Colloquy Podcast: How to Change—for Good

“If you buy one self-help book this year, pick this one,” wrote New York Times writer Tara Parker-Pope about How to Change: The Science of Getting from Where You Are to Where You Want to Be, adding the book to the paper's list of eight favorites for healthy living in 2021. What sets How to Change apart is its focus on the internal obstacles that plague efforts to develop healthier habits and move forward in our lives. The author, University of Pennsylvania professor Katie Milkman, PhD ’09, says she organized the book this way in order to enable readers to identify common barriers and the best strategies for overcoming them. This month on Colloquy, Milkman discusses her research and how insights from economics and psychology can be harnessed to change consequential behaviors for good.

I want to ask a really basic question to make sure that we've got the right terms here and to see which are interchangeable. Are behaviors and habits the same thing, at least with the way that you talk about them in the book, and what are the obstacles that make them so hard to change?

I would say that habits are distinct from behaviors, and the way that they're distinct is that habits are repeated behaviors that have begun to feel automatic, that we engage in without necessarily deliberating on, you know, do I want to bite my nail right now? Is that the right thing to do? Do I really need to take another bite of this snack or sip of this coffee? When we are doing something without a lot of conscious deliberation just because we've done it before, and it's the natural progression of things, that looks like a habit.

But behaviors are any action we take, whether they've been deliberated on extensively or result from this sort of automaticity that habits tend to have.

So why are they so hard to change, whether they're a habit or not?

Yes, both are hard to change. Habits are particularly hard to change because we're literally not thinking about them. And so we have to re-alert ourselves to the fact that this is a decision point, and I might not want to take this action that is self-destructive in some way. So, habits are particularly hard to change, but behaviors writ large tend to be hard to change for quite a number of reasons. 

And really, the sort of central thesis of my book is there's no one answer to that question. In fact, there are several different answers, and figuring out which one is holding you back is critical to employing the right scientific tools to help you succeed. So, a common barrier, though, and one that I actually really first discovered and that helped me fall in love with behavioral science by taking classes as a graduate student at Harvard, is present bias. 

So present bias is our tendency to care much more about whatever we can get right now, in the present moment, than anything we'll have to wait to get, any delayed rewards. So I care much more about how good it will feel to eat the pizza right now than all the long-term implications of having had an unhealthy meal. 

Habits are particularly hard to change because we're literally not thinking about them. And so we have to re-alert ourselves to the fact that this is a decision point, and I might not want to take this action that is self-destructive in some way.

I care much more about how great it will feel to skip the gym right now than how good it will feel in the long run if I stay in shape. So that's present bias. It's a major barrier to taking steps that are good for us in the long run and making the kinds of decisions that we generally prefer to make if we're thinking globally, but we often don't make when we're focused narrowly on the here and now. 

There's also just the motivation we need to get started on something new and figuring out, you know, how do I convince myself that now is a good time to make an adjustment or make a change? Because we're inertial; we like to sort of keep doing what we've been doing. And so getting that sort of escape velocity for making a change itself is a barrier because we can always put off until tomorrow what we’d rather not do today. Very related to present bias is procrastination. Because of present bias, we often put things off. And so again, this is related to that escape velocity. So these first three ideas that I've mentioned, they're all tightly interwoven. Some of them start to get a little bit more differentiated though here. So one is forgetting. 

That's very distinct from present bias. Often it's just not top of mind to take action at a moment when we can to make a change. And if we forget, if it's not salient enough to us when we can do something, then we'll miss out on important opportunities. Another barrier is confidence. If you don't actually believe you have what it takes to make a change, you may never bother to try. 

So that's important. You started this question by asking about habits. And habits are a barrier to change, but they're part of a category of barriers that I like to think of as laziness, which sounds insulting, but you should remember actually that, when I was in the graduate school program at Harvard, it was a program where I was spending a lot of my time in the computer science department, and in the computer science department, laziness is a virtue. And I don't mean that the students are lazy. I mean the algorithms are lazy. You want to build the most efficient possible algorithm to solve any problem. And so it's, I think, no wonder that the human mind and the human condition involve a lot of laziness and a lot of desire for shortcuts and easy solutions, because that's what a good algorithm does. It looks for the easiest path. 

So habits are a form of laziness. It's our mind putting on autopilot a decision that we don't want to have to deliberately make over and over again. Those habits are hard to change because you have to short-circuit this loop that's built. Also, just anything that takes more effort, we're less willing to do, and a lot of the time change does take effort. It requires deviating from the status quo, deviating from the set of priorities that you've laid out day in and day out. So that's a barrier.

And then finally, the last thing that I have studied and written about is: do we see other people like us doing this, and are we getting social support? We look to other people around us for cues about what is possible for someone like us. And when the people around us are showing us that change is hard or they're sticking to old patterns, it's much more difficult for us to break out of our ways and make a positive change than if we surround ourselves with people who are showing us, “Yeah, actually getting in shape is straightforward. I ran a marathon recently. You know, I get straight A's. I'm an excellent student.”

So, who we surround ourselves with shows us what is possible. And that can be a barrier or a solution. But when it's a barrier, it's an important one to recognize so that you can try to make adjustments that will support your goals.

You've done some pioneering research on something called the fresh start effect. What is that and what does your work on it tell us about the timing of habit change?

This is one of my favorite topics that I've ever studied. The fresh start effect is going to sound intuitive, I hope, when I mention the first instance of it, which is New Year's. Everybody's familiar with the phenomenon where at New Year's people suddenly put down what they've been doing and set resolutions. About 40 percent of Americans do this, and it's sort of like a weird communal ritual that every news program that you watch on TV or listen to on the radio or, you know, whatever newspaper you pick up, they're all covering New Year's resolutions like clockwork, every New Year's, and people are talking about them. Then it goes away.

That is an example of the fresh start effect, the phenomenon that at moments that feel like new beginnings, we are more motivated to pursue change and set new goals. But what my collaborators and I have shown is that New Year's is just one such date. There are actually many moments in our lives that feel like new beginnings and propel the same kind of behavior, maybe not to the same communal degree. They don't maybe get the same headlines, but anytime we cross a boundary into what feels like a new chapter in our lives—and we do think about our lives as if we're characters in a book living through chapters as opposed to completely linearly—we think these chapter breaks really matter.

Chapter breaks come when you move to a new job, when you move to a new city or a new home. They come when you become a parent. They come when you celebrate a birthday. They can even come on Mondays and at the start of a new month, or after the celebration of a meaningful holiday that you associate with new beginnings. So think something like Easter if you're religious and Christian.

These are moments that come up consistently in our lives, and they give us this sense that because of this chapter break—who I was before, what I was not accomplishing, say, last year—that was the old me in the last chapter. And this is the new me. And things could be different. So any failures we can set further behind us, and they also tend to kickstart us into thinking big picture about our lives and about, you know, what do I want to accomplish this week or this year or in this new decade I'm entering as I enter my 40s or my 50s or my 60s, or in this new job or in this new community.

We think bigger picture at these fresh start moments, and we feel disconnected from our failures like a new me and more capable of optimism. Yeah, last year I didn't get in shape, but that was the old me, and this will be the new me. So what we've shown in our work on fresh starts is that all these moments that feel like new beginnings give rise to this increased motivation to pursue our goals, and also by highlighting fresh start dates and giving people the opportunity to goal set on fresh start dates.

We can increase the attractiveness of behavior change because people are drawn to moments that feel like fresh starts for making a change; we can make it more attractive to change by highlighting them.

I watch a lot of videos of the Buddhist monk Thich Nhat Hanh and also the Buddhist nun Pema Chödrön. They talk about the ability to come back to your breath at any moment as essentially starting again. And you can do it at any time of the day; any time can be a fresh start for you.

I think it's an idea that's in many religions. And it resonates very deeply with people, in part, I think, for that reason. What I find really interesting is that, you know, in data set after data set for every type of behavior that people might be trying to change, whether it's behaviors related to the environment, their education, or health, we see these consistent patterns where there’s upticks at the start of a new week in goal setting around these behaviors. 

There are upticks at the start of a new month; there are upticks following people's birthdays. There are upticks following holidays that we associate with fresh starts. So think Memorial Day and Labor Day in the US much more so than Valentine’s Day; we just see it so consistently. And then I think the other interesting part, which is what I think you're highlighting with this point about we can make fresh start moments by taking a breath, is that we do have the ability.

In our research, we've shown we can increase goal pursuit by drawing attention to the opportunity to create a fresh start. In one of my favorite experiments, we let people who weren't saving for retirement begin saving by checking a box on a postcard and mailing it back and signing their name. We’d take care of the rest; we’d get them enrolled in their employer-sponsored savings program, but we varied the offering. They could either start saving right away, always, or they could start saving after an upcoming birthday or after the start of spring. So fresh start dates—they could check that box, or in another condition, they could start saving in two or three months. You know, they could check that box, and it's literally identical. So if you have a birthday in two months, we're flipping a coin, and you're either invited to save now or in two months—that’s condition one—or now or after your upcoming birthday. So I've offered you exactly the same thing. But by associating that opportunity with a fresh start date, that upcoming birthday, I make it more attractive. 

And we see that people save significantly more over the next eight months just because we've put a different framing of that opportunity in front of them that highlights this fresh start. So there is an ability to create a fresh start simply by drawing people's attention to a different moment that they might have otherwise not noticed that they associate with a fresh start. 

We haven't been able to, or, I should say, we haven't run studies where we say, you know, turn today into a fresh start moment or this moment into a fresh start moment by taking deep breaths. So we have not been able to do that, nor have we tried to do things like say, hey, it's the 100th day of the year. It could be a great fresh start moment for you. Don't you want to start pursuing a goal? And we found that when we try to do things like that, we haven't been successful. Whereas when we peg the date to a moment that already naturally has fresh start associations for people, like the start of spring or an upcoming birthday, we get big effects. But not for these sort of "we just tried to convince you it was a fresh start" moments.

But I do think there's an interesting line of research that can be done on how we help people feel that there is a fresh start and a sense of renewal, even on an ordinary day? And are there ways, through ritual or other practices, that you can create a fresh start truly out of any ordinary Thursday?

I've come across the work in recent years of two scientists who study habits and how to change them. Wendy Wood is at the University of Southern California, and she emphasizes the importance of changing the cue or the context of a behavior. Judd Brewer is an addiction psychiatrist and behavioral neuroscientist at Brown who says it's more important to get people to pay attention to the consequences of the behavior itself. So, on the one hand you have “change the context.” On the other hand, it’s more like “change the reward value.” What have you found in your research on behavior and habit change?

It’s a great question. First, I want to say that I respect both of these points of view and both of these streams of research. And Wendy Wood is a member of the Behavior Change for Good Initiative team of scientists and a friend and mentor. And I think her work is incredible, in particular on habits. I don't think there's anything wrong with either of these perspectives, and I don't think they're really in conflict.

I think there are many things that you tend to need to do in order to make a change stick. Changing a situation to remove cues and to give yourself essentially a clean slate is incredibly powerful. It can be very powerful to change incentives, which is how I would interpret this idea of making people more aware that their behavior is actually being destructive, making it clear to feel that this is not good. It's not rewarding instantly the way that you'd want it to be. We know from lots of research that that’s the case.

In my own work on habits, I would say the most interesting thing we've found is really around habit start-up. And how do you build good habits? In terms of habit disruption, which is what you're more focused on? That’s actually not an arena where I've done as much work. Most of my work is around trying to build healthy habits as opposed to disrupt harmful ones. But when I look at that literature, the most impressive findings that I point to are findings around using what’s called a commitment device, or a tool that allows you to constrain your future self in some way.

So this is very much in line with Wendy’s work, but I think it’s also aligned with Judd’s work of saying, let’s, you know, change the way you perceive the incentives. My favorite study using commitment devices, which I did not do, is a study that invited people who were smokers to put money on the line that they would have to forfeit if they failed a urine test for nicotine or cotinine six months in the future. And that was an experiment. Some people were offered this opportunity to put money they would have to forfeit on the line, essentially to fine themselves for failure, and others were not given that opportunity. And being able to fine yourself, being just given this opportunity, led to a 30 percent higher quit rate because people had the tool they needed to create a constraint.

I think it's very related to Wendy's work and very related to Judd’s work because you're giving yourself the opportunity to change the situation, and you're also changing the value proposition of the behavior. And doing those two things together through this kind of a commitment device, that's very effective. Or creating any kind of friction, making it more costly in any way. You know, you lock up the unhealthy foods in a pantry with a padlock, and now you have to take the time to unlock the pantry in order to get to those foods. Well, now you’re going to have to deliberate about it and reflect: do I really want to do that? And that kind of friction could be very helpful. Or you can reduce friction and make it easier to get on your workout machine by buying one and putting it in your house instead of joining a gym that's a five-mile drive and having to get all the way there in order to get your exercise in.

I think there are many things that you tend to need to do in order to make a change stick. Changing a situation to remove cues and to give yourself essentially a clean slate is incredibly powerful. It can also be very powerful to change incentives, which is how I would interpret this idea of making people more aware that their behavior is not rewarding instantly the way that you'd want it to be. We know from lots of research that that’s the case.

So I believe in the importance of both of these ideas. And in my own work, I have focused less on that disruption question and more on building habits.

So let’s talk about building positive new habits. And this goes back to your original experience as a graduate student. Why does “fun” trump “effective” when it comes to behavior change?

This is really building on an insight that I will say, while I had it as a graduate student, I didn't have the clarity that I have now. And there’s some wonderful research that was done in the interim by Ayelet Fishbach at the University of Chicago and Kaitlin Woolley at Cornell University, showing that we persist longer on activities that we enjoy pursuing than on activities that are good for us in the long run but are unpleasant in the short run.

Most people think if there is some big goal I have, I should look for the most efficient route to getting there. Like I want to get in shape. Let me find the maximally punishing workout machine that will help me do it. That’s what we think we should do when we have some big goal: efficiency.

What [Fishbach and Woolley] show is that’s a mistake. If you actually encourage people to pursue a less efficient but more fun route to achieving a goal, you end up with better results because people are less likely to quit. So instead of getting in shape on the maximally punishing StairMaster, I take Zumba classes with a friend. Well, yeah, it takes more visits to the gym to get in shape, but I don't quit after my first visit because I actually enjoy it. I look forward to doing it.

Ultimately, this is the insight at the heart of my work on temptation bundling, that we have to find a way to make it instantly gratifying by combining a temptation with a chore so that we all want to engage in that chore. But I don't think that it was until after I'd done my research and found that temptation bundling was effective, and then years later, Ayelet and Kaitlin wrote their paper about the importance of fun, that I sort of put it all together and said, oh yeah, this is the fundamental principle, is that we misunderstand ourselves in this way where we think, let me structure the path to get to my end goal as efficiently as possible. Instead of recognizing that we're better off structuring the path to get to the end goal a little more slowly but enjoying the journey. And that, ultimately, is what will keep us on the path and get us to the end goal.

Back in the early ’90s, when PBS used to do their fundraising drives, they had a popular figure on there, a man named Covert Bailey, who wrote a book—I guess it was a bestseller—called Fit or Fat? And I remember seeing him, you know, he used to do these shows in front of these big audiences, and he said people all the time would ask him, "What’s the best exercise?" And he would say to them, "Whatever one you will do."

That’s right. And he would have been even more useful if he’d said, "Whatever one you’ll enjoy," because that’s the one you’ll do. And we can misforecast what we’ll do because we don't appreciate the importance of enjoyment.

You mentioned the person who picks the most punishing workout because that's supposed to be the most efficient way to get into shape. Going hand in hand with that, then, is this notion that an iron-willed discipline and a rigid adherence to routine are the keys to developing good habits and good behaviors. But what did you find about the importance of flexibility when you and your Wharton colleagues studied Google employees who wanted to start exercising?

I love that you brought up this study. And I do want to make a shout-out because Harvard professor John Beshears, who was one of my graduate school classmates at Harvard as well, was a key collaborator on this project. And we were both really interested at the time in this question of whether or not it would be more effective when building habits to have real consistency in the timing of when you perform the behavior. We thought, you know, maybe that would be incredibly helpful if at the same time each day, you knew, this is my gym time, and we can put things on autopilot more easily if the cue is really consistent.

So we randomly assigned Googlers—that’s what they like to call themselves—who wanted to build exercise habits to different incentive schemes during a month when they were trying to build a new workout habit. We either invited them to exercise any time of day and earn cash rewards for showing up at the gym, or they could only earn cash rewards if they showed up at a consistent time of day.

Now, everyone across our study chose their preferred time of day for workouts and signed up for reminders at that preferred time, but only for a subset of those people would their payment depend on showing up in that narrow window. It was a two-hour window that they picked that was their preferred time. This study involved, first, a 30-day payment period.

What we were interested in was not what happened for that month when we were paying them, but what happened when the rewards go away, the habit-building period is over. And now it’s time to look and see who has built habits that stick, who's still going to the gym even when they're not being rewarded for doing so. And also, when are they going? Are they going during that sort of sweet-spot time that they'd identified as their best time?

We went into the project thinking the people who had been paid for going at this consistent time, if we look at them, they're going to have built a more consistent habit, and they’re also going to have built a longer-lasting habit compared to another person who went just as often during our reward phase—because we know frequency matters for forming habits—but in a more variable structure, right?

So we had, in the end, two groups of people that we were able to identify from our random assignment study. One group went about half the time during that sweet spot, and the other group went at the same frequency, but 80 percent of their visits were during that sweet-spot time. And we compared who is more likely to make it to the gym in a given week after the program ended, and we actually found it was the people who had had more variability in when they visited the gym, the people who only half the time went in that sweet spot. 

And initially, I will tell you, we were puzzled. We were like, "What? This doesn't make any sense. Consistency does not breed habit? How could that be?" But then we dug into the data a bit more. What became clear was where this effect was coming from. It was that the people who we had trained to be all about that sweet spot, we saw that they visited the gym after the rewards period a little bit more often during their sweet-spot time. But if they missed that window, they didn't go at all. It was all or nothing.

However, the people who had been rewarded less and visited the gym more variably—they’d been rewarded for going whenever—they had developed a more robust habit such that they’re a little less likely to go in that magic time, but if they missed that time, they would still find some other time to go. So they'd gotten used to a Plan B and a Plan C, and that meant they were more likely to make it, because life does not always serve us a simple menu of, you know, yes, everything went exactly according to plan, and you can make it to the gym at your preferred time. Sometimes there's a traffic jam, an emergency, you know, something comes up that gets in the way. And the people who'd built those habits that were more variable ended up being more flexible and less rigid about exercise and more likely to stick to it.

So for us, this was a really surprising finding, given what we knew about the importance of cues and habits. And I've talked to Wendy Wood about this, whom you mentioned earlier in our conversation, who's an expert on habits and cues. And, you know, we have slightly different interpretations of what this teaches us. But I think everyone who reads this study agrees that it was not what we would have expected based on our theory of how habits form, but a really helpful study to point out that when we are building habits, we want to avoid rigidity and build flexibility so that we have fallback plans as opposed to a one-and-only plan.

Durable change is all about having systems that are resilient to the inevitable slip-ups that we will face, and that can be having a growth mindset or a learning mindset.

You’ve alluded in your discussion of the Google study that a lot of the time attempts at positive habit change don't take, you know, at least not at first. When you were talking about your fresh start research, you started out with New Year's resolutions. But of course, so many New Year's resolutions die after January. So, as we wrap up, I'd love you to talk about the importance of having a learning orientation to the inevitable slip-ups and trip-ups of building new habits in general.

When we look at how we help people sustain behavior change in the long run, which is the holy grail of all of this work, and as you point out, fresh starts are only helpful for getting us started. The reason that I study behavior change for good is because we need to find durable solutions, not just springboards that get us moving but then we freeze moments later.

It is so critical that we be resilient to setbacks. One of the ways we can be resilient is by practicing flexible routines so that we are ready to fall back when something goes wrong. Another way we can be resilient is in having a growth mindset, recognizing that any setback is just a learning opportunity.

Another way we can be resilient is by planning for those kinds of setbacks and even giving ourselves what my colleague Marissa Sharif calls emergency reserves: psychologically setting tough goals but giving yourself "get out of jail free" cards if you don’t, say, meet those tough goals. Not so many that you've given yourself freedom to fail all the time, but enough that you forgive yourself when there's a slip-up. You know, “I mean to exercise seven days a week, and that's always my goal. But I get one emergency reserve per week.” And in my colleague Marissa’s work, she shows that that actually helps people achieve more because they keep pushing and aiming for the highest rung, the highest achievement. But when they have a setback, they can forgive it and keep pushing.

It’s very much like the idea of a growth mindset. When I have a setback, I don't interpret it as meaning I'm incapable. Instead, I interpret it as an opportunity to learn and grow and do even better next time. So I do think durable change is all about having systems that are resilient to the inevitable slip-ups that we will face, and that can be having a growth mindset or a learning mindset.