How habits work and how you may control them

by Kaj_Sotala12 min read12th Oct 201376 comments



Some highlights from The Power of Habit: Why We Do What We Do in Life And Business by Charles Duhigg, a book which seems like an invaluable resource for pretty much everyone who wants to improve their lives. The below summarizes the first three chapters of the book, as well as the appendix, for I found those to be the most valuable and generally applicable parts. These chapters discuss individual habits, while the rest of the book discusses the habits of companies and individuals. The later chapters also contain plenty of interesting content (some excerpts: [1 2 3]), and help explain the nature of e.g. some institutional failures.

(See also two previous LW discussions on an online article by the author of the book.)

Chapter One: The Habit Loop - How Habits Work

When a rat first navigates a foreign environment, such as a maze, its brain is full of activity as it works to process the new environment and to learn all the environmental cues. As the environment becomes more familiar, the rat's brain becomes less and less active, until even brain structures related to memory quiet down a week later. Navigating the maze no longer requires higher processing: it has become an automatic habit.

The process of converting a complicated sequence of actions into an automatic routine is known as "chunking", and human brains carry out a similar process. They vary in complexity, from putting toothpaste on your toothbrush before putting it in your mouth, to getting dressed or preparing breakfast, to very complicated processes such as backing one's car out of the driveway. All of these actions initially required considerable effort to learn, but eventually they became so automatic as to be carried out without conscious attention. As soon as we identify the right cue, such as pulling out the car keys, our brain activates the stored habit and lets our conscious minds focus on something else. In order to conserve effort, the brain will attempt to turn almost any routine into a habit.

However, it can be dangerous to deactivate our brains at the wrong time, for there may be something unanticipated in the environment that will turn a previously-safe routine into something life-threatening. To help avoid such situations, our brains evaluate prospective habits using a three-stage habit loop:

From behind a partition, for instance, it’s difficult for a rat to know if it’s inside a familiar maze or an unfamiliar cupboard with a cat lurking outside. To deal with this uncertainty, the brain spends a lot of effort at the beginning of a habit looking for something— a cue— that offers a hint as to which pattern to use. From behind a partition, if a rat hears a click, it knows to use the maze habit. If it hears a meow, it chooses a different pattern. And at the end of the activity, when the reward appears, the brain shakes itself awake and makes sure everything unfolded as expected.

This process within our brains is a three-step loop. First, there is a cue, a trigger that tells your brain to go into automatic mode and which habit to use. Then there is the routine, which can be physical or mental or emotional. Finally, there is a reward, which helps your brain figure out if this particular loop is worth remembering for the future.

Over time, this loop— cue, routine, reward; cue, routine, reward— becomes more and more automatic. The cue and reward become intertwined until a powerful sense of anticipation and craving emerges. Eventually, whether in a chilly MIT laboratory or your driveway, a habit is born.

Unused habits disappear very slowly, if at all. If a rat is trained to find cheese in a particular section of the maze, and the cheese is then moved to a different location, it will obtain a new habit. But once the cheese is moved back to its original location, the old habit re-emerges, almost as if it had been active for the whole time. This is part of the reason why it is so hard to start exercising regularly, or to change one's diet: the habit of relaxing in front of the TV, or snacking on a meal, will still be activated by the old cues and engage the behavioral pattern. On the other hand, if one does manage to establish a habit of ignoring the snacks or going out for a jog, it will eventually become as automatic as any other habit.

Habits are crucial for our ability to function. People with damage to the basal ganglia, the parts of the brain responsible for habitual behavior, often become mentally paralyzed. Even basic activities, such as opening a door or choosing what to eat, become difficult to perform, and they may need to pause to wonder whether they should tie their left or right foot first, or whether to brush their teeth before or after taking a shower.

In one set of experiments, for example, researchers affiliated with the National Institute on Alcohol Abuse and Alcoholism trained mice to press levers in response to certain cues until the behavior became a habit. The mice were always rewarded with food. Then, the scientists poisoned the food so that it made the animals violently ill, or electrified the floor, so that when the mice walked toward their reward they received a shock. The mice knew the food and cage were dangerous— when they were offered the poisoned pellets in a bowl or saw the electrified floor panels, they stayed away. When they saw their old cues, however, they unthinkingly pressed the lever and ate the food, or they walked across the floor, even as they vomited or jumped from the electricity. The habit was so ingrained the mice couldn’t stop themselves.

It’s not hard to find an analog in the human world. Consider fast food, for instance. It makes sense— when the kids are starving and you’re driving home after a long day— to stop, just this once, at McDonald’s or Burger King. The meals are inexpensive. It tastes so good. After all, one dose of processed meat, salty fries, and sugary soda poses a relatively small health risk, right? It’s not like you do it all the time.

But habits emerge without our permission. Studies indicate that families usually don’t intend to eat fast food on a regular basis. What happens is that a once a month pattern slowly becomes once a week, and then twice a week— as the cues and rewards create a habit— until the kids are consuming an unhealthy amount of hamburgers and fries. When researchers at the University of North Texas and Yale tried to understand why families gradually increased their fast food consumption, they found a series of cues and rewards that most customers never knew were influencing their behaviors. They discovered the habit loop.

Every McDonald’s, for instance, looks the same— the company deliberately tries to standardize stores’ architecture and what employees say to customers, so everything is a consistent cue to trigger eating routines. The foods at some chains are specifically engineered to deliver immediate rewards— the fries, for instance, are designed to begin disintegrating the moment they hit your tongue, in order to deliver a hit of salt and grease as fast as possible, causing your pleasure centers to light up and your brain to lock in the pattern. All the better for tightening the habit loop.

However, even these habits are delicate. When a fast food restaurant closes down, the families that previously ate there will often start having dinner at home, rather than seek out an alternative location. Even small shifts can end the pattern. But since we often don’t recognize these habit loops as they grow, we are blind to our ability to control them. By learning to observe the cues and rewards, though, we can change the routines.

Chapter Two: The Craving Brain - How to Create New Habits

A basic rule of marketing, based on the habit loop, is to attempt to identify a simple obvious cue, and then offer a clear reward from one's product. An early success was in the marketing of Pepsodent, where the marketer instructed people to run their tongue across their teeth and notice the existence of a "film" on the teeth. He then argued that by using his toothpaste, people could get rid of the film and obtain beautiful, clean teeth. (In reality, the "film" is a harmless membrane that builds up on teeth regardless of how often one eats or brushes their teeth.)

However, other toothpaste companies had tried similar marketing tactics before, without much success. Another part of Pepsodent's success was that it happened to contain citric acid, as well as other chemicals that act as mild irritants. Their effect is to create a cool, tingling sensation on the tongue and gums of people. This acted as the real reward for the habit - although the sensation itself only happened to occur by coincidence, people came to associate it with having brushed their teeth, and of having a clean mouth. It was when people began craving this reward that tooth brushing really became a habit. When other toothpaste companies realized what was going on, they all proceeded to add similar irritants to their products.

“Consumers need some kind of signal that a product is working,” Tracy Sinclair, who was a brand manager for Oral-B and Crest Kids Toothpaste, told me. “We can make toothpaste taste like anything— blueberries, green tea— and as long as it has a cool tingle, people feel like their mouth is clean. The tingling doesn’t make the toothpaste work any better. It just convinces people it’s doing the job.”

When a habit becomes sufficiently established in the brain, the cue no longer just activates the routine - it also makes us crave the reward that is associated with completing the routine. If the cue is present, but we can't engage in the routine or try to prevent ourselves from doing so, the craving will increase in strength until it becomes almost overpowering. Various cues - the sight of a pack of cigarettes, the smell of food, a computer or smartphone chiming to signify the arrival of a new message - can activate the anticipatory mechanism, and the craving to take a smoke, eat a bite, or check one's messages.

Scientists have studied the brains of alcoholics, smokers, and over-eaters and have measured how their neurology— the structures of their brains and the flow of neurochemicals inside their skulls— changes as their cravings became ingrained. Particularly strong habits, wrote two researchers at the University of Michigan, produce addiction-like reactions so that “wanting evolves into obsessive craving” that can force our brains into autopilot, “even in the face of strong disincentives, including loss of reputation, job, home, and family".

The same mechanisms can also be used to encourage good or healthy habits. One chooses a cue, such as going to the gym as soon as one wakes up, and a reward, such as smoothie after each workout. Then one thinks about the smoothie, or the endorphin rush that follows during the exercise. As one allows oneself to anticipate the reward, a craving will begin to ensue, which will make it easier to get oneself to the gym every day. (See also PJ Eby on this.)

Cravings are what drive habits. And figuring out how to spark a craving makes creating a new habit easier. It’s as true now as it was almost a century ago. Every night, millions of people scrub their teeth in order to get a tingling feeling; every morning, millions put on their jogging shoes to capture an endorphin rush they’ve learned to crave.

Chapter Three: The Golden Rule of Habit Change - Why Transformation Occurs.

The Golden Rule of Habit Change is that one cannot extinguish a bad habit, only change it. One keeps the old cue and the old reward, but changes the routine. Almost any behavior can be changed if the cue and reward stay the same.

For example, alcoholics rarely crave the actual physical state of intoxication itself. Rather, people drink in order to obtain escape, relaxation, companionship, blunting of anxieties, or an opportunity for emotional release. Organizations such as Alcoholics Anonymous1 build a system of "sponsors" and group meetings, allowing a person in need of relief to talk with their sponsor or attend a group meeting. The cue, a need for relief, stays the same, as does the reward: getting relief. What changes is the behavior: instead of drinking, one obtains their relief by talking to others.

Habit reversal therapy is the formal version of this technique. In one example, Mandy, a 24-year-old graduate student had a compulsive need to bite her nails. The therapist asked Mandy to describe what she felt right before bringing her hand up to her mouth to bite her nails: Mandy described experiencing a feeling of tension. This was the cue for the habit. After some discussion, they established that Mandy bit her fingers when she was bored, and after she had worked through all of her nails, she felt a brief sense of completion. The physical stimulation acted as the reward.

At the end of their first session, the therapist sent Mandy home with an assignment: Carry around an index card, and each time you feel the cue— a tension in your fingertips— make a check mark on the card. She came back a week later with twenty-eight checks. She was, by that point, acutely aware of the sensations that preceded her habit. She knew how many times it occurred during class or while watching television.

Then the therapist taught Mandy what is known as a “competing response.” Whenever she felt that tension in her fingertips, he told her, she should immediately put her hands in her pockets or under her legs, or grip a pencil or something else that made it impossible to put her fingers in her mouth. Then Mandy was to search for something that would provide a quick physical stimulation— such as rubbing her arm or rapping her knuckles on a desk— anything that would produce a physical response.

The cues and rewards stayed the same. Only the routine changed.

They practiced in the therapist’s office for about thirty minutes and Mandy was sent home with a new assignment: Continue with the index card, but make a check when you feel the tension in your fingertips and a hash mark when you successfully override the habit.

A week later, Mandy had bitten her nails only three times and had used the competing response seven times. She rewarded herself with a manicure, but kept using the note cards. After a month, the nail-biting habit was gone. The competing routines had become automatic. One habit had replaced another.


Say you want to stop snacking at work. Is the reward you’re seeking to satisfy your hunger? Or is it to interrupt boredom? If you snack for a brief release, you can easily find another routine— such as taking a quick walk, or giving yourself three minutes on the Internet— that provides the same interruption without adding to your waistline.

If you want to stop smoking, ask yourself, do you do it because you love nicotine, or because it provides a burst of stimulation, a structure to your day, a way to socialize? If you smoke because you need stimulation, studies indicate that some caffeine in the afternoon can increase the odds you’ll quit. More than three dozen studies of former smokers have found that identifying the cues and rewards they associate with cigarettes, and then choosing new routines that provide similar payoffs— a piece of Nicorette, a quick series of push-ups, or simply taking a few minutes to stretch and relax— makes it more likely they will quit.

For some habits, though, this is not enough. The alcoholics who replace their old behaviors with new ones may manage to stop drinking for a long while, until they run into some particularly stressful event in their lives. At this point, the stress becomes too much for many, who start drinking again. Not everyone does, however, and the difference seems to be in whether people are capable of genuinely believing that things will become better.

However, those alcoholics who believed, like John in Brooklyn, that some higher power had entered their lives were more likely to make it through the stressful periods with their sobriety intact.

It wasn’t God that mattered, the researchers figured out. It was belief itself that made a difference. Once people learned how to believe in something, that skill started spilling over to other parts of their lives, until they started believing they could change. Belief was the ingredient that made a reworked habit loop into a permanent behavior.

“I wouldn’t have said this a year ago— that’s how fast our understanding is changing,” said Tonigan, the University of New Mexico researcher, “but belief seems critical. You don’t have to believe in God, but you do need the capacity to believe that things will get better.

“Even if you give people better habits, it doesn’t repair why they started drinking in the first place. Eventually they’ll have a bad day, and no new routine is going to make everything seem okay. What can make a difference is believing that they can cope with that stress without alcohol.”

Appendix: A Reader's Guide to Using These Ideas

There isn't a single formula for changing habits, but rather thousands. Different people are driven by different cravings, and different habits require different approaches: stopping overeating is different from giving up cigarettes, which is different from how one communicates with their spouse. That said, the author attempts to provide a general framework for changing habits. It consists of four steps: Identify the routine, experiment with rewards, isolate the cue, have a plan.

The routine involved in the habit is usually the most obvious aspect. For example, maybe somebody always gets up from their desk at afternoon, walks to a cafeteria, buys a cookie, and eats it while chatting with friends. What exactly is the reward here? It could be the cookie itself, the change of scenery, the temporary distraction, the opportunity to socialize with colleagues, or the burst of energy that comes from the blast of sugar.

To identify the answer, one needs to experiment with rewards. On one day, instead of going out to a cafeteria, they might instead take a walk around the block. Another day, they might go to the cafeteria and buy an apple or chocolate bar and return to their desk without talking to anyone. On yet another day, they might walk to someone's desk to gossip for a few minutes and then return to work. When they do return to their desk, they should take a moment to quickly write down their thoughts or feelings - even just in the form of three random words in their head, like "relaxed", "saw flowers", "not hungry" - and then set a fifteen-minute alarm. If, after fifteen minutes, they still feel the craving, they know that whatever it was that they just did, it didn't give the desired reward. On the other hand, if they replaced the cafeteria visit by going to chat with a friend and the cafeteria craving vanished, then they've identified the reward as being a desire for temporary distraction and socialization.

Then there is the task of identifying the cue. Experiments have shown that almost all habitual cues fall into one of five categories:

  1. Location
  2. Time
  3. Emotional state
  4. Other people
  5. Immediately preceding action

So when one notices themselves engaging in a habit, they can write down the state of each of these variables. For example, here's one of the notes that the author made while trying to diagnose his own snacking habit:

Where are you? (sitting at my desk)

What time is it? (3:36 P.M.)

What's your emotional state? (bored)

Who else is around? (no one)

What action preceded the urge? (answered an e-mail)

After making such notes for three days, the pattern became clear: he got an urge to snack sometime between 3:00 and 4:00. The reward was temporary distraction, the kind that comes from gossiping with a friend.

Now he needed to have a plan for overriding the old habit with a new one, while maintaining the old cue and reward. So he wrote down the following:

At 3:30, every day, I will walk to a friend's desk and talk for 10 minutes.

To make sure I remembered to do this, I set the alarm on my watch for 3: 30.

It didn’t work immediately. There were some days I was too busy and ignored the alarm, and then fell off the wagon. Other times it seemed like too much work to find a friend willing to chat— it was easier to get a cookie, and so I gave in to the urge. But on those days that I abided by my plan— when my alarm went off, I forced myself to walk to a friend’s desk and chat for ten minutes— I found that I ended the workday feeling better. I hadn’t gone to the cafeteria, I hadn’t eat a cookie, and I felt fine. Eventually, it got be automatic: when the alarm rang, I found a friend and ended the day feeling a small, but real, sense of accomplishment. After a few weeks, I hardly thought about the routine anymore. And when I couldn’t find anyone to chat with, I went to the cafeteria and bought tea and drank it with friends.

That all happened about six months ago. I don’t have my watch anymore— I lost it at some point. But at about 3:30 every day, I absentmindedly stand up, look around the newsroom for someone to talk to, spend ten minutes gossiping about the news, and then go back to my desk. It occurs almost without me thinking about it. It has become a habit.



1: How effective is the AA? The book admits that the effectiveness is hard to evaluate, but notes that An estimated 2.1 million people seek help from AA each year, and as many as 10 million alcoholics may have achieved sobriety through the group. AA doesn’t work for everyone— success rates are difficult to measure, because of participants’ anonymity— but millions credit the program with saving their lives. It also comments that although scientists have been critical of the AA's unscientific methodology in the past, increasing numbers of researchers have recently become interested in the organization as its methodology fits other findings about habit change.