Katy Milkman on How to Change
Everyone, we assume, wants to be their best person. Few of us, perhaps, none, hits all their marks in this pursuit even if the way toward the goal is generally apparent. If you want to know how to do a better job hitting those marks, whether its walking 10,000 steps, learning Esperanto, or quitting smoking, a good person to consult would be Katy Milkman. Working at the nexus of economics and psychology, Milkman – the James G. Dinan Professor at The Wharton School of the University of Pennsylvania and co-founder of the Behavior Change for Good Initiative at Penn – studies the almost alchemical process of turning good intentions into solid actions.
In this Social Science Bites podcast, she details for interviewer David Edmonds some of the biases and some of the critical thinking processes that both define and then overcome the obstacles to changing our behavior. These range from concepts with such academic names as present bias and temptation bundling to the more colloquial ‘what the hell effect’ and its antidote, the emergency reserve. But the point of her research – especially as it gets translated to the public through her podcast Choiceology or her 2021 book How to Change: The Science of Getting from Where You Are to Where You Want to Be – is to find practical ways to change yourself.
For example, she explains that “it’s important for goals to be measurable and achievable, although they should be a stretch. You know, if your goal is ‘exercise more,’ how can you measure that? How could you even set a commitment device, for instance? … It’s also important to have a plan of, sort of, when will I do it, where will I do it, how will I get there. These are called “implementation intentions.” I think the most important part of them is they associate a cue with the action. So just like an actor needs a cue to know when to say their lines, we need to not forget to take action on our goals.”
Her influence in turn is felt practically. Choiceology, for example, is sponsored by the brokerage house Charles Schwab, and Milkman has been a consultant for organizations ranging from the U.S. government and Walmart to 24 Hour Fitness and the American Red Cross. She is a former president of the Society for Judgment and Decision Making and a fellow of the Association for Psychological Science.
To download an MP3 of this podcast, right-click this link and save. The transcript of the conversation appears below.
Before we begin this episode, we’d love to hear your feedback on the Social Science Bites series. Please let us know your thoughts on Social Science Bites by taking our short survey, and you’ll be entered to win one of five free copies of the Social Science Bites book, Understanding Humans.
David Edmonds: You want to exercise more. Simple, just do it. Unfortunately, it turns out that changing our behavior – even when we want to – is not straightforward. There are many psychological obstacles to change. But Katy milkman can help you. Professor Milkman teaches at the Wharton School of the University of Pennsylvania and presents the podcast Choiceology. She is also the author of How to Change.
Katy Milkman, welcome to Social Science Bites.
Katy Milkman: Thank you so much for having me. I’m excited to be here.
Edmonds: Today we’re talking about change. How to change, if we want to. Perhaps you can explain what forms of behavior we’re talking about. What sorts of things do people typically want to change in their lives?
Milkman: Yeah, well, for this, it’s, I think, easiest to think about new year’s resolutions. That’s a moment when a lot of people are thinking about this kind of change, trying to improve some aspect of their personal, their work, their spiritual life, their finances, right? Maybe people are thinking, I want to save more, or I want to run a marathon and get in shape, or maybe they have professional goals that have been unachieved. They want to climb the corporate ladder or become a serious academic and figure out how to complete a dissertation. So, whatever those goals are, they’re all aspirational goals of changing some aspect of your life through greater achievement.
Edmonds: So obviously there are lots of obstacles to change. Some of these obstacles are external; they’re things I can do nothing about. There might be laws against what I want to do. But what you’re talking about is understanding internal obstacles, psychological obstacles. These are the ones you’re mainly focused on, obstacles in the brain, in the mind.
Milkman: That’s right, yeah, there can be many, many external obstacles to change. Maybe you don’t have the financial wherewithal to make the changes that you dream about, or maybe your physical health is an obstacle. And I don’t study those obstacles. They’re very important, but they’re outside of my area of expertise. What I study is, even if you have everything lined up magically, which is very hard to do of course, in your external situation, there are still things inside you that can be barriers to change. A lot of people who have all of the resources you could imagine, still struggle to stay fit or to save sufficiently or to achieve their goals at work or their relationship goals. So, it’s really about understanding what can get in your way that’s inside you, and how can science help you overcome those barriers.
Edmonds: When I think of those psychological barriers, I’m principally thinking of weakness of the will. Part of my brain wants to diet. Another part loves chocolate. Is that the main mechanism preventing us from achieving our goals, or there are others?
Milkman: I don’t call it ‘weakness of the will,’ I’d call it ‘present bias,’ but that’s certainly a big one. Or you could call it impulsivity, and this is this very well studied challenge we have where we have long-term objectives and then we have short-term impulses, and they’re not often aligned. And specifically when it comes to thinking about the way our brain works, what present bias means is I overweight instant gratification over a long-term reward. So if I’m choosing between something that will be good for me now and something that will be good for me later, you might think that the balance would be even. I’d sort of weight those evenly. Or you might think that the balance would be tipped in one direction or other. Maybe I overweight the future because the future looms large. But that’s not what happens at all – I dramatically overweight the present. It looms much larger. It’s called present bias, and this can be a major challenge. You know, tomorrow I’ll go to the gym. Tomorrow I will get my finances in order. Tomorrow I will get back to all of the folks waiting for email replies and so on.
Edmonds: So that’s present bias. But that’s not the only obstacle.
Milkman: That’s right. Present bias is just one of many obstacles that can get in our way. Some of the others that I think are important are forgetfulness, which, of course, we tend to forget how much this can get in our way, but if something is not top of mind, we’re very unlikely ever to get around to it. So we have to find ways to keep things high on our level of attention. Habits are really critically important. I think of this barrier as having to do with our focus on efficiency. We look for shortcuts and sort of what’s the easiest way out.
In my book, I call this laziness, but I think some people take that as an insult, and so I’m describing it in a little more academic terms here. In general, our operating system is designed to prefer the easiest, shortest path to our goals, and that means we fall back on habits when there’s a little bit of friction that gets in our way, we are easily deterred, and so we have to overcome that feature of the way our mind works.
Some other barriers that aren’t always present, but can be present depending on the circumstances, are things like confidence. If you don’t believe you can achieve your goals, then it’s going to be very difficult for you to actually put your mind to it and put your effort towards goal achievement. And there are situations where confidence can be an obstacle. There are also plenty of situations where people have plenty of confidence and they still don’t succeed. But confidence is an important barrier to be on the lookout for in certain cases, and it can be particularly detrimental to groups that have faced negative stereotypes about their abilities to achieve.
Another barrier that bridges the internal and the external, but that social psychologists have thought a lot about, and so have behavioral economists, and so I think about it too, is social support. Do you have sort of the right people around you? And this is because we tend to conform to the behaviors of those around us, and they also signal to us what’s possible. And so, I think of it as an internal and an external barrier. It’s external in that you have to structure your social supports deliberately if you want to succeed, but it’s also internal because it reshapes the way we see ourselves and what we believe is possible.
Edmonds: OK, so let’s look at what might work to change behavior. This podcast will be posted in February, but we’re recording this conversation in January. You’ve already touched on this January the first is a date when people promise themselves that they’ll change. They have these new year resolutions, and it seems a rather arbitrary point, really. Why January the first and not January the second, or January the third? Or if it’s not New Year’s Day, people might choose a birthday from the age of 40, they’re going to give up smoking. They’re going to give up eating chocolate after supper. Do these kinds of commitments to oneself with these relatively arbitrary starting points, do they work?
Milkman: Often no. Absolutely there’s no magic in starting your commitment on January 1st or on any other special day. What I find interesting, and have studied about these dates is that they do propel many more of us to begin to pursue a goal than would on an arbitrary date. You still need the same set of strategies in place to overcome all the barriers to change, whether you start on January 1st or your birthday or an arbitrary date. So that’s the problem. And many, many people who set resolutions around a special date just think that that date and the motivation they feel in that moment will be sufficient to propel them to success. And that is a huge mistake, because we actually need to set up structures and systems in order to be successful.
I’d love to say a little bit about the psychology of fresh starts. This is one of the most interesting things I have studied, which is why it is that there are these moments in life when we are more motivated to set goals, and we see many more people setting goals and beginning goal pursuit. Whether they succeed isn’t, like I said, as sort of another matter, that’s where all the systems come in. And choosing a date doesn’t do anything beyond get you started, but we’ve studied this. This is work with Hengchen Dai of UCLA and Jason Reese, a former Wharton colleague. We looked at a series of moments. It’s not just New Year’s, as you alluded to, but any moment that feels like a new beginning in our life. So, this includes birthdays, it includes Mondays, it includes the start of a new month. For students, it’s the start of a new academic semester, or coming back from a school break. All of these dates, they give us the sense that we have a chapter break. And the way we think about our lives, the whole literature on autobiographical memory is not completely linear. We think about ourselves like we’re characters in a book. And these bookends that give us the sense that one era is coming to an end and a new one is beginning are very meaningful to us psychologically. They give us the sense of a disconnect. So on January 1st, we can feel more separated from who we were last year then we would feel that disconnect if it were January 5th, and we were looking back on January 4th, third, second and first. This disconnect actually ends up being really important. The reason it’s important is if we have had failures in the past, say, last year, you meant to quit smoking, but you didn’t manage to. Or last year you meant to get in shape, but you didn’t pull it off, you can say to yourself on January 1, “Well, that was the old me, but this is the new me, and the new me will be different.” And this sort of new me idea is around many other chapter breaks, right when you move to a new job or a new role, when you celebrate a birthday and you can say, “You know, in my 40s, it’s going to be different. It’s a new me.” All of those moments have the same psychology.
We also tend to step back and think big picture about our goals at these landmark moments, and so that’s what we find is propelling the “fresh start effect,” as we call it. And we’ve looked at lots of data sets, we just see these natural blips. We’ve looked at in terms of people showing up at the gym more frequently when they search for the term ‘diet’ on Google, which, for better or worse, is the most popular New Year’s resolution, and even when they set goals on a popular goal setting website about everything from their health to their finances to their education, all of these data sets show the same pattern where we see spikes in goal setting around these fresh start dates.
Edmonds: My favorite study that you cite in your work, I don’t think it’s your research, but this incredible data set they’ve uncovered distinguishing between baseball players who’ve been transferred within leagues and baseball players who’ve been transferred to another league. Can you explain what that study is and what it shows?
Milkman: Yeah, so this is work I supervised by my dissertation advisee, Hengchen Dai. I’m so proud of her. She’s amazing. She’s a tenured professor at UCLA now, but I vividly remember when she knocked on my office door and sort of started talking about this idea. And the idea was to turn what we’d already studied on fresh starts on its head a little bit. We’d been focused on the positive side of fresh starts exclusively, that they motivate people who are struggling to achieve a goal to begin this disconnect is psychologically beneficial. If you look back at your past and say, “Ugh, I haven’t gotten where I want to be, but now I have this fresh start.” And she wondered if, actually, it might be that fresh starts could also be harmful if you’ve been performing well, because then you don’t want to disconnect. You don’t want to feel separated from your past successes that could be very demotivating.
And so she found a really interesting way to study this. In addition to doing a bunch of clever laboratory experiments, she gathered data from professional baseball players, and it turns out that baseball has some really interesting facets of the way it works that allowed her to look at this question. So, players are often traded in the middle of a season to new teams, and some of those players have been having a great season, and some of them have been having a rockier season. Some are in the middle of the pack, of course. But there’s two different ways you can get traded across teams. Some trades are within league, and this is actually not nearly as big of a change. It’s not as much of a new beginning, because your season to date statistics all remain with you. They just keep tracking. But if you’re traded across leagues, you have a really big fresh start, because not only are you on a new team, but you’ve also got to begin again on all your season statistics, everything’s wiped clean. You have a clean slate there. And so what she looked at is what happens comparing players who are traded across leagues and get that pure fresh start. How do they compare with players who are performing equivalently but are traded within league? So it’s a tight comparison, and she looks at players who’ve had a great season to date and also players who’ve had a really weak season to date, and what she sees is that difference between the cross league and within league trade. It’s a really big positive for the people who’ve had a crummy season so far, they’re really well served by the pure fresh start. But it’s actually really bad news for the players who’ve been at the top of their game to end up with a cross-league trade as opposed to a within league trade.
And so the nice thing about this comparison is you don’t have to worry about things like regression to the mean driving this effect, because anyone who’s at an extreme is going to sort of revert to the mean, but she is able to take care of that by looking at the difference between equivalent players just as a function of whether their trade is across versus within league. And they’re very nice results combined with laboratory data showing this harmful effect of fresh starts for top performers.
Edmonds: As you mentioned, a fresh start is just the beginning. Then you’ve got to stay on the straight and narrow if you’re trying to change something in a positive way. And there are various tricks for achieving that. You’re well known for talking about something called “temptation bundling.” So explain to us what temptation bundling is.
Milkman: Yeah, temptation bundling is one of my favorite things I’ve studied, and I have to say it’s also “me-search,” because it’s something I did and do to help myself achieve goals. So this was the rare case where sort of a quirk from your own life can be turned into a science experiment. So let me tell you what I do, and then I’ll tell you about the science. This really started for me in graduate school. At the end of a long day of classes, I would find it very difficult to motivate myself to exercise. I am a former Division I athlete; exercise is critically important to me for many reasons, identity and also just truly mental health. I am a mess when I’m not physically active. But after a long day of classes, I just couldn’t get myself off the couch and to the gym, and all I wanted to do was sort of curl up with a cozy novel or binge watch TV. Hopefully this is relatable. And so I came up with a solution to get myself to the gym, and the solution was I only allowed myself to indulge in the entertainment I was craving at the end of that long day while I was at the gym. It was a gym-only treat. I started with audio books. I listened to the Harry Potter books this way. Now I do it with, I will admit, Netflix TV shows. Right now I’m actually binge watching the latest season of Bridgerton, but only when I’m on the elliptical.
And what that does for me is one at the end of a long day, instead of dreading going to the gym, I crave a trip because I can’t wait to find out what’s happening to all my favorite characters. When I get to the gym, time flies because I’m so engrossed in the entertainment, and finally, because I have this rule that I’m only allowed to indulge in that when I’m at the gym, I am really motivated. When I come home, I feel energized and ready to get my work done, instead of dying to procrastinate by binge watching TV shows or reading a novel. So it ends the sort of wasteful time. It gets me to the gym, and it changes the experience of the gym in this positive way.
So I’ve studied it. We call this temptation bundling, combining a chore that you dread with something you look forward to in order to make the experience more positive and one that you will crave instead of something you dread. We’ve run a couple of studies now. One was an experiment at the University of Pennsylvania’s gym with undergraduates and staff, and we randomly assigned some of them to condition where they were able to pick four audio books from a menu of 80 books that people said were really engrossing once they started listening. So these are books like The Da Vinci Code, The Hunger Games, the Twilight series. They put these on an audio device, then we brought them over to the gym, had them do a 30-minute workout while listening to the start of one of the books. And after they finished, we said, we hope you really enjoyed that if you want to hear what happens next, you’ll have to come back to the gym, because this device that we’ve loaded up with this book for you is going to be locked in a monitored locker you can only access when you’re here at the gym. And then we had a control group that also did a 30-minute workout, was also encouraged to exercise more regularly. They were also actually given a gift card to Barnes and Noble. So if they had the insight to temptation bundle, they would have the same resources to do so. So we wanted to hold gift exchange constant, but they didn’t get this idea from us to try to listen to audio books or enjoy some form of entertainment only at the gym. And then we measured gym attendance, and lo and behold, what we found was that the group that could only access these temptations that the gym exercised significantly more for the next seven weeks.
There’s another sort of fun twist in the study that I’ll mention too, which is in the very first project we did on temptation bundling. There was a holiday break – Thanksgiving break – after seven weeks in the semester, and this break lasted five days. The gym shut down, no one could access their audio books, and no one was on campus. And after that break, when people came back, the effect was completely wiped out, which I find really interesting and actually really aligned with what we just discussed regarding Hengchen’s research on how fresh starts are good when we have not been achieving our goals, but harmful when we are on a roll. And in fact, that insight is part of what led to Hangchen’s dissertation work, sort of talking about this study.
We’ve done other work with temptation bundling, showing that it can last up to months and months if there’s no such disruption. But it does highlight that when we are on a roll when we’ve built a habit that feels like it’s working, and then we have a period of disruption, we actually need to be quite deliberate about making sure we can sustain that habit and we have a strategy for sort of getting back on the horse after we return from a sickness, a vacation, whatever it might be.
Edmonds: Temptation bundling is a sort of reward that I give myself for going to the gym. What about punishment that I could inflict on myself if I fail to do so?
Milkman: Yeah, thanks for pointing out that there’s sort of the carrot and the stick approach, and of course, temptation bundling is tackling this present bias, making it more instantly gratifying. The flip side, if you’re worried about present bias getting in the way, is you could actually make it more painful to fail to do what’s good for you in the long run, and you could do that by, for instance, imposing monetary penalties on yourself and so on. And of course, this is the way parents, managers and companies and even the state often deal with present bias. So you might be assigned a deadline — teachers, I should say — so that you won’t procrastinate on your work, and you might be told you’ll get a fine if you are speeding, which might be very tempting to do, and that would sort of align present bias with the goals of your teacher, your boss, the state. What those all have in common, right, is they’re imposed on you by someone else. But what’s really interesting is it turns out, if you want to manage your own goals, you can also create constraints for yourself, and these are called “commitment devices.” And economists find them very bizarre, because it’s very counterintuitive to imagine finding yourself or constraining yourself in any way. Why would a human ever want to do that? But if you recognize that you may be tempted to do things that aren’t in your long-term best interest, in a moment of planning, you may actually decide you’re better off if you say, sign yourself up for one of these self-exclusion lists, that will mean you get kicked out. If you try to go to a casino. If you’re a gambling addict, you might recognize that that’s going to be a really good thing for you. And so you might say, you know, “I actually want to be constrained. I will choose to opt into that.”
There are also cash commitment devices, which are really quite interesting. This is where you put money on the line that you agree to forfeit if you fail to achieve some goal, and you define a referee or a way of evaluating whether you’re succeeding. So you might say, you know, “If I don’t go to the gym three times this week, I forfeit $10 I’m going to give it to a charity I hate to make it really sting!” And there are actually websites like beeminder.com, and stickK.com – which I have no affiliation with – that will let you set these things up formally and put a credit card in the line. She’s a referee, but you can also do it casually with a friend. And what the research shows is that this is extremely effective. So commitment devices help people sustain habits around exercise.
But most impressively, there’s a really nice, randomized, controlled trial run by Dean Karlan and collaborators, that shows smokers who are given the opportunity to put money in a commitment device account, an account where their money will disappear if they fail to pass a nicotine or cotinine urine test six months in the future. If smokers are given that opportunity, they are 30 percent more likely to quit than those in their randomized controlled trial who were just given standard smoking cessation tools. So that’s really powerful. Most of the change research I think about and talk about has nothing to do with addiction, because that’s a chemical dependency. It’s a totally different ball game. And a lot of these that are powerful enough to say, get me to the gym or make me a little more productive at work, can’t counter a chemical dependency, but commitment devices can, which I think is really noteworthy.
Edmonds: Yeah, that’s fascinating. Does it help to have very clear and rigid goals – “always run 20 minutes at the gym?” That seems like an obvious tactic. Is there research on that?
Milkman: Great question. Absolutely, it’s important for goals to be measurable and achievable, although they should be a stretch. You know, if your goal is ‘exercise more,’ how can you measure that? How could you even set a commitment device, for instance? But if your goal is 10,000 steps a day, and you’ve got one of these little I’m going to show you my Fitbit now, I realize that listeners can’t see it, but you know, if you got one of these little things on your wrist, well now you can actually be held accountable for whether or not you achieve that goal. And by the way, the little jiggle I get at 10,000 steps is its own remarkable reward. There’s really fascinating graphs you can see of sort of the huge spike of people hitting just 10,000 because of that jiggle, because of that goal. The goal itself and the jiggle that comes with it are compelling enough to attract behavior change. So setting concrete, measurable goals where you can see, did I do it? That’s really important.
It’s also important to have a plan of, sort of, when will I do it, where will I do it, how will I get there. These are called “implementation intentions.” I think the most important part of them is they associate a cue with the action. So just like an actor needs a cue to know when to say their lines, we need to not forget to take action on our goals. And so creating these implementation intentions, like 4 p.m. on Tuesdays, Wednesdays and Thursdays is when I’ll go for my workout, or 8 p.m. every weekday is when I practice Spanish on Duolingo for 15 minutes. Then when you see that moment, it triggers remembrance that you have to take this action and you can no longer procrastinate without recognizing, you know, it’s not sort of a vague intention, it’s a specific thing you’ve failed to do. And that is a much more momentous thing for us to ignore.
Edmonds: But I’m also wondering about the rigidity of it, because if I fail to meet that goal, that’s kind of depressing, and I might fall off the wagon altogether.
Milkman: Great. Yeah, you’re absolutely right. There’s something in marketing called the “what-the-hell effect,” which I think is one of the funniest names and best names for any phenomenon in behavioral science. And the what-the-hell effect describes what happens psychologically and practically, when we have a goal and we fall down on the job. Actually, I think, was first used to describe dieters. And what happens when, you know, in the morning you go into work, you’re on this strict eating regimen, and somebody’s brought in the donuts for a birthday, you have one, and you go, “Shoot — so much for my calorie counting goal today. You know, what the hell I guess I might as well have pie for lunch and deep-dish pizza for dinner. Everything’s out the window.” So the what-the-hell effect is something we have to be very concerned about if we set these goals, if it could demotivate us from continuing.
There’s wonderful research by my Wharton colleague, Marissa Sharif, and also Cornell’s Suzanne Shu, that shows one strategy that can help us with the what-the-hell effect. We can give ourselves a little bit of wiggle room. They call it an emergency reserve. And I think the reference to emergencies actually is quite important, because it highlights we shouldn’t use these all the time, but it’s essentially a Get Out of Jail Free card. So Marissa does this in her own life, and that’s what inspired it. She’s a runner, and she found that if she set the stretch goal of running every day, that was great and motivating for her. But if something got in the way on Wednesday, maybe a friend was in from out of town, or she was a little sick and felt like she needed to take the day off, then she’d say, “What the hell,” and she wouldn’t even try to hit her seven days straight. She just sort of [say], “Next week I’ll restart.” And so she would have a very bad week with running, and she realized she could maybe counter that by aiming for seven days. But saying seven days is my goal, but I’m going to give myself two Get Out of Jail Free cards, or two emergency reserves. And then if I miss Wednesday because of these untoward circumstances, well I can still get to seven because I’ll just sort of use one of my emergency reserve cards, and we’ll call it even.
So they did these research studies where they showed that if we give people a goal of five days a week, a goal of seven days a week, or seven days a week with two emergency reserves, there’s a huge difference in performance. People, by far, do the best with the seven-day goal with two emergency reserves, because they’re aiming for the high goal, but they’re not experiencing the what-the-hell effect if they happen to slip up.
So we should be thinking about in all our goals, what’s the type of emergency reserve we want to give ourselves? It’s a little bit of slack, not so much that you get lazy, and it needs to be reserved for emergencies, right? It’s not an entitlement to take a day off. That’s a different thing. But if we can give ourselves a little bit of slack but still stretch and aim for these tough goals, that seems to be the best of both worlds.
Edmonds: Right. You’ve touched on habit already in this interview, but I think you cite somewhere a study about people who are paid to go to the gym. Some are paid to go many times, some only paid to go. I think once or twice I probably got the details. Well, can you expand on that?
Milkman: Yeah, this is wonderful work that was done by or Uri Gneezy of UCSD and Gary Charness, who recently passed but was a professor of economics at UCSB, Santa Barbara. And they did this very clever study published in 2009 where they invited people who wanted to exercise more to let them track their gym attendance, and they offered these very large payments to everyone for being in the study, $175. But the way the payments were constructed was a little different. Some people just got paid for being in the study and allowing gym attendance tracking, so that’s a pretty sweet deal. Some were paid 175 for visiting the gym once, and this was sort of just to make sure they literally knew where it was, and then some were paid $175 only if they visited the gym eight times in a month. Now the first part of the study is really boring and obvious, which is during that month, when some people were paid $175 for nothing, and some were paid $175 for eight gym visits, you see very, very different patterns of gym attendance. The people who have to go eight times to get $175 they go a lot more than the people who only have to go once, for instance, or the people who don’t have to go at all.
But that’s not what the researchers were interested in. What they wanted to see is what happened in the following months. Did the activity of going to the gym repeatedly leave any habit? Did it create a change? So now no one’s being paid, but people have had a different pattern of behavior for the prior month, and what they find is there’s a very large increase in subsequent gym attendance. The people who had been paid to go eight times continue to exercise because they’d formed a habit. I’ve personally replicated this result multiple times. It seems like there’s about a 30 percent, I would say, rough carry over. It’s not like you go as often as when you’re being paid really large amount to go regularly to the gym, but you get about 30 percent of the benefit you had carrying over, because we do form habits. The more we repeat a behavior when it’s rewarded. That’s exactly what leads to habit formation.
Edmonds: We’re rattling through these effects now, but I’m also interested in the ambition of a goal. My brother is trying to persuade us all to sign on to this app whereby we all have to walk two and a half million steps in the first six months of the year. And I’ve crunched the numbers, and that turns out to be 14,000 steps a day, something like that. So how do you compare those two goals? 14,000 steps a day, or two and a half million in six months?
Milkman: Well, the first thing I’d say is you were very clever to crunch the numbers. And interestingly, most people seem to maybe skip that step, or they’re not that focused on that step. Knowing what the goal is each day is really, really important. So we did this one experiment where we were working with an organization called Crisis Text Line, and this was led by Aneesh Rai of the University of Maryland. And Crisis Text Line is an organization that provides crisis counseling to people via text message when they need it, and it’s staffed entirely by volunteers who’ve been trained to deal with these crises. They rely on this volunteer army, and everyone goes through extensive training to become a qualified crisis counselor, and then they commit to 200 hours a year of volunteering. But many do not reach that goal. They get a lot of reminders from the organization. “Don’t forget, you committed to 200 hours a year. Please do a little every week. Please sign up for your hourly commitments this week.” And we looked at those messages, and we said, you know, a lot of people probably aren’t doing the division and recognizing what the concrete, near-term, bite-size goal is that they need to achieve to stay on track. And that really matters. People need to be able to see what’s the bite-size thing. What can I do today? What can I do this week? Not, what am I going to do all year?
So we just did an experiment, an A/B test where we tweaked the messages from that sort of, “do a little every week to reach your 200 hour yearly goal,” to say, “you know, do four hours of volunteering every week to reach your 200 hour yearly goal.” And what we see is about an 8 percent increase in volunteering when we’re clearer and give people that bite size, approachable goal. So I think that’s a really important thing to recognize when we have these big, ambitious, long-term goals/ It’s intuitive, but we don’t always say to ourselves, “What does that mean I need to do today?” Not just what does it mean, I sort of vaguely kind of need to do regularly more. We need to know how much more, and be precise.
Edmonds: The other aspect of my brother’s challenge is that there’s a kind of competition, there’s more than one person involved, and I guess that motivates us in some way.
Milkman: Absolutely social accountability and visibility and camaraderie are really important. We did an experiment about this that we actually published last year. I’m talking about a lot of gym studies. We do a lot of work at the gym when we’re studying habits, not because we’re only interested in the gym, but it turns out to be really easy to track behavior, so I think of it as like the fruit fly of habit research. So this was a gym study, and it was about social accountability and the benefits of sort of pursuing goals with others. So we randomly assigned people, and this was led by Rachel Gershon at UC Berkeley. Everybody signed up for our program with a friend. So everybody wanted to exercise more. Everybody signed up with a friend. But then we randomly assigned them to two different conditions. Some people were paid $1 every time they went to the gym, regardless of whether their friend showed up, and other people could only earn that dollar if they and their friend checked into the gym within a few minutes of each other. Economists would say the second incentive, where you are contingently paid, is inferior. If we’re just trying to maximize gym attendance, you always pay directly for what you want, which is showing up. Now we’ve made it harder to earn this dollar. But if you believe that social pressure and accountability and the camaraderie we build from pursuing goals with others could add value, you might have a different theory. And in fact, we did, and that’s what we found, people actually went to the gym 35 percent more when their payments for gym attendance were contingent on showing up with a friend.
When we surveyed them afterwards to try to sort of look under the hood, the key things that came out is this feeling of accountability and not wanting to let someone else down, as well as the fact that it was more fun to work out with someone else. That’s sort of like temptation bundling here. We’re not, you know, combining a steamy TV show, but instead an enjoyable conversation and interaction. So absolutely, there’s benefits to pursuing goals in tandem with our friends.
Edmonds: Related to this is the study that looks at classmates and which room you’re allocated when you first arrive at college.
Milkman: Yes, absolutely. So this is work by Scott Carell, and he and his collaborators looked at students at the Air Force Academy, and what was the group of cadets with whom they were randomly assigned to live as freshmen. Did that impact their own study habits? And what he showed, what he and his collaborators show, is that we’re very influenced academically by the habits, essentially, of the people around us. So first-year students who were randomly assigned to have fellow cadets who were sort of more studious – they had done better on their sort of verbal SATs – that ended up pulling up others’ performance. And if you ended up with a group of roommates, right, who were sort of less studious, your performance suffered. So we were shaped to some degree by those around us, which makes total sense, right? If your roommate is staying in on studying on Thursdays and Fridays and Saturdays, you’re like, “Hey, I guess that’s the thing to do.” And if your roommate is going out and partying and trying to drag you with them on Thursdays and Fridays and Saturdays, you’re gonna think, “Hey, I think that’s the thing to do.” And so it’s going to affect your performance.
And this is true in all parts of our life. We are shaped by the people around us, by their habits, by their attitudes. And this can be a challenge or it can be an opportunity, if we actually choose to craft relationships and groups of friends who inspire us and push us to achieve more.
Edmonds: Let me end with something personal. Tell us a goal that you’ve had, which, despite being a world expert on goals, you’ve been unable to achieve.
Milkman: Oh my gosh. Where to begin? I mean, I’m a parent, and I would say that’s been the hardest challenge in terms of so many goals. You know, “Never yell.” How many weeks can I say I got through the week and I didn’t yell at my child? Probably count it on one hand, unfortunately. So maybe that’s one. I have plenty of professional goals that I also haven’t achieved, one that I’m working on right now that I feel is a little more achievable. I’d really like to find more uses for AI in my research. I do not feel like I am using large language models optimally, and one of my goals is to try to find a new use of a large language model for my research each day this year, and I’m particularly concerned about the gender gap with large language models are so powerful, and I’m sort of afraid, am I conforming to gender norms by not relying on them and not developing expertise in them? Enough? So that’s a goal, I would say, for the coming year, and I have not achieved it yet. I do not feel that I am adept an expert with LLMs, but I hope to become so.
Edmonds: Katy Milkman, thank you very much indeed.
Milkman: Thank you very much for having me.