run these mega studies. That's what we call them, these massive collaborative tournament science endeavors where you can make apples to apples comparisons about the cost effectiveness and overall effectiveness of different approaches to solving a problem. And let's see what happens. And so the short story of the MacArthur competition is we made it to the semifinals, I think. Maybe the quarterfinals.

34:25Katy Milkman:I don't know. We made it a round or two of elimination before we got kicked out. We didn't make it to the round of eight, but we made it to the round before that. There were thousands of teams. So we felt, you know, it was respectable. Hopefully we didn't make the University of Pennsylvania look too bad. I think the video we submitted got some prominent placement on the MacArthur website. So that was good. Like, OK, we got some attention, but we didn't win the competition. We didn't get$100 million, which, by the way, is also probably good because we were just starting on this journey. And maybe we could have spent$10 million responsibly, but$100 million at that moment, probably not.

34:58Although, interestingly, the team that won was Sesame Street to design programming to try to improve the lives of Syrian refugee children.

35:06Katy Milkman:So a pretty wonderful and really different mission. But interestingly, ironically, Angela was later contacted by Sesame Street with the question from the organizers. They said, we're really trying to figure out how to change behavior in these kids and these populations that we're targeting with our entertainment. Could you help? And we thought that was funny because we were like, yeah, that is the ultimate question. How do you change behavior? That's what we were proposing. But you can't argue with giving Sesame Street funding. It's a very different goal than advancing science and behavior change.

35:36Katy Milkman:Great organization. But we were lucky that even though we lost, Penn had committed to give us some startup funds. So we did end up getting a little infusion from Penn. And then we built the thing on more of a shoestring budget. And slowly through various grants we've been able to obtain over the years and donations from Penn alumni, we've been able to create an organization that's now at the eight-year mark. We were founded in 2017. We now have over 180 affiliated scientists across disciplines. We've held a couple of big convenings. We've run seven megastudies, published multiple papers in Nature, which is the world's oldest prestigious scientific journal by many measures.

36:22Katy Milkman:I don't know, maybe you'd argue Science Magazine. I think, whatever, it's neck and neck. and change the way people think about, I think, what's the best approach to accelerating insights that are policy-relevant to target a specific outcome. The mega study is sort of clearly a better approach than the one-at-a-time way we've been doing science before to solve a policy challenge, like how do we get people to save more or how do we get people to get vaccinated? And it's been really exciting to see not only that this method we developed as a result of that MacArthur grant, you know, landing in top journals and that we've honed it and it's become a thing, but it's been more exciting to see its wider adoption.

37:10Katy Milkman:And there's lots of mega studies and development right now that we are not leading. And we love that. There was one that came out in Science last fall focused on enhancing democratic norms. And there's one that's being launched shortly that Angela and I are both advising, but we're not running, related to trying to figure out how do we improve happiness. There's been work on pro-environmental behaviors. And it just keeps growing, the number of people who are launching these things, and we're really pumped about it.

37:43Misha:this is just the coolest thing to me genuinely not only is it impactful but just such a cool method i think at least personally and hopefully the viewers listeners also think so it's kind of we're very excited about it too yeah i want to give you a chance to shout out the work that you've been doing because i think it's so cool so i know you talked about cbs but i know you've also done work with the gyms and correct me if I'm wrong, but freshmen following freshmen through college and Walmart. So if you want to talk a little bit about that.

38:17Katy Milkman:Yeah. These mega studies have been our bread and butter for the last eight years. The first one was with 24-hour fitness gyms. We built a 28-day program to try to help people build healthy habits around exercise and tested 53 different programs designed by different team scientists to test about 20 different hypotheses about what works. So even though there were 53 programs, we sort of have little sub-studies embedded within the mega-study. So scientists might design a treatment condition and a control condition that are paired to test a very specific hypothesis about the way to communicate with someone, for instance, or incentivize someone.

39:00Katy Milkman:So we have these sub-studies all embedded in the mega-study. And we launched that and we tested what actually changed the rate at which people went to the gym. We had about 60 over 60 ,000 participants in this first mega study. And we found some things that really were beneficial and at low cost. We deliberately made the project constrained so that nothing cost more than like a dollar or two a week to deploy. And most things cost less than a dollar over a month because we were looking for low cost, scalable interventions that could be used to try to improve population health and increase exercise.

39:42Katy Milkman:Some insights that came out of that that I still think about. The best performer, though, I should note there's a statistical tie really at the top of the heap. So we more identified a bunch of promising things worthy of more testing, which couldn't necessarily be differentiated from one another in terms of their exact performance. The top of the tie was an intervention that emphasized not missing more than one planned workout. So most of the interventions involved some implementation intentions type component, which is a very jargony term for making a plan. So an implementation intention means like, when will I do it?

40:19Katy Milkman:Where will I do it? How will I get there? Those are the kind of details you lay out. It's an if-then plan for getting something done. So most of these interventions, they had people say like, which days of the week and at what time do you plan to go to the gym? And we'll send you reminders. That's the beginning. And then the program that was so effective said, if you miss one of your planned workouts, we want to give you a little extra reason to come the next time so you don't ever have two misses in a row. So never let one miss become a streak of misses, if you will, or more than a one-off thing.

40:51Katy Milkman:So normally you got some points which were convertible for about 20 cents in Amazon rewards, Amazon cash, for going to the gym. if you missed a day and then came back so you'd get 30 cents for going to the gym instead of 20 cents so not enough that you'd strategically miss a workout in order to earn that magic extra 10 cents but enough that it's psychologically a little different and we're talking to you about it differently don't let this miss accumulate come back there's a little extra sweetener on top that performed really well that actually performed as well as super charging incentives and paying people more like two bucks a gym visit.

41:32Katy Milkman:So I thought that was super interesting because it highlighted that we really want to prevent multiple misses. We don't want people to feel like they've totally fallen off the wagon when they have a blip. We want to help recover if we want to build long-term healthy habits. So that was one insight that I took away from that project. Another really well-performing arm was designed by the great Bob Cialdini, like the godfather of influence, original field researcher, one of my, I think, the people who most has inspired me. And he developed an intervention that focused on just giving people the information that there's a growing number of Americans who are interested in exercising and are exercising more.

42:16Katy Milkman:So there's this growing trend. And there'd been some research done, actually some at Stanford, also some at Arizona State, by different teams of researchers showing that not only is it important to say, hey, this is the normal thing. Lots of people are doing something. But actually, when you can see a growing trend, even if it's a minority of people, but you can see that fad building, people want to follow along. And so we had a majority, we had this condition where we said the majority of people are doing it and the numbers are climbing. And that was very compelling. Of course, we didn't lie.

42:46Katy Milkman:You know, we use real statistics and we're truthfully telling them about growing trends and exercise. But that was also. So anyway, that's an example of one of our projects. We've done a bunch more since. We did a bunch of work during COVID-19. In the fall of 2020, before the vaccines were developed, we partnered with both Walmart Pharmacy and two large healthcare systems, Penn Medicine and Geisinger Health, to test ways of communicating about the flu vaccine and measure actual vaccination decisions in order to come up with what we thought the best candidates were for how we should communicate with people to try to encourage COVID vaccine take-up.

43:22Katy Milkman:So we ended up identifying a really effective strategy was to use ownership language. And I should nod to John Bogard from WashU and his team of UCLA scholars Noah Goldstein and Craig Fox, who came up with this insight about ownership language saying, we have a vaccine reserved for you or waiting for you. And by making it feel like it belonged to you, it felt exclusive. It felt like something you didn't want someone else to get because it was mine. Felt recommended. And so there's probably a whole lot of psychology sort of bundled together there. But that was more effective than a lot of things that we thought sounded clever.

44:00Katy Milkman:And then after we did these tests where we validated in multiple settings that that seemed to be a really effective way. And by the way, also multiple reminders. Like the more we nag you, the better. multiple teams, including one of my former students, Hank Chen Dai, and Sylvia Sicardo, went out and tested with COVID vaccines to make sure that that same language was useful there. And indeed, that seemed to outperform standard messaging. So we were really excited to be able to contribute in a timely manner to that important policy issue. And we've done work not just on vaccines at this point, but on increasing savings, encouraging students to pursue or to complete more math programming by communicating with their teachers with Zurn Math.

44:41Katy Milkman:And we are launching right now, we're in the middle, we have about 36 ,000 students who've gone through it already, and we'll hopefully have 100 ,000 by the end of the summer. A mega study in partnership with 30 universities around the United States that is designed to basically provide digital orientation programming as well as support throughout the fall term to test what works to help students stay enrolled and improve their grades if we're trying to use the best science available. And we're testing some things that past research suggests should work, as well as some new ideas, ranging from things like paying kids to meditate regularly to providing them with an AI chatbot coach to help support their social-emotional needs.

45:26Katy Milkman:It's a really wide range of different things that we're testing. And so that's super exciting. I think that's the most involved mega study we've run yet with potentially really large implications if we can find things that at low cost and that are easy to scale improve these outcomes and we'll see if anything

45:43Misha:works yeah i mean i'll be really excited to see it all happen okay so i do i do want to wrap up in a bit here but before doing so maybe is there anything that you want to talk about So in the past, I've asked people if there's a piece of advice that they have, either for grad students, professors, undergrads, etc., etc., or something we could talk about as future directions for you that you're hoping for, or maybe for the field more generally. Is there anything that kind of speaks to you? Sure.

46:17Katy Milkman:I could say a couple things to those broad themes. One thing I'd say is one of the most rewarding parts of this career and something I hope everyone listening will invest in and look forward to is mentorship. It turns out that being a mentor to young scholars is one of the ways you grow the most as a scholar yourself. And in fact, there's some really wonderful work by Lauren S. Chris Winkler at Northwestern University that I got to be involved in a little bit of showing that when you coach or teach someone else on something, a skill that you want to build, it actually helps you grow in part by growing your confidence, but also growing your competence.

46:58Katy Milkman:mentoring has been just such an amazing part of my career it's such a gift I've gotten to mentor a bunch of different PhD students at this point I mentioned Hengchen Dai who's a professor at UCLA and her work on COVID-19 vaccines that's so incredible I have students you know Erica Kyrgios at UChicago Edward Chang at Harvard and Yish Rai at University of Maryland and a bunch of of wonderful PhD students I'm mentoring currently. And just say that the chance to work with these young scholars and help them build careers and figure out, you know, what does great research look like and how do they take their insights to the field and have impact, there's really nothing more satisfying.

47:42Katy Milkman:And it's exciting, you know, it's been really exciting to build the Behavior Change for Good initiative and test my own ideas and so on. But I think a really wonderful thing about this career is that opportunity that you're constantly coaching and mentoring the next generation. And it builds these amazing relationships. Like these are some of the people who mean the most to me in the world. I feel the same way about my mentor and I'm like tearing up and I'm still really close to my mentor. And also I call them my academic siblings. The other people who were in my research group at the same time I was are still a huge part of my life.

48:17Katy Milkman:And so I think that's a really special thing about academia, especially when it goes well. Obviously, there's relationships that can go poorly and that's its own mess. But tell the listeners, I would say, you know, look for great mentors, people who have a track record of building strong relationships and being good mentors. And it's not always it's not always like perfect compatibility and social interests and so on. Sometimes that's not there. But someone who's supportive and caring and excited to be your mentor and to invest time and effort and energy in you. And I also just want to encourage when you're on the other end of it to realize what a gift it is that it's not a responsibility, but a gift to get to mentor and coach the next generation.

48:56Katy Milkman:And there's really nothing more rewarding in a career like this than having that opportunity.

49:03Misha:Yeah, I couldn't agree more. I've had some really amazing mentors that have made it possible to be where I am today.

49:09Katy Milkman:I'm so glad. I'm glad you've had great mentors.

49:12Misha:This has just been a wonderful conversation. So thank you.

49:17Katy Milkman:Thanks for having me. I really appreciated it. Enjoyed the conversation and hope your listeners will find it useful as they're charting their own paths.

49:26Misha:Thanks for listening. Following Robert Cialdini's advice on this podcast, let's see if I can convince you to take about five seconds of your time and leave us a review on Spotify, Apple Podcasts, or wherever you're listening to this on. This podcast has been a labor of love by several wonderful young folks here in the department. and we've been surprised by the ever-increasing reach the podcast has had. Help us make even more people excited about psych by leaving us a review or subscribing to our no-spam, all-fun sub-stack at stanfordpsychpod or shoot us an email with your thoughts or suggestions at stanfordpsychpodcast at gmail.com.

50:13Misha:Thank you so much and have a wonderful psyched day. Thank you.

From the publisher

This week, Misha chats with Katy Milkman, the James G. Dinan Professor at The Wharton School of the University of Pennsylvania. A Fellow of the Association for Psychological Science and former president of the Society for Judgment and Decision Making, her research explores how insights from economics and psychology can be harnessed to change consequential behaviors for good. Her work, published in journals like Nature and PNAS, has been recognized by Thinkers50 as among the world’s most influential in management thinking.

In this episode, they discuss Katy’s influential work designing “megastudies” to generate new insights about behavior change, as well as lessons from her bestselling book, How to Change. Katy also shares her perspective on translating scientific findings for a broad audience and the vital role of mentorship in academia.

If you found this episode interesting, subscribe to our Substack and consider leaving us a good rating! It just takes a second, but it will allow us to reach more people and excite them about psychology.

Links:
Katy's book: How to Change
Katy's Website: Link
Choiceology Podcast: Link
Behavior Change for Good Initiative: Link

Misha’s website: Link

Podcast Twitter: @StanfordPsyPod
Podcast Bluesky: @stanfordpsypod.bsky.social
Podcast Substack: https://stanfordpsypod.substack.com/

Let us know what you think of this episode or the podcast! :) stanfordpsychpodcast@gmail.com

More from Stanford Psychology Podcast

All 29 episodes
156 - Katy Milkman: The Art and Science of Lasting Behavior ChangeStanford Psychology Podcast · 51 min
Listen in VO