Roy Baumeister [ Dartmouth ] Understanding Self-Control and the Limits of Willpower. I’m particularly delighted to come and talk about the research we’ve been doing on willpower and what we’re trying to conclude from it. First of all, just what is it? What are we talking about? Self-control, I’ll start with that.

WillpowerSelf-control is basically the faculty we have for changing ourselves. It’s a way to override one impulse or one response and substitute another. It can be changing your thoughts, changing your emotions, resisting impulses, maximizing your performance, any sort of thing. But, again, it’s a matter of changing yourself. And the term used in the research literature is often self-regulation, and I like that term because to regulate means to change but not just any change.

It’s to regulate according to a particular idea or a standard.

So when the government tries to regulate how you make sausages or make buildings or whatever, it doesn’t just say do it differently, but it says, no, they have to measure up to certain criteria. There has to be a certain amount of actual meat in the sausage or a certain number of windows in the building or whatever. So when you self-regulate you change yourself to bring it up to ideas of how you could or should be or would be better. It’s vital for human social life and culture. Indeed culture is hard to imagine if people failed to inhibit their impulses and, indeed,  culture is a biological strategy of our species .

It’s been very successful for us.

And you say, well, why don’t other  creatures  pick it up if it works so great? But, the capacity to live in culture, to change your behavior, to conform to the norms and the values and the morals and the laws and so forth doesn’t appear to be well enough developed in  the others . Morality, we sometimes call self-control the moral muscle because the capacity to do what is right and what is expected morally rather than what you feel like at the moment is a crucial aspect, a crucial application of self-control. More grandly ideas of free will, self-control is one of the psychological realities behind the idea of free will.

The particular term is contentious depending on competing definitions, but what people have that gave them the idea of free will is this capacity to change yourself according to ideas and values and so forth. And there are both  state and trait aspects  of self-control which I will get to quite shortly.

Let me say a bit about the value of self-control and compare it to some extent with self-esteem. I said in a number of conversations today I started my career studying self-esteem, and we thought self-esteem had great promise for making people better, and if we could boost peoples’ self-esteem they would do better in school and in work and make better partners and be better citizens and so on. Most of that has failed to materialize.

Self-esteem seems to be  more a result than a cause , and so self-control, however, evidence suggests it is the real thing, the real deal that really does produce these benefits. And so I’ve kind of shifted some of my focus partly because it does seem to work better.

Longitudinal studies have really confirmed the value of self-control compared to self-esteem. For example, we’ve known that the kids with high self-esteem tend to get better grades in school, and this encouraged many of us for a while to think that, well, if we could just raise kids’ self- esteem they would become better students. But the correlation is because the causation is the other way, that when you track people over time their grades gradually, if they get good grades that leads to having higher self-esteem.

It doesn’t work the other way around.

Whereas with self-control, and that’s measured early, that does predict better outcomes later on even years and decades later. Just to go through some of these have shown success in work, success in school both in terms of grade point average, in work even salary and objective achievement. Positive relationships, people with good self-control have better, more lasting relationships, are liked and loved better by their friends and partners and so forth.

People with good self-control are  happier  than others.

They report  lower stress  in life. Partly, sort of trying to untangle that relationship, I mean to some degree stress affects self-control because it takes away your resources. But  one of the best ways to reduce the stress in life is to stop screwing up so badly  [laughter] and self-control is an important trait for enabling you to do that. Adjustment, mental health and so on, again, show these people with good self- control have fewer mental and physical problems. Even physical health, taking your medicine, flossing, all these things take self-control.

And the bottom line is that people with good self-control even live longer.

Socially and behaviorally, plenty of other benefits have been documented. People can overcome prejudices and biases and treat others more fairly. Criminality is associated with poor self-control and so forth. The last line here, it’s difficult to identify any major sort of problem that people have that doesn’t have some element of self-control failure added. And this is something Todd and I wrote  a preliminary scholarly book about a quarter century ago  back during our post doc, and that’s one thing we struck. It wasn’t just here and here and here. Like I said, it’s very hard to find any major problem that people have that don’t have some degree of self-control failure. Alright. Now, I said there’s both state and trait. Let me talk briefly about the trait as a stable aspect of personality.

It’s clear some people are better at self-control than others. Really ask your friends, check them out.

So evidence shows people with good trait self-control do better. There was recently a meta-analysis that took all the studies published and unpublished they could find looking at personality differences and trait self-control. Not surprisingly, again, the positive outcomes went with the good self-control. But a couple findings were more surprising. One was many people associate trait self-control with dieting and eating because we have to resist temptation and perhaps quit smoking. Well, it works there, there is some benefit that people with good self-control are better at dieting and maybe at quitting smoking, but those effects were among the weakest.

And so it’s a little bit unfortunate that people associate in their minds self-control with that because that’s not its best application.

In contrast much stronger effects with work and school, again, grade point average, salary, career achievement and so on, and medium level effect sizes are based on the statistically criteria, how big the effects should be averaged across many different studies. Medium effects on quality of relationships and on personal adjustment, things like that. Also what the researchers did was code all the behaviors for whether this is an automatic or a controlled [su_highlight background="#efebcf"]inaudible[/su_highlight]. Here the prediction was fairly straightforward because self- control should be controlling the controllable behaviors, controlling what’s controlled, and the automatic things should be exempt from it. They were, therefore, rather  flabbergasted when the results came significant in the opposite direction , and colleagues were writing this up and saying, well, maybe controlled behaviors can’t be  controlled and automatic , I don’t know, why is this happening.

So I said let’s go back and look at what those automatic behaviors were.

And they tended to be a matter of habits. And so this is creating a bit of a new approach or a new understanding for how self-control works. That, among the people that have good self-control, it’s not what they use to bail them out of a crisis or help themselves deal with the temptation in a desperate moment or whatever. But rather people with good self-control use it to  set up their lives  to run along smoothly.

They use, in a sense, self-control to  create good habits , and even though habits, once they’re habit, they become automatic, they are the avenue by which self-control can make a very good positive difference in your life. This goes back to, remember I said the strongest effects and findings were with work and school. Some people might think where does self-control enter into it when you’re a student doing your school work, and that would be when you have to stay up all night and finish that paper that’s due tomorrow morning. However, in the long run that’s probably not the best strategy. What  people with good self-control do is get started early and get the work done ahead of time and not be in the position where they have to draw on their tremendous reserves to bail them out of trouble  and get the paper done with an all night marathon.

Now, let me say a bit about desires.

We want to look at how people use self-control in everyday life. And most of what I’m going to say is  based on laboratory studies , but here’s one out in the real world. We wanted to see do some of these effects happen in which people use self-control to resist desires and resist temptations every day. Psychology has studied cognition very extensively. We know less about motivation, even the question of how often people experience desire throughout the day. This was something we didn’t know how to predict or guess when we set this up.

So in the procedure we had people, a sample of some students, some  citizens working in business  and so forth wear beepers for a week. And whenever the beeper went off they’re supposed to stop and report what they were doing at that time, were they having a desire, had they recently had a desire, if so how strong. Was the desire a conflict with other things and so forth. Are they resisting that desire, and end up are they going to do what they desire. There’s plenty of other measures, too. But we thought of a basic thing, a sequence like this, is there a desire or no desire.

People were reporting more desires most of the time, either they had one currently or they’ve had one in the last half hour.

If you take the narrow measure are they having a desire at the moment the beeper went out, then I think it’s about half and half. So  about half the time when people are awake they are experiencing some desire . Do they resist these desires? Well, the majority of the desires, the slight majority are not resisted, but a hefty minority, about two out of five desires people do, indeed, have to resist and try to resist them.

And then you see over on the right whether they acted out or not.

Even if they don’t resist that doesn’t mean they always do what they want. It’s only 70 percent. That’s because sometimes you can’t fulfill your desire not because you’re resisting it because the person you want to do it with says no [laughter] or it’s raining outside or whatever.

Still, the dramatic finding is  if you take that 70 percent as the base rate then it drops from 70 to 17 percent . So, again, this is important.

We’re using self- control quite a bit to  restrain  our desires. This is part of being a civilized human being and perhaps what sets us apart from other animals who don’t have to do it. But it’s important, to live in culture, to curb your desires, when you walk through the restaurant and somebody else has something that looks good, you don’t just grab it off their plate and gobble it yourself, although most other animals would do that at least if they thought they were bigger than the person who had the food. So restraining our desires is a big part of everyday life and is mostly successful.

Or, to put things another way extrapolating from what proportion of the beeps were resisted and so forth, people spend about 8 of the 16 waking hours per day in a state of at least mild desire.

Mild desires are more common than severe ones. They spend  three to four hours a day, depending on how you calculate, using self-control to resist desires . And that’s not the only use of self-control, but still that means self-control is used a lot, several hours a day. And then there’s about a glorious half hour on average every day of succumbing to temptations that you had previously resisted.

So, again, self- control works a lot of the time but certainly not always. It’s possible to take these data and do various things like  mapping . Did they report a conflict between their desire and others, such other goals and so on? Also, how strong was the desire? Some are stronger than others, and you get a map like this. And there’s a lot of information in this one.

The size of each circle is how often the desire was reported.

The red part means actually doing it, but the higher it is the more conflict people reported. My sense with the conflict is people spend a lot of time at work, and a simple formula for explaining a lot of the conflict results is is this activity compatible with work? So, things like well, it turns out if you break it down further the least conflicted desire that people report is the desire for a cup of  tea  [laughter].

So when you have a cup of tea you’re participating in  one of the few activities that isn’t a problem . But, partly, again, employers don’t mind if you have a cup of tea on the job. Employers have a little bit more of a problem if you do leisure activities on the job, if you go to sleep, have sex on the job [laughter], things like that.

So those are more associated with conflict.

The extremes, as I said,  the strongest desires, Freud said sex and aggression , and I’ve had people look at this and says where is aggression and violence, but I think that may have revealed more about them [laughter]. So you didn’t hear a lot of aggressive impulses in there. There were some. But Freud was right on sex. It’s one of the strongest desires people report, but topping that was sleep, and we kind of didn’t expect that. Apparently a lot of  modern citizens  are going through their daily lives wishing they could take a nap [laughter] or go back to bed.

The most conflicted, as I said, relaxing, things incompatible with work.

The least conflicted was the tea and other desires to drink something. And the weakest desires, and this is a bit of a surprise, are the desires for tobacco and alcohol. These are supposedly addictive patterns, in the addiction literature one reigning theory is that these are irresistible impulses that people simply cannot overcome. And so we’re actually trying to understand what this means. We have in mind to probably do another study.

We’re going to do some more closer analyses on these data to see if we can understand what’s happening. One idea was that, well, when you have the impulse for a cigarette you always just give in right away, but really the rates of giving in were just about average. They weren’t any different from others. Then we wondered, well, maybe on average the desires are weak but they do get strong, too.

So we looked at just the desires that were rated as irresistible. I put up this graph. Now, irresistible desire, this is something when Todd and I wrote the scholarly book, we wrestled with the idea how much, are there really irresistible desires and how common are they. People will say this fairly often that they are irresistible desires and, again, the addiction literature is full of that, that you can’t help it once you’re addicted and once you’re a smoker and so on. But you notice tobacco has one of the lowest rates of irresistible desires. They rated strength from 1 to 7 and 7 was irresistible, impossible to resist this.

Again, it doesn’t look like a minority of the irresistible desires is responsible for the problems with addiction.

So we have to, as I said, we want to do some followup work on that. Clearly there are irresistible desires a lot more than we thought in our previous book. On the other hand, just because they rated them as irresistible, then we went and looked do they try to resist them, and people spend a lot of time trying to resist irresistible desires, and they’re surprisingly successful when they do. Most irresistible desires are successfully resisted it turns out.

So perhaps the term irresistible desire is used somewhat loosely when people describe it, first was strong desire but, again, the truly irresistible desires are probably quite few and far in between as we thought back in 94.

In these data people high on trait self-control, one straightforward prediction what’s trait self-control for? Self-control is for resisting your desires. So we thought if nothing else it should be clear that they resist desires more than others.

And yet we found significantly opposite again. And this kind of surprised us, so what is happening, but it appears to be, again, that  people with high self-control don’t get themselves into problematic situations . They certainly feel less guilty so, again, they seem to screw up less and they have lower stress as well.

So that’s the idea of what we were saying playing offense rather than defense with self-control, this is one of the conclusions we had in the book that people with good self-control use it rather than playing defense to bail yourself out of trouble in the desperate moment of crisis or temptation or whatever, but to set your life up so that you don’t have these problems in the first place.

Or, to put it more pictorially  many of us think of self-control in the sense of Ulysses on the deck coming home from the Trojan War and passing by this place where he hears the song of the sirens  and they are tempting him to go closer to hear it and, of course, the sailors would sail their ships onto the rocks and drown there. So he wanted to hear the song and here’s a pictorial image that people are trying to resist this and using some  pre-commitment or helping devices . He’s got himself tied to the mast, and I guess a little cross-dressing seems he thinks will help. [Laughter].

I don’t know what that’s all about. The rowers there have stuffed up their ears so they can’t hear the sound and won’t be tempted. But apparently people with high self-control just took a different route home [laughter] and avoided the problem altogether.

So that’s an important clue. Instead of thinking of situations situation by situation, we think of self-control as something, again, creating good habits,  setting up your life and avoiding trouble in the first place . Alright, let me turn now to how self-control works.

There are three basic ingredients you can analyze it into. One is your goals, what are you trying to accomplish, what are the standards or rules or, again, morals or values or performance levels or whatever. You have to have some commitment that I want to do this with myself. I want to live up to this or that goal, this or that value. That’s an important first step there. Now, second, is monitoring which means keeping track of what you are trying to change. This often is the easiest way, if you’re looking for a quick fix, something to get started on improving your self-control,  keeping precise records of something works better .

The study in  Todd’s undergraduate  on people that go into eating binges they stopped keeping track. When dieters break their diets it’s not only that they decided to eat more, but dieters normally keep very scrupulous tally of how many calories and how much they ate and so on. But once they break their diet they stopped keeping track, and they sort of lose what’s going on which means it’s, again, hard to regulate.

So trying to take up jogging or whatever, keep explicit records. Write down every day whether you exercise or not because it’s easy to fool yourself, but if it’s written down in a diary or a calendar or whatever then you have an objective record and that works better. And then the third ingredient is willpower in the sense of some energy that you use. It’s  the capacity to change to override one response and to substitute another in its place .

For the psychologist it’s related to several other topics, inhibitory control, working memory and so on. But I want to focus for a while on willpower here. So,  the idea of willpower is an energy that implies that you have a limited amount of it . And so contrary perhaps to the Victorian stereotype that your willpower is a stable aspect of your character, rather it’s a fluctuating capability.

The same person will have more at one time than another, and more that it’s finite, that it can be used up to a certain point and then perhaps it’s gone. It turns out to be a little more complicated than that, but still the fact that it’s a  limited resource  has been fundamental to the way we’ve been thinking about it for a long time. Now, these effects, called  ego depletion effects , suggest this is what happens after you’ve used up some of your willpower, that after you expended some of this energy you’ll be less able to succeed at self-control in another context.

But one of the early studies we did that made this point we looked at would people keep working and keep going at a difficult task when you feel like giving up and it’s unpleasant to do?

The way we set up this study, we can’t tell people they’re setting their self-control because then they’ll be all nervous and then do weird things, so we told them we’re studying their ability to remember taste sensations, and that meant that they shouldn’t eat anything for several hours before the experiment.

So they come to the experiment, they’d skip lunch or whatever, they’re hungry, and then in the laboratory we set up this little microwave oven and baked chocolate chip cookies there, and you know how that blows out the aroma and stuff, so it really made the laboratory smell delicious. We knew we got complaints from people across the hall saying I’m trying to do my statistics here and you’ve got me smelling chocolate all day. So we brought them into the laboratory, they’re hungry, but it smells really good, seat them at a table, there’s a tray with all these cookies, and if they weren’t into cookies or chocolates were there, too, all very tastefully arranged.

Also on the table is a bowl of radishes.

And the experimenter, the one says, well, you’ve been assigned to the radish condition, so your task will be to taste the radishes and don’t touch the cookies,  they’re for other people in the experiment  [laughter]. We had two other conditions where they were supposed to eat the cookies and the chocolate, and one was in no food control, but the ones we cared about were the ones who had to sit there seeing those chocolates and smelling those chocolates and wanting those chocolates and instead had to make themselves eat those stupid radishes [laughter].

So we left them alone to maximize the temptation.

Of course we didn’t trust them, so we secretly observed and we saw people longing glances there, people would pick them up and sniff them and put them back [laughter], all this stuff.  People dropped them on the floor and put them back .

But anyway, everybody made it. Nobody bit into the forbidden food and everybody managed to eat the better part of a radish or two. So they successfully used up some of their willpower doing that. And then we took them to a different room where there was no food and no tempting odors and so on and administered this test which we borrowed from stress research which is how long do people keep going on this difficult, frustrating puzzle.

We actually had to make it unsolvable because it messes up the experiment if they actually solved it, but how long did they keep going before they gave up. And fairly dramatic differences that  people in the radish condition quit way faster than the others . Those of you who do behavioral experiments these are quite large effects for what you usually get in a laboratory measure.  Five minutes of resisting the chocolate temptation took ten minutes off their perseverance on the unsolvable puzzle .

So the point is resisting that temptation in the radish condition that took something out of them that they didn’t then have available to help them with the next challenge that came along, namely this task. So this study and many other like that gave us the idea that there is really some energy, some form of, if you will, willpower that people seem to use up as they go about their daily activities as they exerted to show self-control and resist temptation.

And this then impairs their ability to succeed at the next tasks that come along. So there are a great many studies showing this kind of effect now. There is a meta-analysis out last year that some people did in confirming it in the laboratory.

The only question is  how often does this happen in real life ? Well, given that we went back this last year when we had the beeper study to see is there any sign of it there. Now, we clearly were only beeping people once every two hours. We were missing a lot of what’s going on, but still we could say were they, as far as we could tell, were they resisting other desires as the day went on. So we calculated a  score of their depletion  by how often previously that day have they resisted other desires and how recently that happened. And we find, well, in terms of acting out the next desire that comes along, if they’re not resisting it it makes no difference.

Okay, that’s good, it shouldn’t.

But if they are trying to resist it their resistence gets weaker and weaker to the extent that they’ve resisted other desires that day and especially doing so recently. So,  ego depletion  is not something that you can cook up in the laboratory and get people to do it, but it has no relation to real life.

This seems to happen in the daily lives of lots of citizens even with this fairly crude measure that, yes, your capacity to resist temptation and so forth does seem to get weaker as you use it as the day wears on. Okay, couple of general points. First of all, we find  there’s one willpower resource, not many . Some people want to say, well, I have good willpower for my work and not so good for cleaning up my house or apartment. It’s the same resource for using all these different things, so you have a certain amount or not. Because it’s limited you may choose it, you may allocate it for one thing and not for another.

So, yes, we do see people who are successful in one sphere and not so self-controlled in other spheres, but that’s an optional choice. They’re allocating their resource to put it into one thing, not another. And, as I said, the main spheres controlling your thoughts, controlling your emotions, impulse control and task performance, they all depend on the same resource, and you can use it or allocate it or not however you want.

Now, more generally we like the analogy that it works like a muscle. I showed the depletion effect, so like a muscle gets tired after you use it, then self-control after you use it it looks in a sense like it’s fatigued.

It’s not as effective at the next challenge that comes along. Two other ways in which it resembles a muscle. One is that  the depletion effects are often conservation effects. It’s not that your willpower is completely used up, but rather you are saving what’s left . And, again, like a muscle, you see an athlete perhaps that doesn’t use the muscle to maximum exertion until it’s totally exhausted and then collapse and be carried off the playing field. More commonly once your muscles start to get tired you automatically start to conserve your energy, start to pace yourself.

The  shots that you can’t get anyway you don’t run after  and so on, using your energy more judiciously. That appears to be what happens with depletion of self-control because  you really never want to get down to zero in terms of your willpower . This was of some concern to us just ethically. We have people come to the laboratory and do these little tasks which seem ordinary, but if we were really using up all their willpower we’d have to have qualms about sending them out into the world where people who want to sell them things or seduce them or get them to join a cult [laughter] or who knows what then they wouldn’t have any capacity to say no.

But, no, they are conserving. As soon as it starts to get somewhat depleted automatically you start to conserve what’s left. And then the third finding is that you can  make it stronger by exercise .

Again, it’s resembling a muscle, and this, too, is important in terms of the message we want to send out. If we just told the first thing, well, every time you exert self- control it uses up some energy, some willpower, we’re afraid people would hear that and say, oh, well, I’m never going to exert self-control, don’t want to waste that. This would be a very socially undesirable message for us to put out.

But instead we can say, no,  if you exert self-control on a regular basis you will become stronger and improve your capacity , and so that would be better for yourself and society and that seems a better, healthier message.

Alright. In terms of the exercises, the idea of exercise is important. I’ve talked to various ones of you about intelligence and self-control. There is some overlap,  self-control and intelligence  being the two main traits that psychology has found it predicts success in many different walks of life and many different contexts. Well, improving intelligence has proven quite elusive especially to produce lasting gains, but self-control can be increased even in adulthood.

Given that there’s  only one willpower resource , any exercise you do that strengthens it will help anything else. It just strengthens the capacity in general. And there are about a dozen published studies that have shown this sort of effect in various contexts.

Usually they have people do one sort of self-control exercise for a couple weeks, and then the laboratory people measure them on some tasks have no relation to what they were exercising and they show yet significant improvements. Some of the most dramatic effects were these Australian researchers who had in one study they had people work on managing their money, so they would make financial plans and learn to curb their spending and try to save a little more and so on.

And, sure enough, it improved their spending.

But more impressively when they took laboratory tasks on having to track moving figures and so they were better able to discipline their attention and stay focused and concentrate. They also reported all sorts of peripheral improvements in their lives.

They were  less likely to leave the dirty dishes in the sink at night , and their study habits got better. They even ate better which  healthy food is more expensive than junk food, so it kind of went against what they were trying to do . And yet, the point is strengthening the one thing improved self-control which then filtered through all features and all aspects of life. To do exercises the key thing is the efforts you put in to make a change in your behavior. So the more often you’re doing that the better exercise it is. In some of our studies we have people who are right handed, we tell them to use their left hand for a variety of randomly chosen things like brushing their teeth and opening doors and so forth and just try to do that for two weeks.

And so we’re trying to get them since you’re in the habit if you’re right handed to reach for the door or the cup or the toothbrush with your right hand, to override that then your left hand instead.

Obviously, if they went on for a couple months doing that then after a while it would become habit to use your left hand, and then it would cease to have any exercise benefit once that’s an automatic response. Alright. Some of the behaviors are showing these depletion effects. Eating, studies are done here at Dartmouth where  depleted dieters will eat more than non-depleted dieters . If you’re non dieters being depleted doesn’t really seem to change how much they eat. So it’s only the kind of behavior that you’re typically trying to control, trying to restrain your desires or impulses.

That’s what gets weaker.

So the dieters are trying not to eat. When their self- control is depleted then they’re more likely to give in and eat more. Non dieters, since they’re not trying to restrain it, it doesn’t change their behavior. Drinking alcohol, many sorts of impulsive behaviors, this sort of thing has been shown. Perhaps interesting suppressing stereotypes is tied strongly into self- control.

Trying to  deal with people of another race  or suppress right about targets, gay people or whatever without using stereotypes,  this is something that people don’t do as well once they’re depleted .

Even issues of threat of death, effects that people have done in the laboratory of making people think about what’s going to happen when you die, and this produces a variety of effects. But these very much seem to be depletion effects that once you started thinking about your death and it upsets you, and shutting that out of your mind takes self-control, and that uses up some willpower.

On a more mundane everyday level spending money, again,  people spend more impulsively when their willpower has been depleted. They’ll pay more for the same products, they’ll buy more stuff that they don’t need . So don’t go shopping after you’ve used up your willpower in other tasks. A couple words about  intelligent performance. I know that’s of some concern here .

I’ve mentioned intelligence and self-control is the two trait psychologies found. Well, some of self-control performance has to do with or some intelligent performance has to do with self-control. We had people deplete their willpower by an  attention control task  where they’re supposed to keep focused on this and not look at these other distracting things. And then we gave them, there used to be a logic test in the graduate record exam, they got rid of it because apparently graduate students don’t have any use for logic [laughter].

But we like that test, and that’s a certain kind of reasoning. And then, sure enough, people did substantially worsen. I mean, look, they’re getting half as many right as they were depleted as opposed to not. It was just from watching a video and trying to control your attention, and yet that  depleted your willpower  and  impaired  the  intelligent performance . Not all intellectual tasks are impaired.

There’s a reading comprehension one where you had to think and extrapolate. There people did, that’s supposed to be moved over, significantly worse here. But a memorization task, memorizing a list of words,  rote memory was fine despite the depletion . No significant difference there. So it goes with some intellectual operations are fairly automatic like memorizing and retrieving from memory.

Those are unaffected by self-control and willpower, but depletion will affect the kind of things where you have to think. Or, another way of looking at it is the fluid versus the crystalized intelligence. The fluid measured by this CET where you have to think, they’ll give you one and then you have to extrapolate to something else. They’ll give you the population of Boston and you have to estimate the population of Chicago.

That kind of thing  where people have to move from one thing to another, that people do worse when they’re depleted . The GMAT, they’re actually two tests by that name, it’s a little confusing, but this one is a brief  IQ test , and that a sort of  general knowledge test like vocabulary  and so forth, and that’s no effect at all. Again, so this retrieving stuff from memory unaffected. Now, we long associate willpower with self-control, but one of the big steps in my view in building this theory was that this same energy resource that’s used for self-control is used for other things as well. So decision making seems to be affected by it.

And intuitively I think people can get the idea self-control depends on willpower, they feel the temptation is strong or weak and so I have to be strong or weak to resist that temptation.

Intuitively people can understand that. They tend not to have any intuitive idea that  making choices and making decisions will deplete the same resource and have an effect on their self-control . And more recent work initiative. Let me just quickly run through a couple of these studies. Here’s a simple one. We had people either make a series of choices or look at the same products and rate whether they’ve used them or not but not choose.

And then a standard laboratory test of self- control, the  cold pressor task , how long can you hold your hand in ice water before you give up. And the ice water is cold and unpleasant and you have this impulse to pull it out, and you have to override that and keep your hand in the water. As you see, making choices, again, you used up their willpower and so they quit substantially faster on the cold pressor task. Not all choices are equally difficult. If there’s an objective best answer people can sometimes figure that out, and that’s not nearly as depleting.

So here’s if you’re thinking of an apartment to buy, a long or short commute and the rent is high or low, well, here they’re fairly linear so there’s no one is better than others. This was depleting. This choice not so much because here it’s a pretty short commute, it’s not that much longer than the short commute, and not that much more expensive than the cheapest one. So objectively in a sense that’s the best answer.  If there’s a clear winner it’s not as depleting . So more struggling, a way to make decisions that seems to be what depletes people. So these studies show that  making choices uses up willpower , and self-control is affected thereafter. Now what about in the other direction?  If people exert self-control first does that impair their decision making ?

And the answer is  yes .

There are a variety of kinds of effects on decision making that have been documented. For one thing, peoples’ willpower is used up, so after exerting self-control they are more prone to want to not make a decision. If you have an option of just not deciding or postponing a decision, I don’t care, you choose and so on, that’s what people do when they’re depleted. Compromise like the commute versus rent or price versus quality or whatever, many people are compromising when they have their willpower, when they’re fully intact, but once they get depleted just give me the cheapest or just give me the best or whatever.

Their  compromise goes out the window .

There’s a  tendency to take the default option . Again, taking the easy way out once your willpower is depleted. And a study Jonathon Lavov [phonetic] did with  Audi dealers , and they very nicely controlled the series of decisions you make when you buy a car, arranged them in different sequences and so on. And it showed  as you make more decisions you get willpower used up, and so you’re more and more likely to take whatever is standard . And  marketers sometimes use this to take advantage of people . They’ll have you buy an Audi and there’s something like 240 different interior fabric options of the different colors and so on, and you waste a lot of energy doing that, light gray or dark gray or whatever.

And then you get to the time of you want the rust proofing and all this stuff and I’ll just take whatever’s standard, and you end up spending a lot more money for those expensive things just because it comes with it.

Impulsive, self-indulgent choices more likely,  irrational bias more likely to happen , a nicely controlled study that showed people let stuff that shouldn’t really logically enter into the decision do so when they are depleted. They’re using the trust game of economic decision making process.

A lot of  people are less likely to trust others and less likely to reward other peoples’ trust in them when they’re depleted . Going with the general tendency people are less fair and more selfish when they’re depleted.

So a variety of kinds of changes in decision making occur when willpower is down. Essentially the idea is there are two different ways of choosing. One is more expensive than another where you can put a lot of effort in thinking it out and figure it out and reasoning and comparing and so on, and a more superficial way where you can just go with what’s easiest or pick what looks best or whatever. And as people get depleted they make their decisions in the cheaper, easier way to do it. I mentioned initiative also or the idea that  people become more passive when they’re depleted .

This work not yet published, but a series of experiments suggesting the answer is yes.

In one we had people first deplete their self-control or not watching this movie with special instructions to focus and not look at the distracting stimuli. And others they just watched it with no instructions. And then experiments, okay, the next task the computer will tell you what to do, just sit here and follow the prompts on the computer. The experimenter hits starts and leaves the room.

The computer goes to blue screen and just stays there, and we see how long does the person sit there staring at that screen before doing anything or, look, the thing is not working. And, sure enough,  the depleted people sat there twice as long just staring at the blue screen  wasting their time thinking something might happen [laughter]. So, again, depleted people more passive and more prone to do nothing.

Now, a couple things suggest that when people are mildly depleted that there are various things that can happen to overcome this. And we started to become sensitive to the difference between being slightly depleted and more severely depleted. So a paper on assigning somebody to be the leader of the group, if you have a group of laboratory subjects and just say, well, you’re the leader and you others are the subordinates.

You  choose them at random .

There’s no real basis for it, but the  leaders  really rise to the occasion and they put out and they try to  perform better . Their performance at baseline is better, they show  no depletion effect , they keep going really hard. But you don’t magically get more willpower just because the experimenter says that you’re the leader. Eventually this catches up to them, so in the long run they’re super depleted. In fact, it also turned out that what leaders do is they kind of appraise a task that’s given to them. Is this worthy of me or should I hand it off to someone else?

And so they  conserve their energy by essentially disdaining certain tasks , and if they’re supposed to do them they just kind of screw off on those. But that appraisal process stops when they’re depleted. So once the leader starts to get depleted and you’ve used up your willpower and so on, then you start doing whatever’s put in front of you so you deplete yourself even more.

This can set up a kind of vicious circle where everything that comes along I’m responsible, I’m the leader, I got to do it and you put in more and more energy, and eventually you become extremely depleted and exhausted and it creates a vulnerability. So, again, being the leader at a mild level it seems to improve performance but at a cost of severe depletion. There’s another interesting challenge to this work published just a year and a half ago.

These people said, well, how about if we convince people that their willpower is unlimited, would they actually do well? And [inaudible] has this strong belief in peoples’ theories can shape their outcomes and so on, and so if they just had the right theory they would do great. Well, and sure enough  they found they could convince some people that their willpower was unlimited, and they didn’t show depletion effects .

They wrote up this  inflammatory article , look, it’s all in your head, it’s just a matter of what you believe. And we had some skepticism of it at first. I mean if believing in unlimited willpower could actually give you unlimited willpower why don’t all cultures in the world already believe. You’d think there would be no incentive to believe anything else. But perhaps it worked mainly just at the borderline.

So the paper now impressed where we replicated their findings actually, but got some people mildly and some people severely depleted. And the severely did four different depleting tasks, and then we measured their performance. And as you see the black bars are the ones where people were led to believe their willpower is unlimited. It does work for a little while, but it actually is significant in the opposite direction when you really need it.

When you’re severely depleted it actually made people that much worse there. So I think that probably explains why all cultures in the world don’t already believe and tell everybody your willpower is unlimited because  the help is just at the margin  and  the harm there could be substantial . So, willpower is indeed limited. Your beliefs and your designation as leader of the group and so on can make a difference but only up to a point.

Alright, two more things I want to cover in the final few minutes before we quit. First, what it is that gets depleted. We’ve been using willpower indeed for the first 10 or 15 years, we used it as a metaphor, and then we began to wonder is there something more physically real that gets used up here. How we got on to this is sort of a funny story. We were actually testing another hypothesis, but the work led us to the conclusion that this  willpower is tied into the body’s basic energy supply .

Glucose, which is a chemical in your bloodstream, carries energy around to the brain and to muscles and other organs and so on, and perhaps some processes consume it more than others.

We got the initial findings and we looked at the literature, and nutritionists have collected a lot of data about glucose effects without having any real theory for it. But, sure enough, they found things like juvenile delinquency, just been arrested had very low levels of glucose. Well, maybe that was the low willpower is what got them doing whatever got them arrested.

There are nicely controlled studies of children eating breakfast in school. They tell everybody skip breakfast, randomly give half the kids a breakfast right away, the others don’t get anything to eat, well then that morning the  kids who had breakfast they learned better , they behaved better than the others. At 10:30 everybody gets a snack and all the differences vanish.

So this and many other studies suggesting that this physical  level of glucose is, indeed, tied into capacity for self-control . In our studies in the lab we found some evidence that self-control reduces the glucose in the bloodstream. I’m not as confident about those findings. We did get them, but there are a couple other replications that found it and others more difficult.

There’s a lot of noise and fluctuation in the glucose levels which seems to be more durable.  Low levels of glucose predict poor performance . When your blood glucose is down your self-control tends to be off as the nutritionist data suggests. And then if we give people a dose of glucose in the middle of the experiment after they’re depleted, that seems to enable them to come back and perform quite well again.

So, trait self-control or state self-control studies we show people are depleted,  self-control gets worse but then it gets better again if they consume glucose . Decision making effects likewise can be reduced if people get a — the way we manipulate glucose we give people a glass of lemonade.  We’re in the south, it’s hot, they’re glad to have that. We mix it with either sugar or Splenda , and it tastes equally good to them.

They were happy either way.

But the  sugar provides a dose of glucose, the Splenda doesn’t really give you anything nutritionally  and, sure enough, the sugar has a beneficial effect, not the Splenda. Obviously I’m mindful of the irony of consuming sugar as a way to improve your self-control [laughter], but that’s just a lab thing.

Don’t try that at home.

I mean we just have people in the lab for a short time and we need something that works fast. If you want to actually have energy to make decisions and have self-control in your life you probably want something that your body can burn over a longer period of time like protein rather than sugar. Alright. Let’s see, we’re coming to the end here. Let me come to the last point here.

What does depletion feel like? Is there some way you can know when you’re in this state? Well, best evidence I have, this is  from the London Times review of our book , pictorial representation. But actually it doesn’t really feel like that. I do with my other energy in assembling that stuff, but actually the problem is there doesn’t seem to be any signature future.

A meta-analysis I mentioned last year on these depletion effects, the effects behaviorally very strong and clear.

Subjective effects of what it feels like there was hardly anything, maybe a little bit of a tendency to feel worse, but that could be carried just by a few studies where there was some negative emotion and it was felt a little more strongly.

And that got us to thinking, well, what is this all about?

We assume that self-control is all about the restraint, and that when you’re depleted your restraints are weakened. But if it’s impulse versus restraint the impulse itself should be exactly the constant. Of course, the same impulse with a weaker resistence you’re more likely to enact it.

But this was never tested.

However, when we put that to the test in the laboratory we started to get findings that  people feel things more strongly when they’re depleted . So we depleted people by having them suppress thoughts, and my friend Dan Wagner had this procedure to tell people not to think about a  white bear . Once you’ve planted that idea it’s hard to get out of your mind. So they use up their energy shutting down that thought. And then we showed them a movie that was kind of upsetting about wildlife, effects of radiation. Well, they were more upset if they were depleted than not.

So negative emotions were there.

Well, could that be negative, but no positive emotions, too. So if we showed people puppies  they thought the puppies were cuter if they were depleted . The shark was uglier and meaner, a more negative reaction there, so both positive and negative reactions were there. Was this mainly associations because everybody knows puppies are cute? No,  we showed people Chinese characters which they have never seen before, and they reacted more extremely to them , formed more stronger positive and negative reactions to them. We had an ice water study, depletion as usual people gave up faster on holding their hand in ice water.

That was just a standard. But what was novel is they reported it hurt more when they were depleted, had a prior task use up their willpower. The water actually seemed colder and more unpleasant to them. We had about the urge for a cookie. We said, well, you can eat a cookie when you’re done. How much do you want another cookie? Well,  depleted people wanted cookies more , and then they went on and ate more. And so we showed people you’re going to get this gift, how much do you want to open it? And instead of getting a one shot measure they had to control a joy stick, say moment to moment how much do you want to open that gift right now?

Oh, I really want to. I guess I don’t care. No, yeah, I do want it. [Laughter]

We could track what happened. And the data looked something like this. The people who were already depleted they wanted it more. It wasn’t just a one shot measure or a delayed effect. Throughout the whole period, indeed, with these sort of data you can get all these different measures. Their peak desire was higher, they were faster to reach their peak, more got stuck, however you analyze the data stronger desires there.

And this sent us back one last time to the experience sampling data. Would people experience their desires as stronger to the extent they’d resisted other desires during the day? And, yes, there was a tendency that the more recently they had and the more frequently they had resisted other desires that day the stronger whatever next desire came along the stronger it seemed to them. So, what does depletion feel like?

There’s no signature feeling.

There’s  no way that you can tell when you’re depleted. Instead it seems to just turn up the volume on life . When everything seems to hit you harder than others maybe that’s a sign that your willpower is down. So, to conclude, self-control depends on this limited resource of energy or the popular term is willpower. It is important and the glass is both half full and half empty.

Maybe people think they don’t have much self-control. Compared to other species we have quite a bit. Like I said, maybe we’d be better off if we had a lot more.

But we do have a lot and nature has done well by us. So this is powerful, widely used, as I said frequently every day, but it’s a limited resource. Used also not just for resisting temptation or for dieting or whatever, but for all manner of acts of self-control, for making decisions and choices, for initiative. We’re suspecting probably also activities like  planning and following plans that that will also draw on willpower . And ultimately [inaudible] the greatest human strength. I want to suggest the idea here is our species has succeeded by virtue of culture, by cooperating, by sharing knowledge, by working together in these large groups and units with interlocking roles and joint action and so forth.

Self- control to make yourself conform to the rules and demands of the group [background music] that’s crucial for the success of culture and thus for the very success, the biological success of our species and thus ultimately one of the key parts I think of what makes us human.


Tagged on:     

8 thoughts on “Willpower

  • 14/10/2015 at 08:54

    Best Quote: “Turns out diet and exercise aren’t the best application for this technology [of self-control]”

  • 16/10/2015 at 04:55

    It’s worth pointing out here, specifying, that Baumeister is talking about short-term “desire”. Not stuff like a desire to be svelte, or athletic or go down in history as a brilliant mathematician. Desires that you would have to keep monitoring and feeding, particularly if your environment is not doing it for you.Reference

  • 16/10/2015 at 06:47

    By the same token, popular culture puts a lot of energy into associating a chocolate chip cookie with feeling or being good. If as much popular culture messaging went into associating a radish with feeling good – propagating images and vignettes where radishes are consumed in special rituals involving salt on little bits of bread with sweet butter in environments associated with Success – or just lots of money – then maybe the chocolate chip cookie “people” would be caught surreptitiously dropping forbidden radishes on the floor when they thought no one or no thing was looking.

    What is being presented here as emblematic of short term desire is actually a deep system, set up over a long period of time on a very large scale.

  • 18/10/2015 at 07:12

    Interesting to compare Baumeister’s ideas about Desire with those of Anna Fels. One might argue, coming from the Fels camp, that what Baumeister has been studying, with his beeper exercises in Germany, is in fact none other than the infamous  Monkey Mind. Aka the opposite of Focus.

    I’ve just picked up Baumeister’s book, but whenever I come across his use of the word Desire, I’m going to try translating it as Monkey Mind and see if that doesn’t make everything make more sense.

  • 18/10/2015 at 07:28

    I think you should zoom in on the Splenda Defect. GREAT research study. So badabing.

  • 18/10/2015 at 07:30

    Why is the work of Anna Fels classified as Evolutionary, and that of Roy Baumeister as Creative?

  • 18/10/2015 at 07:35

    Good point. Excellent point. Been there, fixed that.

  • 18/10/2015 at 08:38

    In fact, what Baumeister et al seem to be isolating would probably be better described as Entropy.

    Clearly, foythamoah, there’s a Heisenberg Effect in action. Isn’t there?

    The fact that his subjects are charged with allocating memory to constantly keep track of whether a beeper is beeping could in fact be CAUSING what Baumeister is calling (unwanted or inconvenient) Desire(s).

    The Beeper “Effect” is right up there with White Bears and Chocolate Chip cookies, I submit to you, when it comes to solving seemingly impossible puzzles — and refraining from grabbing the meat off some stranger’s plate.

    When, exactly, did Dr. Baumeister conduct this research? Maybe there’s a great opportunity for a timely longitudinal.


Leave a Reply to asm Cancel reply

Your email address will not be published. Required fields are marked *