At Penn State University, drinking is as much a part of college life as dining-hall food and choosing a major, which is one reason Penn State topped the 2009 Princeton Review list of best party schools. On Friday nights students line up outside of East Halls, a complex of freshman dorms, to take shuttle buses to fraternity parties across the campus. Many students have already “preloaded” in their dorms before going out, drinking shots of hard liquor or mixed drinks to prepare for the evening. When reporters from the radio show This American Life visited, and boarded the bus to interview students, they found one group of women who were on their second loop around campus because they had forgotten to get off at their stop the first time around. Many of these women were wearing cheap, ratty jackets over their mini–tube dresses, an incongruous combination until you realize what often happens to those jackets later—they get soaked with beer or worse. “They get like puked on—you don’t want that back,” said one student. These disposable garments have acquired a name of their own—“frackets.”

It is not just the students who drink. As at many universities, tailgating before football games is an institution at Penn State. Alumni gather as early as 8:00 a.m. on game days to start drinking and socializing. One Saturday, after a huge snowstorm, Penn State University did something it had never done before, at least for as long as anyone could remember: it canceled all tailgating outside of its football stadium before a game. Although conditions were treacherous, many fans were not deterred, tailgating in their homes or on porches instead. Several families took refuge on the fourth story of a parking garage, setting up their tables and grills in the frigid spaces next to parked cars.

One middle-aged couple was there celebrating their daughter’s twenty-first birthday by, to put it bluntly, getting her drunk. Another couple in their fifties, who were also tailgating in the garage, decided to join the fun by giving the birthday girl a drink from their stash, which she readily accepted. When a This American Life reporter asked the couple why they had joined the party, the husband replied, “We have to keep ourselves entertained somehow, right?”

“Watch the twenty-one-year-old get plastered, that’s the entertainment?” asked the reporter.

“Exactly, exactly!” the man replied, accompanied by gleeful giggles from his wife.

Before going any further, I should point out something that is probably obvious to many readers—people at Penn State have a lot of fun. This comes through clearly in the This American Life episode. Students sing together, party together, and blow off steam together, at a time in their lives when the burdens of adulthood are just over the horizon. As we’ve seen, hanging out with close friends is one of the most important ingredients of personal happiness, and although people don’t have to drink to achieve this, one has the sense that at Penn State it sure helps. As Ira Glass, the moderator of This American Life, put it, it seems like the motto at Penn State is “Drinking together is what we do.” Yes, some people overdo it, just as some people drive their cars too fast and cause accidents. But that doesn’t mean we should return to the days of Prohibition or the horse and buggy.¹

But drinking seems to have reached a new level from what I remember from college. In my day, there were keg parties and some people overindulged. Occasionally, someone drank to the point of getting sick, but not so often that we felt the need to wear frackets. Today, the number of students who abuse alcohol is staggering. About 40 percent of college students engage in binge drinking (the figure is above 50 percent at Penn State), which the National Institute on Alcohol Abuse and Alcoholism defines as consuming enough alcohol to reach a blood alcohol concentration (BAC) of .08 percent or above. The number of drinks it takes to reach this level depends on your gender and body weight, and how long you have been drinking, but a good rule of thumb is five drinks in a two-hour period for men and four drinks in a two-hour period for women. People who binge-drink have an increased prevalence of serious injuries, accidents, cardiovascular disease, liver damage, neurological damage, sexually transmitted diseases, unwanted pregnancies, and sexual dysfunction.²

At Penn State, one-quarter of the students say that drinking has interfered with their academics. For some, it even interferes with their ability to find their way home. It is not unusual for intoxicated students to enter the houses of town residents and pass out on empty beds. So common, in fact, that the local police have a name for it: “drunk in the house.” And, like many college campuses, Penn State has its share of medical problems due to alcohol poisoning. In the first three months of 2007, for example, 109 students were taken to the emergency room with an average blood alcohol level of .245—three times the legally defined level of intoxication.

Many kids start to drink long before they get to college. Seventy-five percent of teens have tried alcohol at least once, and 23 percent have started drinking before the age of thirteen (think about it—one out of every four thirteen-year-olds you see at the mall has already begun to drink). On a recent survey, one in four high school students said they had engaged in binge drinking at least once in the previous thirty days and nearly one in three said they had driven with someone who had been drinking. Car accidents are the leading cause of death for teenagers in the United States, and 23 percent of teens who die in crashes have blood alcohol levels of .08 grams percent or more (.08 is the legal limit in all fifty states).³

Numbers only tell part of the story, of course, and one only has to type “teenage drinking stories” into a Web browser to read tragic accounts of what alcohol has done to the lives of many young people. There is no point in repeating those stories here; I imagine that readers are all too familiar with the havoc alcohol abuse can cause, and many have undoubtedly lost loved ones in drunk-driving accidents or other alcohol-related tragedies.

There is one death, however, that was so senseless and tragic—and close to home—that I haven’t been able to get it out of my mind. It happened in 1997 at the University of Virginia, where I teach. Now, before I talk about this incident, I feel compelled to point out that the University of Virginia (UVa) has done a lot to combat alcohol and drug abuse over the years. A 1996 report identified UVa as having one of the twelve best campus alcohol strategies in the nation. In 2003, John Casteen, then president of UVa, was the first recipient of the President’s Leadership Group Award, given by the Higher Education Center for Alcohol, Drug Abuse, and Violence Prevention to college and university presidents who have excelled in combating alcohol and drug abuse on their campuses. President Casteen initiated a number of programs, including one that is now named the Center for Alcohol and Substance Education in the Dean of Students Office. In short, UVa has done more than many universities to reduce alcohol and drug abuse.

But in any large group of eighteen-to-twenty-two-year-olds, there will be some who abuse alcohol. This is epitomized at UVa in a practice called the fourth-year fifth, whereby some seniors attempt to drink a fifth of alcohol on the day of the last home football game. To put it in perspective, that’s seventeen one-and-a-half-ounce shots of alcohol. A 185-pound male who drank that much in three hours would have a blood alcohol level of .33—close to the level at which coma and death occur. A 125-pound woman who drank that much in three hours would have an astounding blood alcohol level of .57, well above the level that causes death.

Most of the students who subject themselves to this self-poisoning survive it, probably by spreading out the drinking over several hours. But not everyone does. Leslie Baltz was a talented art major at UVa who had spent a semester abroad in Florence and was writing an honors thesis on the sculptor Elie Nadelman. She had been on the dean’s list every semester and her GPA in her major was a nearly perfect 3.97. By some reports, Leslie was not much of a drinker. But on November 29, 1997, the Saturday of the last UVa home football game, Leslie was drinking heavily. It is unclear whether she was attempting the fourth-year fifth, but by the time the football game started she was so intoxicated that her friends left her on a couch in an upstairs apartment. When they returned that evening, they found Leslie comatose at the bottom of the stairs, where she had fallen headfirst. Her blood alcohol level was .27.

That same afternoon, another UVa senior, Ryan Dabbieri, went to a tailgate party and started drinking shots of bourbon. By the time he got to the football game, he was so drunk that he began vomiting in the stands. Two female friends managed to get him to their car, intending to drive him home and put him to bed. But one of the women, who had had first-aid training as a lifeguard, changed course at the last minute and took Dabbieri to the emergency room. It was a decision that saved his life, because just after she and her friend maneuvered Dabbieri into a wheelchair and pushed him through the doors of the ER, he stopped breathing. The doctors managed to get him on a respirator and resuscitate him, and by the next day he was out of danger. Not so Leslie Baltz, who lay critically ill with severe head injuries in the same intensive care unit where Dabbieri was being treated. Her injuries were not survivable, and Baltz’s parents made the wrenching decision to take her off life support that Sunday and signed the forms to donate her organs. Her body was cremated later in the week.

Thankfully, the allure of the fourth-year fifth appears to be diminishing. One survey found that 13 percent of fourth-year students attempted it in 2001, but that number had dropped to 10 percent in 2004. In 1991, Peer Health Educators—students trained to educate their classmates about health matters—started an alternative activity called the Fourth-Year 5K, a benefit race that takes place on the morning of the last home football game. In 2007, 143 fourth-year students participated in the race. But, if the 10 percent figure is correct, then twice as many of their classmates were drinking themselves into a stupor (and possibly an emergency room visit) that very same day.⁴

DON’T YOU D.A.R.E.

I wonder how many of those UVa students had D.A.R.E. T-shirts tucked away in their dressers at home, left over from middle school? D.A.R.E. (Drug Abuse Resistance Education) is by far the most common school-based program to combat the use of alcohol, drugs, and tobacco. It began in 1983 in Los Angeles, as a joint venture of the police department and school system. Police officers from the LAPD came into elementary schools for an hour a week and taught kids about the dangers of alcohol and drugs and how to resist peer pressure. The program soon achieved phenomenal popularity. It was expanded to junior high and high school grades and was adopted by school districts throughout the country, including the schools my children attended. Both of my kids sported black D.A.R.E. T-shirts that they received after completing the program, and they were not alone. According to the D.A.R.E. website, 75 percent of the school districts in the United States use the program, and it has spread to forty other countries as well. Recent presidents of the United States seem to agree that D.A.R.E. is a fine program; they have all designated a day in April as National D.A.R.E. Day.⁵

There is only one problem—D.A.R.E. doesn’t work. A number of well-controlled experiments have shown that the program has no effect on students’ use of tobacco, alcohol, or other drugs. In 2003, the United States General Accounting Office surveyed the evidence and reached the same conclusion. (Apparently, the White House doesn’t pay close attention to reports from the General Accounting Office, or at least not this one.) Amazing amounts of time, effort, and money (more than one billion dollars annually) have been devoted to a program that was not adequately tested and, when it was, turned out to be ineffective.⁶

Some D.A.R.E. officials have chosen not to believe the scientific evidence. When fifth graders at the St. Joseph School, in Medford, Massachusetts, attended a D.A.R.E. graduation ceremony, Domenic DiNatale, the Massachusetts state D.A.R.E. coordinator, asked the children, “When people say D.A.R.E. doesn’t work, what do you say?” Unsatisfied with the muted responses, he said, “Louder!” To which the children replied in unison, “It does!”⁷

But shouting won’t make it work, and in 2001 the D.A.R.E. organization decided to revamp the program. With funding from the Robert Wood Johnson Foundation, D.A.R.E. developed the Take Charge of Your Life curriculum, which was tested experimentally in six U.S. cities. Eighty-three schools were randomly assigned either to a group that used the new curriculum or to a control group that did not. The results were decidedly mixed. Among students who were already using marijuana, those who took part in the program showed a greater decrease in use of that drug than did students in the control group. That’s the good news. Among students who were not smoking cigarettes or drinking alcohol at the beginning of the study, those who took part in the D.A.R.E. program were more likely to be smoking and drinking by the eleventh grade than were students in the control group. This qualifies the program for my bloodletting award, because it seems like it should work but actually does more harm than good. Based on these results, D.A.R.E. abandoned the Take Charge of Your Life curriculum in favor of yet another approach called keepin’ it REAL, which has yet to be tested.⁸

On the one hand, I admire the D.A.R.E. organization for (finally) paying close attention to the evidence and trying new approaches. But doesn’t it seem odd that a program would first be implemented in 75 percent of our schools and then tested to see if it works? This is especially mind-boggling given that there are many other programs that have been proven to work. Although most of these programs were not conceived with the story-editing approach in mind, we will see that they have common elements that are consistent with that approach. And, by looking at them carefully, we will get some clues about why the D.A.R.E. program has failed so miserably.

LIFESKILLS TRAINING

LifeSkills Training (LST) was developed in the early 1980s as an antismoking program and was soon expanded to target the use of alcohol and drugs as well. It is a school-based program that begins in the sixth or seventh grade, the point at which kids are most likely to begin experimenting with alcohol, cigarettes, and drugs. Students attend fifteen class sessions of forty-five minutes each, followed by ten sessions the next year and five sessions the year after that. The curriculum is multifaceted, including units on personal self-management skills, social skills, and information about various drugs and their dangers. In the personal self-management unit, for example, the students engage in a self-improvement project in which they choose something about themselves that they want to change and then work on doing so. In the social skills unit, students work on how to get along better with others and receive tips on how to overcome shyness and be more assertive. In the drug-related information unit, students learn about drugs and how to resist influence from their peers and the media (e.g., cigarette advertising). In addition, students receive information about two kinds of social norms: the actual levels of drug use by adolescents and adults (which are often lower than people think) and the fact that social approval of cigarette smoking and other drug use is declining. As we will see shortly, conveying these norms to people is an important part of story-editing interventions.

LST has been evaluated extensively with positive results. One study took place in fifty-six suburban junior high schools in New York State, some of which were randomly assigned to implement the LST program and others to a control condition that did not. Students who took part in LST were significantly less likely to smoke cigarettes, drink alcohol, or use marijuana or other drugs at the end of high school. Similar results were found in a study that took place in inner-city schools in New York City that primarily enrolled African American and Hispanic students.⁹

LST helps kids avoid alcohol, tobacco, and other drugs in at least three ways. First, it is not enough to tell kids to “just say no”; it’s important to teach them how to say no, and LST does this by teaching assertiveness skills. Second, the program imparts a general sense of well-being in kids, which makes them less interested in using drugs. And finally, as mentioned, the program helps combat the notion that using alcohol, tobacco, and other drugs has social benefits. That is, kids who take part in the program are less likely to believe that their peers think it is cool to smoke and drink and use drugs.

PROJECT TOWARDS NO DRUG ABUSE

Project Towards No Drug Abuse (Project TND), developed in the 1990s, targets high school students. Students attend twelve class sessions over a four-week period. Like LifeSkills Training, Project TND is multifaceted; the twelve sessions attempt to explode myths about drugs, improve kids’ assertiveness and decision-making skills, and correct kids’ impressions about the popularity of drugs. The program has been evaluated in at least three experimental studies in which schools were randomly assigned to receive Project TND or to control groups that did not. These studies found that teens who participated in Project TND were significantly less likely to smoke and use marijuana and were somewhat less likely to drink alcohol than were teens in the control schools.

Enthusiasm about these results must be tempered, however, by a recent failure to replicate them in high schools for at-risk kids in California. Researchers randomly assigned classes to one of three conditions: one received Project TND; one received a new version of Project TND in which small groups of students were formed and led by student leaders who had been nominated by their peers; and one was a control group in which there was no program. The teens who received the standard version of Project TND were no less likely to smoke, drink, or use other drugs than kids in the control condition at a one-year follow-up. It is not clear why Project TND didn’t work in this study when it did in the earlier investigations, especially given that some of the earlier studies also included schools for at-risk kids.

The results for the teens in the peer-led version of Project TND were even more interesting. It turned out that this program helped some kids and hurt others, depending on the extent to which the teens’ friends were substance abusers. If teens hung out with friends who did not smoke, drink, or use drugs, then participating in the peer-led Project TND curriculum was beneficial: they became even less likely to use these substances, relative to teens in the control condition. But if teens hung out with friends who did smoke, drink, and use other drugs, then participating in the peer-led Project TND curriculum backfired: it made them more likely to use these substances. The researchers refer to this as a “peer acceleration” effect, because the peer-led substance-abuse program seems to have accelerated the influence of the peer group. If a kid’s friends were on the straight and narrow, he or she became even less likely to imbibe. But if a kid’s friends were smokers, drinkers, or potheads, the program made him or her even more likely to join them. This finding is reminiscent of the main conclusion from the previous chapter on teen delinquency: when teens hang out with other teens who are engaged in risky behaviors, they tend to imitate those risky behaviors.¹⁰

MIDWESTERN PREVENTION PROJECT

As powerful as school-based programs can be, they are obviously limited by the fact that they can reach kids for only a brief amount of time in the school environment. What if we targeted teens’ parents and the general community as well? This is the goal of the Midwestern Prevention Project (MPP), an ambitious intervention developed in the 1980s. It includes a school-based program like that offered by LST, namely, one that begins in the sixth or seventh grade, teaches kids how to resist peer pressure, and conveys information about the dangers of drugs. MPP goes beyond LST, however, by including interventions outside of the school. There are sessions for parents that teach communication and child-rearing skills, plus community and government leaders are consulted and encouraged to offer drug prevention services and to coordinate those services across community agencies. Finally, there is media coverage of the program in the form of public service announcements, press conferences, newspaper articles, and interviews with the staff on television and radio.

The main evaluation of the program was conducted in the Kansas City school system, which randomly assigned schools either to receive MPP or to a control group that did not. By the end of their senior year, teens who took part in the program were smoking less, using less marijuana, and, to some extent, drinking less alcohol. A follow-up was conducted in schools in the Indianapolis area a few years later with some modifications. Again, the program was effective, especially among sixth and seventh graders who were already smoking, drinking, or using marijuana at the outset of the program. Like LST, the program works in part by changing teens’ perceptions of how popular smoking and drinking are among their friends. They come to believe that their friends would be less receptive to these activities, which makes them less inclined to smoke or drink.¹¹

SOCIAL INFLUENCE INSIDE OUR HEADS

The success of the three programs just reviewed is no doubt due to the fact that they are multifaceted. Because substance abuse has many causes, it is important to attack it from many angles, and each of the programs does just that. One of the main reasons that kids begin smoking or drinking, however, is because their friends pressure them into it. Virtually all teens are desperate for approval from their friends, and if Tyler or Taylor, the coolest kid they have ever met, gives them a cigarette and dares them to smoke it, it can be incredibly hard to say no. Many teens succumb to pressure from their friends to smoke, drink, take drugs, have sex, drive too fast, and any number of other activities that make their parents’ hair stand on end. It is thus no coincidence that all the effective programs attempt to blunt peer influence by providing assertiveness training, role-play exercises, and the like.

But there is another kind of peer pressure that can be equally powerful, and that is the pressure that exists inside our heads. Yes, we may find ourselves in situations in which Tyler is twisting our arms to get us to smoke or drink. But there are plenty of times when Tyler isn’t around and yet we imagine what he would want us to do. We all have perceptions of what the social norms are, namely, what we think the people we care about are doing (“Tyler is probably out partying as we speak”) and what we think they would want us to do (“Tyler would think it would be really cool if I got drunk”). Even when no one is around to twist our arms, we are subject to imaginary arm-twisting by conforming to what we think others would want us to do. One way to reduce smoking and drinking, then, is to change people’s perceptions of what others approve of—in other words, to use the story-editing approach to change people’s perceived social norms.

TYLER AND TAYLOR AREN’T DRINKING AS MUCH AS YOU THINK

Most of the programs just reviewed try to change kids’ impressions of what their peers are doing and thinking and correct misperceptions about the popularity of drugs and alcohol. But because these programs are multifaceted, we can’t be sure how much of their success is due to this kind of story editing. An intervention that has been used primarily with college students, called the social norms approach, focuses exclusively on students’ perceptions of what their peers are doing and thinking, with some success. The idea is that if we can correct college students’ misperceptions about the prevalence of drinking, maybe that will reduce the pressure in their heads to follow suit.

On a recent survey, for example, 75,000 students at 130 colleges reported how many alcoholic drinks they had consumed the last time they had “partied” or socialized. The amount of drinking varied widely at the different colleges, with the average number of drinks ranging from zero to seven. Most schools (56 percent) fell in the three-to-four-drink range. What was especially interesting was that students at virtually all the campuses overestimated how much their peers drank. When asked to guess how many drinks the typical student at their institution had consumed the last time he or she “partied” or socialized, most students—71 percent—overestimated the amount of drinking on their campus. Only 14 percent guessed the correct number (15 percent guessed too low).

This tendency to overestimate how much one’s fellow students drink has been found on survey after survey, on virtually all college campuses, among virtually all subgroups of students (e.g., men and women, different racial groups, students who are or are not in fraternities or sororities). Students also tend to overestimate how much their peers use other recreational drugs. Why do students get it wrong? Those who drink excessively stick out more than those who drink moderately (or not at all), and we thus are more likely to notice and remember them. When walking across campus on a Saturday night, for example, we are more likely to notice the inebriated students who are shouting loudly and carrying on than the sober students who are walking quietly to their dorms. Also, the overt act of drinking is easier to observe than someone’s private doubts about the practice: just because we see someone drink a couple of beers at a party doesn’t mean that that person is comfortable with what he or she is doing or will necessarily drink at the next party. And finally, films, television shows, and advertisements often glorify alcohol use and convey the impression that it is more common than it actually is.

For all these reasons, students overestimate how much others drink. Given the power of peer pressure, this mistaken belief can be quite consequential. In the survey of 75,000 college students, for example, the researchers looked at who drank the most. Some of the results were just as we would expect—members of fraternities or sororities drank more than nonmembers, men drank more than women, and students who attended schools where there was a lot of drinking drank more than students who attended schools where there was less drinking. By far the biggest predictor, however, was how much students thought their peers drank, independent of the actual amount of drinking at their school. Thus, if you want to know how much college students drink, ask them how much they think their peers drink.¹²

In other words, many people overestimate how much their peers drink, which makes them drink more than they otherwise would in order to conform to this imagined (but mistaken) norm. This doesn’t mean that everyone drinks as much as they think their peers are drinking; if that were the case, then the real norm at their university would be the same as people imagined it to be. For example, if people believe that the norm is to have five drinks when socializing, and as a result they have five drinks, then the average would be five drinks and people’s perceptions of the norm would be accurate. Instead, people ordinarily might not drink at all or have one or two drinks, but because they think other students are having five drinks, they up their consumption to three or four to be closer to the norm. If so, then we might succeed in reducing drinking by correcting students’ impressions about how much their peers drink.

This is the reasoning behind the social norms approach, which has become a popular way of addressing alcohol abuse at colleges and universities. College administrators try to disseminate information about the actual amount of drinking on their campuses in the hope that this will correct students’ misperceptions and that they will drink less as a result. One reason for the popularity of this approach is that it can be administered cheaply on a mass scale. Administrators place ads in college newspapers, distribute computer mouse pads with messages printed on them, and put posters up around campus (an example of one used at the University of Virginia is shown on the next page). In one particularly creative campaign at the University of Virginia, called the Stall Seat Journal, posters are placed on the insides of doors in bathroom stalls around the university.¹³

It will come as no surprise to readers that the social norms approach was implemented widely before it was adequately tested. In recent years, though, there have been some rigorous experimental tests of its efficacy. One study was conducted at eighteen colleges and universities in the United States. First, a random sample of students at each institution completed a survey on which they indicated how much alcohol they typically drink. Then, half the institutions (randomly assigned, of course) conducted a social norms campaign for three consecutive semesters, in which they publicized the average amount of drinking on their campus using newspaper ads, posters, and e-mails. Finally, students at all the institutions were surveyed again about how much they typically drink. The results showed that the social norms campaigns worked: students at schools that implemented the campaigns, compared to students at the schools that did not, lowered their perception of how much their peers drink and drank less alcohol themselves.

So far so good. Unfortunately, though, a follow-up study by the same team of researchers failed to replicate these encouraging results. The researchers selected an additional fourteen colleges and universities and again randomly assigned half to a treatment condition that implemented a social norms campaign for three semesters, and half to a control condition that did not. Even though the seven institutions in the treatment condition used social norms campaigns that were as intensive as in the previous study, these campaigns did not succeed in changing students’ estimates of how much their peers drink, nor did they lower the amount of actual drinking. It is unclear why the social norms campaigns failed at the new sample of universities, though one clue comes from the fact that the amount of drinking at the new sample of universities was higher than at the original universities. Perhaps the more people drink initially, the harder it is to correct their misperceptions about how much their peers drink. Another possibility is that some of the universities may not have publicized their social norms programs widely enough.¹⁴

Perhaps the best summary of social norms campaigns is that they are promising but need to be refined and tested further. One problem with disseminating information through posters and the media, for example, is that these messages are easy for students to ignore. And even if students do pay attention, they might not connect the information to themselves. An alternative approach is to present people with individualized feedback, rather than counting on them to see an ad in the newspaper or a poster in a campus building. This has been done by having students answer questions, anonymously on a computer, about how much they drink and how much they think others drink. They then receive feedback about exactly how far off their estimates are. Some colleges have tried this approach with success.¹⁵

Another problem with the social norms approach is that the general message might not be appropriate for all students. Remember the survey of 75,000 students? Although most students overestimated the amount of drinking on their campus, 15 percent of the students underestimated the amount of drinking. If a social norms campaign were conducted on their campus, these students would discover that other students actually drink more than they thought, which might not have the desired effect. “Wow,” these students might think. “If I want to be popular, maybe I should start drinking as much as Tyler and Taylor do.”

To illustrate this point, I need to digress for a moment to another topic, namely, how to get people to conserve energy. Researchers in California used the social norms approach to try to get people to reduce their use of electricity. They took readings from the electricity meters at 290 houses twice within a two-week period, in order to get a baseline measure of how much electricity each house was consuming. Then they left a flyer on the doorknob of each house that showed how much electricity that household had been using and the average amount of electricity that their neighbors were using. Think about what this would be like: you come home one day and see the flyer on your doorknob, and read that you are using more electricity than your neighbors. “Whoa,” you might think. “I guess I’m more of an energy hog than I thought.” This probably makes you feel a little embarrassed, and so you stop leaving lights on when you leave a room and maybe even use your air conditioning a little less. This is just what the researchers found: people who discovered that they were above-average electricity users decreased their use of electricity over the next few weeks.

But what about the people who found out that they were using less electricity than their neighbors? The feedback had the opposite effect, leading to an increase in power use. “Why should I skimp on the air conditioning,” these folks seemed to say, “when the Joneses and the Smiths are pumping out a lot more cool air than I am?” Thus we see the danger of social norms campaigns: they can backfire among people who find out that they are doing better than average. Perceived norms are a powerful thing. If we think we’re conserving more energy than others, we slack off on our electricity use; if we find out we are drinking less than others, we might down a few more beers at the next party.¹⁶

WHAT DO TYLER AND TAYLOR WANT ME TO DO?

Maybe we can get around this problem by examining people’s perceptions of what their peers want them to do, in addition to what they think their peers are actually doing. Social psychologists have, in fact, distinguished between these two kinds of social norms. Up to now, we’ve been talking about descriptive norms, which are people’s perceptions of how others actually behave in a particular situation (e.g., how much others are drinking or how much electricity our neighbors are using). The alcohol studies we’ve seen so far address descriptive norms by trying to correct students’ beliefs about how much their peers drink.

Injunctive norms refer to people’s perceptions of what others approve or disapprove of in a given situation (e.g., the amount of drinking or electricity use that we think our peers find acceptable and approve of). Often, descriptive and injunctive norms are the same: for example, few people crack open a beer and take a swig in the middle of a church service (the descriptive norm is low) and if they did they would be subjected to icy stares and verbal rebukes (there is an injunctive norm against such behavior). But just because a behavior is unusual doesn’t mean people disapprove of it: for example, most people don’t eat a salad for breakfast, but no one would particularly care if they did. And just because a behavior is common doesn’t mean that people approve of it. It seems like more and more drivers in my town run red lights, for example, something that most of us disapprove of. Thus, descriptive and injunctive norms are not always the same and it might help to convey information about both when trying to change people’s behavior.

Maybe we could bolster social norms campaigns by disseminating information not only about what people do (they drink less than you think), but also about what people approve of (your peers don’t condone drinking as much as you think). Would this be a more powerful way to change their behavior? The study of electricity consumption in California suggests that it might be. With an additional group of homes in that study, the researchers made a small addition to the flyer they left on the doorknob: if the homeowner was using less electricity than average, the researchers drew a little happy face on the flyer (), whereas if the homeowner was using more electricity than average, the researchers drew a sad face ().

In this simple way, the researchers conveyed an injunctive norm—people approve or disapprove of what you are doing—and this mattered a lot for the people who had been using less electricity than average. In addition to finding out that they were energy savers compared to their neighbors, they also got a little pat on the back, a smiley face that conveyed the message, “Good job! We approve!” This was enough to keep them on track: they maintained their low usage over the next few weeks, unlike those who got the feedback alone (without the smiley face) and took this as license to become energy hogs. The sad face didn’t have any effect among those who were using more electricity than average, because finding out that they were energy hogs was enough to shape them up all by itself.

Would conveying information about injunctive norms work in a social norms campaign on college campuses? I doubt that many students would be swayed by smiley or sad faces on posters, but it shouldn’t be that hard to find more effective ways of conveying the message that many students disapprove of binge drinking. Some social norms programs have in fact begun to include information about injunctive norms, and while the jury is out on how well this approach works, there is some evidence that this approach is effective.

One study, for example, conveyed information about injunctive norms by associating drinking with an unpopular group, thereby implying that only “uncool” people like to drink a lot. Researchers at Stanford University figured that many students beginning their first year would not want to be associated with older, geeky grad students who spend all their time in the library. At the beginning of the academic year, the researchers posted flyers in a freshman dorm that showed a graduate student holding an alcoholic drink. “Lots of graduate students at Stanford drink and lots of them are sketchy,” the flyer said. “So think when you drink… Nobody wants to be mistaken for this guy.” In another freshman dorm, the researchers posted flyers that portrayed the negative health effects of drinking alcohol but made no mention of graduate students. Then, two weeks later, the students in both dorms completed a survey on which they reported how many alcoholic drinks they had had in the previous week. They also indicated how much they would want other people to think that they were like graduate students.

It turned out that the “geeky graduate student” flyer had a dramatic effect on the drinking of students who did not want to be associated with graduate students. Among this group of freshmen, those in the “geeky” flyer condition had only had two drinks the previous week, whereas those in the health flyer condition had six drinks the previous week. (The flyers didn’t have any effect on freshmen who admired grad students; interestingly, they didn’t drink much to begin with.) Young people have a keen eye for what their peers approve of, and associating drinking with a geeky, disapproved-of group proved to be a powerful deterrent.¹⁷

Two of the programs we reviewed earlier that worked with high school students did so in part by changing injunctive norms. Teens who took part in the LifeSkills Training program were less likely to believe that their peers think it is cool to smoke, drink, and use drugs. And one way the Midwestern Prevention Project works is by changing teens’ perceptions of how receptive their friends would be to smoking and drinking. The teens come to believe that their friends would be less receptive, which makes them less likely to imbibe.¹⁸

These findings offer a clue about why the D.A.R.E. program does not work. A distinctive feature of the D.A.R.E. program is that it is run by police officers from the local community, whereas the LifeSkills Training Program and the Midwestern Prevention Project are administered by teachers. Now, on the one hand, bringing police officers into the classroom in a teaching role might be a wonderful way to improve relations between the police and community members and break down kids’ stereotypes about the police. But if the goal is to change kids’ beliefs about what their peers approve or disapprove of, police officers may not be in the best position to accomplish this goal. Think about a kid who is already feeling a bit alienated and believes that smoking or drinking is going to make him popular with the Tylers and Taylors in his class—the cool kids. It won’t be an easy sell for a police officer to convince him otherwise, because to an at-risk, disengaged kid, the police represent the authority they are rebelling against. It may be easier for a teacher, armed with an effective curriculum, to reach these kids.

I don’t mean to suggest that police officers could never be effective in the classroom. But the fact remains that D.A.R.E. has had little success in reducing alcohol, tobacco, and drug use, whereas LST and MPP are effective. And these successful programs work in part by convincing kids that drinking, smoking, and taking drugs are not going to make them as popular as they think.

VOLUNTEER WORK AND SUBSTANCE ABUSE

What else does the story-editing approach have to offer, other than targeting teens’ interpretations of what their peers are doing and thinking? In chapters 5 and 6, we saw that teens who feel disengaged and alienated are especially likely to become pregnant and commit violent acts. We also saw that an effective way of changing their self-views, making them feel more engaged and connected to their communities, is to get them to do community service. Might this also work to reduce substance abuse? No one would argue that getting teens to volunteer at a nursing home is the sole solution to drinking, smoking, and drug abuse. But it might help.

One study, for example, found that teens in inner-city schools who believed that they had few prospects in life—that is, that they had few friends they could count on and had little chance of going to college or getting a good job—were especially likely to engage in binge drinking. And the more they engaged in binge drinking over time, the more their view of their prospects sank. These kids were caught in a self-defeating thinking pattern that we’ve encountered several times in this book.¹⁹

I don’t know of any research that has tested the hypothesis that engaging in community service will reduce substance abuse. There is a promising hint from one study, however, that surveyed more than 27,000 students at 119 colleges. The more volunteer work students reported doing, the less likely they were to say they abused alcohol. Although this is a correlational finding, it at least raises the possibility that encouraging students to volunteer will reduce problem drinking.

There was another intriguing finding from this study. Independent of the amount of volunteer work students themselves did, the amount of volunteer work that other students at their university did predicted lower alcohol use. Students who attended universities at which there was a culture of volunteering were less likely to abuse alcohol than were students who attended other universities. Thus we can predict whether a given college student abuses alcohol, at least to some extent, by knowing how much other students at his or her university volunteer, independent of how much that particular student volunteers. Colleges and universities have unique cultures, and ones that value volunteer work, it seems, devalue alcohol abuse. (Given that I focused earlier on alcohol abuse at the University of Virginia, I should mention that (a) most students at UVa don’t abuse alcohol and (b) many engage in volunteer work. In fact, in 2009, UVa won an award from the Corporation for National and Community Service for its emphasis on community service and for having a high percentage of students who engage in service activities.)

The fact that an individual student’s decision about how much to drink is influenced by the culture at his or her university is another indication that people are very sensitive to what their peers think and do. Imagine a student who finds out that many of her friends and acquaintances are going to spend spring break doing a service project in New Orleans instead of partying on the beaches of Florida. This student has learned something important about the norms at her university, norms that play a very important role in shaping her behavior.

USING IT

I hope that parents who have read this chapter will be proactive in a number of ways. If your kids go to one of the 75 percent of schools that has adopted the D.A.R.E. program, you might contact your school administrators and ask politely, “Why are you using a program that doesn’t work, when you could easily be using a program that does?” You might direct them to the websites of programs such as LifeSkills Training (http://www.lifeskillstraining.com/).

Another clear message from this chapter and the two previous ones is that parents and teachers should do what they can to prevent kids from feeling alienated and disengaged. This is hardly news, though the ways of accomplishing it—such as getting kids involved in community service—are. Another solution, as we’ve seen, is to correct misperceptions about two types of social norms: what kids think their peers are doing and what they think their peers approve of. If parents and teachers can set kids straight about the prevalence of smoking, drinking, and drug use (or the lack thereof), these activities are likely to seem less alluring.

How can we best accomplish this? No one said it was easy, but there are things we can do. One way that kids learn about what is cool is through television and movies, many of which, unfortunately, glamorize smoking and drinking. About 70 percent of movies made in the United States depict someone smoking, and it is often the sexy, affluent, glamorous stars who light up and appear to enjoy it (we don’t see their yellowing teeth or hear their hacking coughs). And research shows that the more kids watch movies in which people smoke, the more likely they are to try smoking. One study randomly assigned ninth graders to see the movie Reality Bites, either with or without the smoking scenes edited out. In the unedited version of the movie, the lead characters, played by Winona Ryder and Ethan Hawke, smoke in several scenes. In the edited version, the researchers removed these scenes while preserving the story line. The kids who saw the unedited version—with the smoking scenes—reported that smokers were cooler people and expressed more of an intention to smoke themselves than did the kids who saw the edited version.

What, then, can parents do? We can monitor the movies our kids see and boycott the ones that glamorize smoking and drinking. We can also lobby the movie industry to end this glamorization (see, for example, http://smokefreemovies.ucsf.edu/). If our kids do watch movies that glamorize smoking, we can talk to them about the dangers of smoking in advance and show them antismoking ads. The experiment mentioned above, for example, included an additional condition in which ninth graders first saw an antismoking advertisement and then saw Reality Bites with the smoking scenes. The antismoking advertisement worked as a prophylactic: the kids in this condition did not glamorize smoking and reported no more of an intention to smoke than did kids who saw the movie with the smoking scenes removed.²⁰

Kids don’t learn about social norms only from the movies, of course; they are hugely influenced by their peers. We might be tempted to lock our kids in their rooms until they turn twenty, but that is hardly the solution. After all, part of growing up is finding close friends we value, trading information about music, books, and movies, and carving out a niche for ourselves that is independent of our parents. Would you rather that your fourteen-year-old son have a tight-knit group of friends who influence his tastes and preferences, or that he spend every weekend night holed up in his room by himself? Of course we want our kids to have friends, but peer influence comes with the territory. What we don’t want is for kids to have the wrong friends—ones who feel like they have no future or prospects and will tempt our kids with alcohol, cigarettes, and drugs. And although we can’t control this completely, we can monitor who our kids are hanging out with and, above all, talk with our kids. You can ask them how many of their friends smoke and drink, and while you might not always get an honest answer, you can point out that for every kid who is smoking and drinking, there are probably several others who are uncomfortable with it but afraid to speak up. In essence, you could conduct your own mini–social norms campaign, in which you correct any misperceptions your kids have about how cool their peers think smoking and drinking is.

CHAPTER 7

Everybody’s Doing It… Or Are They?

Reducing Alcohol and Drug Abuse