Lack of Willpower? Nope, Just a Realistic View of the World

David Berreby

Praised by politicians from Paul Ryan to Barack Obama, hailed by pundits as the root of all virtue and success, self-control is almost as popular this election cycle as Motherhood. In the prosperous 60s people talked about letting go and loosening up, and in the rich 90s we were more concerned with self-esteem, but now in a tough economy we pity those who can't keep it together. They must be flawed, damaged people, who need help to strengthen their character. All we're arguing about now is how best to give them that help (tax cuts and self-reliance or Head Start and community colleges). But in our worship of self-denial, we may be missing something important: Sometimes the urge for instant gratification makes perfect sense.

That's the message of this study, out this month in the journal Cognition. The authors—Celeste Kidd, Holly Palmeri and Richard N. Aslin—added a twist to a famous experiment that measured small children's ability to resist temptation. The original work assumed that this ability came from within: each child had a certain level of willpower, and the experiment was revealing the differences. But the version by Kidd et al. showed pretty clearly that kids' response to temptation depended on circumstances. Specifically, they showed that when the adults in the room appeared to be untrustworthy, children were far more prone to go for instant gratification. Which means their supposedly poor impulse control was a sensible reaction to the hand they had been dealt.

The famous experiment, conducted by the Stanford psychologist Walter Mischel in the 1960s, was this: A small child is left alone in a room with a marshmallow on a table. S/he's told that she can have two marshmallows, but only if she waits for the experimenter to return in 15 minutes without eating the one in the room. Some of kids gave in and grabbed the treat before their eyes. Others, though, managed to hang on for the full 15 minutes to win the bigger reward.

As the years passed, Mischel followed up with those children. As Roy F. Baumeister and John Tierney wrote in their paean to self-discipline, Willpower, he compared the children and found that those who had waited the full 15 minutes scored an average of 210 points higher on the SAT, were rated more popular by others, earned higher salaries and were less likely to abuse drugs or be obese.

That's striking. But Kidd et al. wonder if that kind of success should be attributed to some innate ability or well-acquired skill on the part of each kid. Maybe, instead, their resistance to temptation was a response to an environment where delayed gratification paid off. If your parents are around to give you the two marshmallows they promised, after all, your efforts to avoid eating one will go rewarded. On the other hand, as they write in the paper (pdf available here), "for a child accustomed to stolen possessions and broken promises, the only guaranteed treats are the ones you have already swallowed."

So they brought 28 young children (aged between three and half and five) into their lab for an altered version of the famous marshmallow experiment. Before the food test, the kids were each given two little art projects (one involving crayons and one involving a sticker). In both cases, the supplies on the table before them were pretty crummy. Each child heard the experimenter offer to go get better supplies. For 14 kids, this promise was fulfilled, when the adult returned with better stuff. For the other 14, the adult proved unreliable, by coming back and saying "I'm sorry, but I made a mistake," and explaining there were no other supplies. (Parents, feel free to cringe in recognition.)

The marshmallow test followed right after, and the results were striking: Children who'd dealt with the unreliable adult lasted an average of three minutes before eating the marshmallow in front of their eyes. But kids who had been with the reliable adult held out an average of 12 minutes. Of the 14 children who had been with the unreliable adult, only one managed to avoid eating the marshmallow for a full 15 minutes to earn the reward. On the other hand, in the group of kids who had dealt with a reliable adult, most (9 out of 14) were able to resist temptation. Which means that the variation in self-control in this experiment wasn't caused by differences in kids' willpower. It was caused by differences in the trustworthiness of the people they were dealing with.

There's a lot of confusion these days about two different claims about the mind: (1) People aren't nearly as rational in their decision-making as we think and (2) When people depart from the norms of the American upper-middle-class, they are being irrational. The evidence for (1), which is very strong, is not evidence for (2). Yet (1) and (2) are easy to conflate, especially for successful middle-class people, whose ranks include most of the researchers and most of the people who write about the research. If someone smokes, or fails to exercise, or blows their work deadlines all the time, it must because they're incapable of being rational about their life, right? After all, such behavior is never sensible in my life.

But the notion that a behavior is always bad is a pretty sure sign of an ideology. Human behavior varies because our circumstances vary. Being weepy and morose might be a sign of depression, or a sign that your best friend just died. Starting at every sound and keeping your finger on the trigger might be paranoia, or a good way to stay alive in a combat zone. Defining and pursuing a long-term goal might be an admirable trait—unless your long-term goal is to kill and eat a bunch of women.

This is why the paper's theoretical question for psychologists—is temptation resolved by self-management or by our understanding of the world around us?—should matter to the rest of us. It's a reminder that people are irrational when science has shown them to be, not when we wish to condemn them for not being more like us.