From the original forbidden fruit in the Garden of Eden to the simplistic advice to “just say no” to drugs, it has been accepted wisdom that self-control is best exercised by an exertion of willpower. However, despite the time-honored history and intuitive appeal of this concept no one has been able to define exactly what willpower is or how it works. What allows one person to pass up dessert while another person can’t resist the temptation? In fact, why can the same person resist temptation one moment but give in to it the next? Various suggestions have been offered to answer this question.
One idea is that the ability to restrain behavior is a cognitive process in which certain information is necessary to help you decide the best choice to make. Think of it as Mr. Spock’s approach to self-control: you choose whatever you decide will work best and is therefore the most logical. Another explanation is that self-control is a skill that improves with practice, which might explain why adults are (usually) less impulsive than children. A third possibility is that self-control is a resource like muscle strength. Some people just have more of it than others, and it can even vary at different times in the same individual. According to this view, it requires a strong “willpower muscle” to overcome temptation.
About twenty years ago, a group of social psychologists, led by Roy Baumeister, tested these ideas . They did some experiments to determine which of these theories is correct by examining the behavior of volunteers in a series of laboratory studies using a dual task paradigm. This simply means that the research volunteers (college students getting class credit, so I use the term “volunteers” loosely) were asked to perform two different tasks, one after the other, both requiring the research subject to use self-restraint.
According to the idea that it is a logical process, the thoughts that are necessary to make a good decision and exercise restraint can be primed or prepared by the first task to improve performance on another self-control task that follows. The idea is that the first task gets them in the state of mind to use self-control, so their performance on the second task should improve. A different prediction would be made if it was a skill that improves with practice. Since one or two practice trials would not be enough to have a measurable effect, performance on the second task should not change. If, on the other hand, self-control is a resource or strength, then restraint should worsen as each attempt to use that “muscle” would weaken it.
Each of the three theories predicts a different outcome, so only one of the three competing explanations could remain standing. The studies showed that when the subjects had to restrain themselves on the first task, their performance was worse on the second one. These findings support the idea that self-control is a limited resource that gets depleted with use. This effect, often referred to as “ego depletion,” was seen whether or not the second task was related to the one that preceded it, as long as they both required self-control.
This basic research design has been repeated frequently in different ways and has consistently shown that self-control gets worse after it has been used. The conclusion that the researchers came to was that there is a supply of willpower that we have at any given time, and it gets weaker the more it’s used. Presumably, just like muscle strength, this supply is restored with rest, so you should be good to go the next day. A second question, though, has the real practical pay-off because it’s the one most non-scientists really want to know: how can self-control be strengthened? The researchers’ answer to that question also follows the logic of the muscle metaphor. If willpower works like a muscle, it should be strengthened with exercise.
Well, as scientists like to say, that’s an empirical question. So to test this hypothesis, the researchers used the same protocol from the original study, consecutive self-control tasks, but this time the same two tasks were repeated two weeks later. In-between these sessions the subjects were given exercises to do, as a kind of home workout to build willpower strength. This study found some improvement in how long it took to deplete self-control for the second task, but, as the authors acknowledge, it did not improve the ability to actually exert self-control. “Following the muscle analogy, our results did not show that the muscle (i.e., self-control) had any greater power after two weeks of exercise, although the results did indicate that it had greater stamina and was less prone to suffer from rapid fatigue after the exercise.”
The results of other studies that examined this question were, at best, mixed. A detailed critique of this research is beyond the scope of this summary, but it is clear that the results do not offer convincing support for the idea that a “willpower workout” can improve self-control. At best, it improves a very limited aspect of control. Unfortunately, this is of little comfort to those who want some practical advice on how to gain some control over their unwanted behavior.
Before I learned about this research about six years ago, I had been treating many patients who had difficulty with self-control. Most of the people that I saw were struggling with their weight due to emotional eating, but there were also many among them who had other behavior control problems as well, such as compulsive shopping, binge drinking, and gambling. My clinical experience was consistent with the laboratory findings that behavioral control does get worse when people had to restrain themselves in other ways. The behaviors often occurred in the context of increased work stress, marital stress, and very commonly, the stress of feeling obligated to take care of other peoples’ needs without complaint. I know that last one sounds very specific, but it was surprisingly common among the emotional eaters.
I did eventually figure out a connection between these stressful experiences and the behaviors that accompanied them, but I had come to a very different conclusion than Baumeister and his colleagues as to why this happens. Most importantly, my understanding of this effect has very concrete and practical implications that led me to a conclusion about how to treat these patients that is the opposite of the one he came to. Rather than improving self-control by practicing more of it, I tell my patients to figure out where they’re experiencing a lot of self-restraint and do less of it. I have been using this approach with very good outcomes over the past six years.
When I eventually discovered this body of research, I read it carefully to understand why our explanations of why this happens were so different if we’re seeing the same thing, and I think I figured it out: the difference is in how we view the concept of self-control. My way of understanding it also helped me solve a mystery that had first bothered me more than twenty years ago. I’ll describe that in more detail, as well as the implications for change, in future posts.