Self-control and finite willpower reserves
Fascinating article in New Scientist by Roy Baumeister, unfortunately locked behind a subscription paywall (but this is a magazine to which it’s well worth your while to subscribe). From the article, some snippets:
After decades of research, psychologists now reckon two traits are most likely to make us successful. The first is intelligence, with smart people doing better at all jobs. Unluckily, there is little evidence that you can make lasting improvements to intelligence.
The other trait is self-control, the ability to change thoughts, emotions, actions and level of performance on duties and tasks. Of course, goals, moral rules, laws, social expectations, personal commitments and other forces play a role, but the more you can change yourself, the more successful you tend to be.
Studies on self-control have their roots in the “marshmallow test” devised by Walter Mischel at Stanford University, California, in 1972. . .
Other studies support this. People with stronger self-control do better at school, earn more and are more respected by co-workers. They are also less likely to be arrested, have fewer personal problems, less stress and live longer.
So what is this amazing thing called self-control? The common sense view is it depends on using willpower to resist temptation and to enable the right action. Our research suggests this notion is not entirely fanciful but that it lacks a key dimension. Research has shown repeatedly that after people exert self-control, they tend to perform relatively poorly on a subsequent, seemingly irrelevant test of self-control. The most plausible explanation is that “energy” was consumed and depleted during the first test, leaving less for more challenges.
Evidence for this depletion of willpower comes from studies like ours in 1998 . . .
Hofmann found that people spend a staggering three to four hours a day on average just resisting temptations and desires.
Not surprisingly, as the day wears on, the more often the person exercises self-control to try to resist what they desire, the more likely they are to give in to whatever temptation comes along: it’s not the time of day that matters, but the cumulative exertion that saps your willpower. If you do not have many temptations to resist, your willpower stays relatively strong, and you may well be able to resist new temptations.
So rather than seeing willpower as a moral quality, the scientific view is that it is like a muscle that tires. After you exert self-control, you have less willpower so you are less able to resist a new demand. Self-control is only temporarily weakened and can recharge after a rest. Willpower resembles a muscle also in that it can be strengthened by exercise.
Two clear facts about willpower have emerged so far. Willpower is what researchers call “domain-general”: controlling thoughts, emotions and feelings, restraining impulses, and performing tasks and duties will draw on one pool of willpower, not, as people tend to imagine, multiple pools with different quantities for, say, dieting or exercise.
The second fact is that the resource is limited. Even a few minutes of exerting self-control is enough to cause a decline in performance on a subsequent, seemingly unrelated test. That might suggest human willpower is scarce, but, again, no: willpower is like a muscle, and when a muscle gets tired, an athlete may cut back effort to conserve what remains. In fact, willpower looks as if it is indeed a kind of energy, tied to levels of the chemical glucose used to carry energy from the digestive system and fat stores to muscles and other organs. Neurotransmitters, that enable brain cells to fire, are made of glucose.
The standard willpower depletion effect, confirmed by a 2010 meta-analysis of 83 studies, shows that after exerting self-control, people perform worse on the next self-control task without being given glucose between tasks. Researchers use lemonade these days: one batch sweetened with sugar (plenty of glucose), the other with diet sweetener (no glucose). After allowing up to 15 minutes for the lemonade to reach the bloodstream, subjects drinking sugared lemonade perform quite well at the next test, while those on diet lemonade fare less well.
This glucose research also suggests why dieting is so fiendishly difficult. In order to resist tempting foods, we need willpower but to have willpower, we must eat. The essence of dieting (restricting food intake) robs us of the psychological strength needed to succeed. Perhaps dieters should concentrate on filling up with healthy food so they have the willpower to resist fattening stuff.
If research continues to implicate glucose in willpower, it could be a powerful key to understanding the human mind since self-control is such a vital part of daily life. But willpower is also used in making choices and decisions, so here’s a startling thought: could daily decision-making impair self-control?
Last year, Jonathan Levav at Stanford University and Shai Danziger at Ben-Gurion University of the Negev, Israel, came up with important clues when they studied parole judges in Israel. The safe and easy decision is to refuse parole since it carries the risk of the convict committing further crimes – and making the judge look bad. The researchers found judges often granted parole in the early morning, but as the day wore on and they made more decisions, they were less willing to take a chance and sent most people back to prison. After a snack, or after lunch, the likelihood of parole went up. In other words, the food seemed to restock the willpower depleted by making many choices, leaving the judge more willing to take a riskier step and grant parole to the next applicant. . .
When I first started planning my weight-loss book, a year or so ago, I gave it the title Weight Loss Without Willpower, because I specifically designed an approach that made as light a demand as possible on my willpower. I did it just because I wanted to avoid the effort of resisting, but it seems to have been a wise decision (as evidenced by my current weight—172.0 lbs this morning—and the research described above).