When Punishment Works Sometimes punishment is unquestionably effective. For example, punishment can deter some young criminals from repeating their offenses. A study of the criminal records of all Danish men born between 1944 and 1947 (nearly 29,000 men) examined repeat arrests (recidivism) through age 26 (Brennan & Mednick, 1994). After any given arrest, punishment reduced rates of subsequent arrests for both minor and serious crimes, though recidivism still remained fairly high. Contrary to expectation, however, the severity of punishment made no difference; fines and probation were about as effective as jail time. What mattered most was the consistency of the punishment. This is understandable in behavioral terms: When lawbreakers sometimes get away with their crimes, their behavior is intermittently reinforced and therefore becomes resistant to extinction.

Unfortunately, that is often the situation in the United States. Young offenders are punished less consistently than in Denmark, in part because prosecutors, juries, and judges do not want to condemn them to mandatory prison terms. This helps to explain why harsh sentencing laws and simplistic efforts to crack down on wrongdoers often fail or even backfire. Because many things influence crime rates—the proportion of young versus older people in the population, poverty levels, drug policies, discriminatory arrest patterns—the relationship between incarceration rates and crime rates in the United States varies considerably from state to state (King, Maurer, & Young, 2005). But international surveys find that, overall, the United States has a high rate of violent crime compared to many other industrialized countries, in spite of its extremely high incarceration rates.

When Punishment Fails What about punishment that occurs every day in families, schools, and workplaces? Laboratory and field studies find that it, too, often fails, for several reasons:

1. People often administer punishment inappropriately or mindlessly. They swing in a blind rage or shout things they don't mean, use harsh methods with toddlers, apply punishment so broadly that it covers all sorts of irrelevant behaviors, or misunderstand the proper application of punishment.

2. The recipient of harsh or frequent punishment often responds with anxiety, fear, or rage. Through a process of classical conditioning, these emotional side effects may then generalize to the entire situation in which the punishment occurs—the place, the person delivering the punishment, and the circumstances. These negative emotional reactions can create more problems than the punishment solves. A teenager who has been severely punished may strike back or run away. A spouse who is constantly insulted, belittled, and criticized will feel bitter and resentful and is likely to retaliate with small acts of hostility. And extreme punishment—physical abuse—is a risk factor, especially in children, for the development of depression, low self-esteem, violent behavior, and many other problems (Fréchette, Zoratti, & Romano, 2015; Gershoff, 2002; Widom, DuMont, & Czaja, 2007).

3. The effectiveness of punishment is often temporary, depending heavily on the presence of the punishing person or circumstances. All of us can probably remember some transgressions of childhood that we never dared commit when our parents were around but that we promptly resumed as soon as they were gone and reinforcers were once again available. All we learned was not to get caught.

4. Most misbehavior is hard to punish immediately. Punishment, like reward, works best if it quickly follows a response. But outside the laboratory, rapid punishment is often hard to achieve, and during the delay, the behavior may be reinforced many times. If you punish your dog when you get home for getting into the doggie biscuits and eating them all up, the punishment will not do any good because you are too late. Your pet's misbehavior has already been reinforced by all those delicious treats.

5. Punishment conveys little information. It may tell the recipient what not to do, but it does not communicate what the person (or animal) should do. Spanking a toddler for messing in her pants will not teach her to use the potty chair, and scolding a student for learning slowly will not teach him to learn more quickly.

6. An action intended to punish may instead be reinforcing because it brings attention. Indeed, in some cases, angry attention may be just what the offender is after. If a mother yells at a child who is throwing a tantrum, the very act of yelling may give him what he wants: a reaction from her. In the schoolroom, teachers who scold children in front of other students, thus putting them in the limelight, may unwittingly reward the very misbehavior they are trying to eliminate.

Because of these drawbacks, most psychologists believe that punishment, especially when it's severe, is a poor way to eliminate unwanted behavior. Consider spanking. A Canadian review of two decades of research found that although spanking may put a halt to a child's annoying or dangerous behavior in the short term, it backfires in the long term because children who are physically punished tend to become more aggressive and antisocial over time (Durrant & Ensom, 2012). Spanking is also associated with later mental health problems and slower cognitive development. No study has ever established a link between physical discipline and any positive outcome. Parents are starting to get the message. Fifty years ago, in many parts of the world, most parents thought that hitting their children was a good approach to correcting bad behavior, but today roughly half hold that belief (Holden et al., 2014; Roberts, 2000; Straus, 2005). A judge in Texas, a state that generally approves of corporal punishment for children, beat his teenage daughter with a belt because she had downloaded video games. She videotaped the beating and posted it on YouTube, where over 100,000 people saw it. As a result, her father was removed from office. To date, 32 countries, including nations as different as Denmark, Israel, Tunisia, Costa Rica, and New Zealand, have outright bans on hitting children.

In special cases, as when children with mental disabilities are in immediate danger of seriously injuring themselves or a school bully is about to beat up a classmate, temporary physical restraint may be necessary. But even in these cases, alternatives are often available. School programs have successfully reduced school violence by teaching kids problem-solving skills, emotional control, and conflict resolution, and by rewarding good behavior (Hahn et al., 2008; Wilson & Lipsey, 2007). And in some cases, the best way to discourage a behavior—a child's nagging for a cookie before dinner, a roommate's interruptions when you're studying—is to extinguish it by ignoring it. Of course, ignoring a behavior requires patience and is not always feasible. If your dog barks all day and night, telling your neighbors that it's best to ignore the racket will not be a winning strategy for long, even if you explain that you learned all about the drawbacks of punishment in your psychology class.

Finally, when punishment must be applied, these guidelines should be kept in mind: (1) It should not involve physical abuse; instead, parents can use time-outs and loss of privileges (negative punishers); (2) it should be consistent; (3) it should be accompanied by information about the kind of behavior that would be appropriate; and (4) it should be followed, whenever possible, by the reinforcement of desirable behavior. Learn more about how reinforcement and punishment affect behavior in the video Physical Punishment: You Decide.

Misuse of Rewards For many years, teachers everywhere have been handing out lavish praise, happy-face stickers, and high grades, even if students don't deserve them, in hopes that students' performance will improve as they learn to “feel good about themselves.” Scientifically speaking, however, this approach is misguided. Study after study finds that high self-esteem does not improve academic performance (Baumeister et al., 2003). Instead, academic achievement requires effort and persistence (Duckworth et al., 2011). It is nurtured not by undeserved rewards but by a teacher's honest appreciation of the content of a student's work, and specific constructive feedback on how to correct mistakes or fix weaknesses (Damon, 1995). These strong findings from psychological science have finally begun to influence some teachers, who are now shifting away from doling out unwarranted “self-esteem boosters” and focusing on helping students appreciate the benefits of effort and persistence.

One obvious result of the misuse of rewards in schools has been grade inflation at all levels of education. Grade inflation at universities began in the 1970s, during the Vietnam War, as a way to help many students avoid the draft. Since then, grades have risen steadily but graduation rates have not. Moreover, the literacy of graduates has declined, as have scores on entrance exams. And today, at many colleges and universities, Cs (which are supposed to mean average or satisfactory) are nearly extinct; in U.S. universities, 43 percent of grades are now As (Rojstaczer & Healy, 2012). One study found that a third of college students expected Bs just for showing up to class, and 40 percent felt they were entitled to a B merely for doing the required reading (Greenberger et al., 2008).

We have talked to students who feel that hard work should even be enough for an A. If you yourself have benefited from grade inflation, you may feel good about it—but remember that critical thinking requires us to separate feelings from facts! The problem is that rewards, including grades, serve as effective reinforcers only when they are tied to the behavior one is trying to increase, not when they are dispensed indiscriminately. Getting a good grade for “showing-up-in-class behavior” reinforces going to class, but not necessarily for learning much after you are there. Would you want to be treated by a doctor, represented by a lawyer, or have your taxes done by an accountant who got through school just by showing up for class? Or who did all the required reading, but without understanding it?

Why Rewards Can Backfire Most of our examples of operant conditioning have involved extrinsic reinforcers, which come from an outside source and are not inherently related to the activity being reinforced. Money, praise, gold stars, applause and hugs are all extrinsic reinforcers. But people (and probably some other animals as well) also work for intrinsic reinforcers, such as enjoyment of the task and the satisfaction of accomplishment. In real-world settings, extrinsic reinforcement sometimes becomes too much of a good thing: If you focus on it exclusively, it can kill the pleasure of doing something for its own sake.

Consider what happened in a classic study of how praise affects children's intrinsic motivation (Lepper, Greene, & Nisbett, 1973). Nursery school children were given the chance to draw with felt-tipped pens during free play and observers recorded how long each child spontaneously played with the pens. The children clearly enjoyed this activity. Then the researchers told some of the children that if they would draw with felt-tipped pens they would get a prize, a “Good Player Award” complete with gold seal and red ribbon. After drawing for 6 minutes, each child got the award as promised. Other children did not expect an award and were not given one. A week later, during free play, those children who had expected and received an award spent much less time with the pens than they had before the start of the experiment. In contrast, children in a control group who had neither expected nor received an award continued to show as much interest in playing with the pens as they had initially, as you can see in Figure7.5. Similar results have occurred in other studies when children have been offered a reward for doing something they already enjoy.

Why should extrinsic rewards undermine the pleasure of doing something for its own sake? The psychologists who did the pen study suggested that when we are paid for an activity, we interpret it as work instead of something we do because of our own interests, skills, and efforts. It is as if we say to ourselves, “Because I'm being paid, it must be something I wouldn't do if I didn't have to.” Then, when the reward is withdrawn, we refuse to “work” any longer. Another possibility is that we tend to regard extrinsic rewards as controlling, so they make us feel pressured and reduce our sense of autonomy and choice (“I guess I have to do what I'm told to do—but only what I'm told to do”) (Deci, Koestner, & Ryan, 1999). A third, more behavioral explanation is that extrinsic reinforcement sometimes raises the rate of responding above some optimal, enjoyable level, at which point the activity really does become work.

Findings on extrinsic versus intrinsic reinforcements have wide-ranging implications. Economists have shown that financial rewards can undermine ethical and moral norms such as honesty, hard work, and fairness toward others, and can decrease people's willingness to contribute to the common good (e.g., by paying taxes and giving to charity). In other words, an emphasis solely on money encourages selfishness (Bowles, 2008).

We must be careful, however, not to oversimplify this issue. The effects of extrinsic rewards depend on many factors, including a person's initial motivation, the context in which rewards are achieved, and in the case of praise, the sincerity of the praiser (Henderlong & Lepper, 2002). Sometimes extrinsic rewards can help boost achievement. When some U.S. high schools started offering large cash rewards to students who got high scores on Advanced Placement tests for college, achievement skyrocketed at inner-city schools. Suddenly, disadvantaged minority students were taking statistics classes on Saturdays and passing the placement tests in larger numbers. But these effects might be limited to a unique combination of already motivated teachers and students. In a study of more than 27,000 students in Dallas, New York City, and Chicago, students were paid to read books, to complete assignments, or for getting good grades. The financial incentives had no effect. Although the students were excited about the money, they did not have the basic study skills to build on in order to achieve the goals (Fryer, 2011).

As for the relationship between intrinsic and extrinsic reinforcement, in general, if you get praise, money, a high grade, or a trophy for doing a task well, for achieving a certain level of performance, or for improving your performance rather than for just doing the task, your intrinsic motivation is not likely to decline; in fact, it may increase (Cameron, Banko, & Pierce, 2001; Pierce et al., 2003). The rewards are apt to make you feel competent rather than controlled. And if you have always been crazy about reading or about playing the banjo, you will keep reading or playing even when you do not happen to be getting a grade or applause for doing so. In such cases, you will probably attribute your continued involvement in the activity to your own intrinsic interests and motivation rather than to the reward.

So, what is the take-home message about extrinsic rewards? First, they are often useful or necessary. Few people would trudge off to work every morning if they never got paid, and in the classroom, teachers may need to offer incentives to some students. But extrinsic rewards should be used carefully and should not be overdone, so that intrinsic pleasure in an activity can blossom. Educators, employers, and policymakers can avoid the trap of either–or thinking by recognizing that most people do their best when they get tangible rewards for real achievement and when they have interesting, challenging, and varied kinds of work to do.