The ability to regulate our impulses and desires is indispensable to success in living and working with others. People with good control over their thought processes, emotions and behaviors not only flourish in school and in their jobs but are also healthier, wealthier and more popular. And they have better intimate relationships (as their partners confirm) and are more trusted by others. What is more, they are less likely to go astray by getting arrested, becoming addicted to drugs or experiencing unplanned pregnancies. They even live longer. Brazilian writer Paulo Coelho summed up these benefits in one of his novels: “If you conquer yourself, then you will conquer the world.”

Self-control is another name for changing ourselves—and it is by far the most critical way we have of adapting to our environment. Indeed, the desire to control ourselves and our environment is deeply rooted in the psyche and underlies human engagement in science, politics, business and the arts. Given that most of us lack the kingly power to command others to do our bidding and that we need to enlist the cooperation of others to survive, the ability to restrain aggression, greed and sexual impulses becomes a necessity.

Social psychologists' appreciation of the importance of self-control reflects a shift in perspective. Thirty years ago many of them mistakenly regarded cultivation of self-esteem as a panacea for personal problems and social ills—an honest mistake. High self-esteem is associated with doing well in life, so it was reasonable to assume that a boost would improve people's lives.

When analyzed more closely, the data suggested that self-esteem does not itself lead to success. It is less a cause than an effect. When researchers tracked students over long periods, they found that getting good grades results in better self-esteem later. But having higher self-esteem does not produce stellar report cards. Self-control, however, is the real deal.

Experiments on self-control began in the 1960s with pioneering studies of delaying gratification conducted by Walter Mischel, now at Columbia University. Using a procedure that came to be dubbed the “marshmallow test,” he offered children a choice between immediately getting the white, cylindrical candies (or another of their favorite treats) or else receiving a couple of those same sweets if they could only wait for a while. More than a decade after these early experiments were published, Mischel and his colleagues tracked down the children, by then young adults, and did so again as they entered middle age. The ones who had the most success at resisting temptation at age four went on to be the most successful as adults.

Recognizing the requirement of self-control for well-being, I and others have set about probing the psychological and biological processes underlying it. The findings indicate that the act of opting to not express anger or of choosing to forgo a marshmallow is akin to drawing on a store of energy that gets you through mile 26 of a marathon. As with any source of energy, it becomes depleted over time and needs replenishing. What psychologists have learned about self-control in recent studies may even provide new ideas for treating the seemingly intractable challenges of drug and alcohol addictions.

Mental Muscle Building

I have spent a quarter of a century doing laboratory studies on self-control with an endlessly fascinating stream of creative colleagues. Over that time, I have come to the conclusion that self-control, which might also be referred to as self-regulation or willpower, works something like a muscle does. In particular, it seems to “tire” after a workout. Several hundred studies in many labs have now replicated the basic finding that, after exerting willpower, people have less left over to complete a second chore. In one early study, we found that people who called on their self-control to resist chocolates and cookies later showed less fortitude on a difficult problem-solving exercise. They gave up much more easily than people who had not worked their willpower “muscle.” In other studies, trying to suppress a forbidden thought—such as being told not to think of a white bear—made participants less able to control subsequent emotional reactions.

We coined the term “ego depletion” to label the state of diminished willpower that follows from expending psychic energy on self-control, be it resisting temptation or forcing oneself to make tough decisions. The term was chosen as an homage to Sigmund Freud, who proposed that the self consists partly of a well of energy. His vague theories about how this energy worked are now mostly obsolete, but he did recognize that some form of psychic energy explains our behavior. Cast aside for decades, this idea reemerged when our experiments found that self-control operates as a mental muscle of sorts, a muscle in which energy stores get depleted with use.

Two other lines of research have extended the muscle analogy. Experiments by Mark Muraven of the University at Albany and his colleagues have shown that after exertion, willpower has not entirely vanished. Rather the body seems to be conserving energy; if an important challenge or opportunity arises, more self-control can be tapped. This finding parallels what happens with physical muscles. As muscles begin to tire, athletes cut back on exertion to conserve remaining energy and strength. But they can marshal concerted effort if needed, calling on reserves for a sprint to the finish.

Muscles do not just become fatigued; they increase in strength when used regularly. Self-control can also strengthen with practice, as shown when people go through an exercise program to enhance it. In several studies, volunteers were assigned for a two-week period to change how they speak—avoiding curse words, using complete sentences, and saying “yes” and “no” instead of “yeah” or “nope.” In another program, subjects were simply asked to improve their posture—sitting or standing up straight. After the exercises were completed, we evaluated the subjects' self-control using lab tests, such as squeezing a handgrip for as long as possible, but without any altering of speech or posture patterns. Those who had practiced the earlier exercises performed significantly better than a control group that had not had to clean up their language or sit up straight.

It has occurred to us from these studies that the Victorian notion of “building character” seems to have some scientific validity. Exerting self-control on a regular basis appears to build up a person's capacity to call on more of this character trait in a pinch.

When we did these studies, we began to ask ourselves whether actual physical energy was getting used up—or whether the notion of energy consumption was in fact just a psychological metaphor. An answer to this question came serendipitously when a failure in one experiment led to a new and useful insight.

Some scientists presumably march from one successful study to another, but I am not one of them. In this case, Matt Gailliot, then a graduate student, wondered whether we could extend the observation that willpower becomes depleted when someone resists temptation. What about the opposite case? Would indulging in temptation actually strengthen willpower?

I had my doubts, but I encouraged Gailliot to pursue the question, which we informally called the “Mardi Gras theory,” in reference to the Christian tradition of indulging in sinful impulses in preparation for a period of self-denial during Lent. First we sapped people's self-control by requiring them to mentally suppress the forbidden thought of a white bear. Then we randomly assigned some of the participants to drink a delicious ice cream milk shake before they took a disguised test of willpower that consisted of searching a matrix of numbers for a particular sequence. In fact, the sequence was not there, and the goal was to see how long people could keep trying before they gave up.

The folks who drank the shake persevered longer on the test than those who got nothing. This apparent victory for the Mardi Gras theory was soon undercut by another result that involved an additional control group. One of the groups, as before, received nothing to drink before the test and, as expected, did badly on that test. The other group drank a milk shake that did not taste good; it contained unsweetened half-and-half rather than ice cream, so it was basically a large, unappetizing glass of dairy glop. Unfortunately for Gailliot's theory, the half-and-half group also did better than the unfed subjects. Gailliot was initially glum because the experiment seemed a bust. But as we talked, another thought occurred to us: If it was not the pleasure of indulgence that restored willpower, could it have been the calories?

We started reading up on glucose, the sugar in the bloodstream that provides energy to bodily tissues, including the brain, the seat of self-control. We ran a large series of studies and came up with two supportive findings that have stood the test of time. One showed that when blood glucose is low, self-control suffers, often substantially. This pattern, by the way, gives credence to the oft-heard complaint that a person is having difficulty functioning because of “low blood sugar”—a conclusion that also jibes with studies from nutritionists.

Credit: Daniel Stolle

The other meaningful finding confirmed that a dose of glucose administered just before self-control is beginning to flag helps to restore the needed willpower to press ahead. These results strongly suggest that willpower is, indeed, more than a metaphor. Further, if exerting self-control diminishes willpower and the energy needed to sustain it, then the remaining energy can be conserved by cutting back on further demands for self-control.

A third result did not hold up. We found in one study that blood glucose levels drop during a task that requires self-control. Such a finding would be consistent with the idea that exerting willpower uses up glucose. But we could not replicate the pattern reliably in later tests. Some studies from other labs have shown, however, that the brain uses more glucose when exerting greater effort—which makes sense, after all, given that it is the brain that controls self-restraint.

A Challenge to our Ideas

Like many scientific theories, our muscle model of self-control has evolved as other researchers have gotten into the act. Some have tried to build on what we have done, and others have wanted to dismantle or challenge our work. These findings—and the debates they have engendered—have helped flesh out our understanding of self-control.

One contentious issue has been whether the brain really runs out of fuel for willpower. Like us, other investigators have confirmed that self-control is impaired when blood glucose is low, a physiological state that affects both body and brain. Some researchers have argued that the human body has extensive reserves of glucose that could be drawn on if an amount allotted to willpower got used up.

Compounding the skepticism over our notion of energy depletion, the brain's glucose consumption does not fluctuate much—still, it does change some. In prehistory, people might have faced a threat of running low on glucose, but few in industrial societies need to worry—certainly not the well-fed college students in our experiments who showed signs of ego depletion or impaired self-regulation.

All these points are well taken. It is possible, though, that exercise of self-control does not necessarily lead directly to the exhaustion of glucose and that when the body senses that available glucose is running low, it makes adjustments to direct the sugar to where it is needed most. In that case, we would still be correct in thinking that willpower is a precious resource—one that needs to be conserved. The simplistic view that ego depletion means that the brain exhausts its fuel supplies is not tenable, but it does seem likely that the urge to conserve a partly depleted resource is powerful and pervasive.

Another critique suggests that any willpower deficit can be overcome by just putting people with declining reserves into circumstances that cause them to call up additional resolve. Studies have shown that assigning people to a position of power and leadership—or even paying them to try harder—makes them continue to show good self-control even in situations where their energy should be depleted by prior exertion of willpower.

This research raises the possibility that willpower is all in your head. No resource is actually depleted, but people simply lose motivation to work hard. It can also mean that when willpower declines, you can still exert effective self-control if doing so is critical. Think of the chief executive who feels the responsibilities and pressures of leadership during a corporate crisis.

In a related criticism of our view, Veronika Job, then at Stanford University, and her colleagues, including Carol Dweck of Stanford, whose innovative theories they built on, have proposed that willpower is limitless and that a person with sufficient motivation can simply keep going. For these researchers, the idea of ego depletion is an illusion based on a false belief.

Our energy-allocation theory does not entirely disagree with the view that people can draw on spare resources for a time. If your willpower is slightly depleted, your body may naturally seek to conserve what remains—but you can still suck it up and perform well if the situation warrants. Tired athletes conserve their energy for the winnable points and the crucial, decisive moments. Ego-depleted people do the same with willpower.

In our own studies, we have found that people who believe in unlimited willpower tap into existing reserves to increase blood glucose levels when the sugar should have otherwise been depleted. The story, though, grows a little more complicated when examined closely.

A crucial test came when people were not just slightly depleted but continued exercising self-control until serious fatigue could no longer be ignored. Kathleen Vohs of the University of Minnesota, Sarah Ainsworth, one of my former graduate students and now at the University of North Florida, and others had shown that cash incentives or leadership responsibilities enable people to sustain self-control even when their willpower is depleted. But these various studies then initiated a grueling series of exercises, which showed that depletion worsened, and self-control started to diminish. Crucially, those who had been led to believe in unlimited willpower actually did worse than others. That belief had been helpful at first, but in the long run it backfired.

Self-control, it seems, can be maintained—but not indefinitely. After all, you do not get an infusion of glucose because you think willpower is unlimited or because you have been promoted to a managerial role. You just become more willing to spend from your reserves. Eventually a limit is reached. The illusion of endless self-control is tantamount to believing that a bank account has infinite funds. At the beginning, you may spend freely, but ultimately you seriously risk running out of money.

Can you will away an Addiction?

Recent studies have revealed newly discovered areas in which self-control plays a pivotal role. Some of these findings overturn prevailing ideas about various forms of addiction. A widely held view suggests that cravings for drugs, alcohol or cigarettes take over an addict's life and that quitting is impossible without complex medical treatments or at least a firm commitment to a 12-step-like program. Alan I. Leshner, former director of the National Institute on Drug Abuse and now CEO of the American Association for the Advancement of Science, has asserted that addiction is a “brain disease.” As he put it, a user may take a puff or inject a substance voluntarily, but at some point, a switch in the brain is thrown. Substance abuse becomes involuntary, and the compulsion lingers even when the addict earnestly desires to quit. Willpower and volition disappear once addiction takes hold.

New findings indicate, however, that any brain changes occurring in addicts do not lead to a loss of control over one's actions; often these people have the power to choose whether to give in to a craving or resist.

More specifically, addiction does not bring about changes in a brain area essential for self-control that governs movement—that is, the motor cortex, where actions, whether brushing one's teeth or reaching for a crack pipe, are initiated. As addiction grows, the decision to grab the pipe does not suddenly become involuntary. Instead addiction brings on a slow and insidious change in desire. Heroin or cigarettes evoke pleasant feelings that develop into a longing for these substances.

The addict can resist for a time but gives in at some point, perhaps sooner rather than later, and must thwart the desire again and again. The desire is not always overwhelming, however. In a study by Wilhelm Hofmann, now at the University of Cologne in Germany, people were contacted at random throughout the week and asked to report on any desires that came to mind. Yearnings for cigarettes or alcohol were rated as weaker than any others.

This and other findings indicate that the addict experiences an intermittent stream of one mild urge after another. The frequently recurring nature of these urges is what makes quitting a challenge. But the addict is not beset by the mythically insurmountable difficulty of resisting an overwhelming desire.

The controversy about whether addicts are still in control will likely persist. Arguments from politicians, drug counselors and others help to sustain the myth that addiction is rooted in overwhelming, uncontrollable urges. Many addicts themselves favor this viewpoint because it exonerates them from personal responsibility. The media often promulgate these arguments, perhaps because actors and other celebrities who develop an addiction want to retain the love and loyalty of their fan base, an easier task if they can attribute their drug use to external demons and uncontrollable psychological drives. Fans might not be so forgiving if celebrities just admitted that they simply like taking drugs.

Psychologists differ as to whether self-control can be an effective antiaddiction medicine. A survey in the U.K. found that addiction-treatment counselors who worked as volunteers tended to think that addicts can regulate their impulses. But those who received compensation for their work preferred to think that addicts are helpless and cannot get better without expert help. This argument is not intended to suggest that clinicians are in it only for the money. But when a controversy arises, financial incentives probably make it easier for people to endorse evidence that goes along with their own interests and to spot flaws in counterarguments.

Another addiction myth holds that cravings grow more acute only when quitting an addictive substance. A clever study by Michael Sayette of the University of Pittsburgh and his colleagues demonstrated that smokers believed that their desire would increase steadily over time, especially if they were told they could not light up.

The study also showed that these beliefs were wrong. Some participants had to abstain for a time and report their desire to smoke as part of the study protocol. Instead of the desire for a cigarette rising steadily, it went up and down unpredictably. Other studies have found that when a smoker quits, the desire to smoke goes down immediately and mostly stays in abeyance. If the addict relapses, as happens frequently, it is not because of an overwhelmingly strong urge for a cigarette. Rather it stems from a rather weak urge to light up at a moment when the smoker's willpower happens to have reached a nadir.

Addiction is for the Strong-Willed

The idea that quitting an addiction requires willpower makes sense to most people. But until recently, few have considered that starting a drug habit and staying addicted also require self-control. Most of us do not really like the first taste of beer or the first puff of a cigarette. Public health warnings about their dangers can also dissuade indulging for the first time. Willpower, in fact, is needed to overcome those negative influences and take the first steps on the path toward addiction. To sustain an addiction over a long period, a user must expend a substantial amount of energy to ensure that a habit does not interfere with work, family and relationships.

Consider smoking. So many restrictions exist today that smokers need to craft elaborate plans to sneak a cigarette. When my former university introduced rules prohibiting professors from smoking in their offices, one colleague struggled heroically to comply. I will not soon forget the sight of her heading out of the building into a Cleveland snowstorm, while holding her tiny baby in her arms, on the way to light up.

Just think about how much self-control she had to muster. First, she had to plan when she would find breaks between classes, appointments and meetings—and where she would go to not violate campus smoking restrictions. Then she had to dress herself and the baby warmly. She also had to remember her cigarette pack and lighter on the way out into the storm.

A study on smoking conducted in the Netherlands by Michael Daly of the University of Stirling in Scotland, myself and our colleagues lends some credence to the anecdotal observation that maintaining an addiction requires good self-control. Daly's group found that a Dutch workplace ban that began in 2004 reduced smoking but with a couple of important qualifications. The ban lowered smoking mainly among people who rated themselves lower on a survey of self-discipline, and this group returned to its old habits within a few years. There was no apparent effect on people scoring high in self-discipline.

Scientists who favor the view that addicts have little self-control might have expected a different initial outcome—high self-control types would alter their behavior in response to the ban, whereas poor self-regulators would keep right on smoking. And they might explain the fact that we found the opposite result by reasoning that people with low self-control needed the strong push from the law to get them over the hump—and interpret the subsequent relapse by suggesting that over time the threat of a legal cudgel somehow faded.

But the explanation for the results appears to be related to the addicts' need to draw on reserves of willpower to preserve their habit. For years people had comfortably smoked right at their desks. Suddenly, that option was gone. To continue their habit after the ban, smokers had to make elaborate adjustments. Each break had to be arranged carefully to plan where and when it could happen.

People with good self-control could handle these demands. But those with low self-control gave up and quit for a while. Over time, however, they noticed the coping strategies of the self-controlling smokers. They saw, for instance, that these diehards headed outside to a particular “smoker's spot” in the park. They then simply copied those strategies.

A number of studies have shown that addicts seem able to consistently plan and execute intricate strategies to maintain heroin or cigarette habits—habits that researchers, clinicians and even users themselves once thought to be unshakable. These findings provide a new perspective on addiction. The possibility exists that these groups may be able to redirect the same sustained willfulness they use to procure a drink or fix toward kicking their habits. But this idea also raises a new set of issues.

A therapist may have difficulty convincing an addict that he or she has taken the wrong path if that person sees nothing permanently damaging with having a few drinks or popping painkillers while continuing to fulfill responsibilities at home and work. This new insight into the nature of addiction provides further evidence of the extent that self-control can influence our behaviors in myriad ways—and how it may even, perhaps counterintuitively, enable us to persist in adhering to self-destructive habits. It demonstrates, once again, that our ability to control our emotions and desires lets us manage, for good or bad, the endless challenge of adapting to the world around us.