Education Working Paper
No. 4 January 2004
Sex, Drugs, and Delinquency
in Urban and Suburban Public Schools
Jay P. Greene, Ph.D.
Senior Fellow, Manhattan Institute for Policy Research
Greg Forster, Ph.D.
Senior Research Associate, Manhattan Institute for Policy Research
For the last several decades middle-class families have been fleeing from the cities to the suburbs, in part because many parents see the suburbs, and suburban public schools in particular, as refuges from the disorder and social collapse they see as endemic to America's urban school districts. Parents believe that suburban public schools provide children with safer, more orderly, and more wholesome environments than their urban counterparts.
This report finds that those perceptions are unfounded. Using hard data on high school students from the National Longitudinal Study of Adolescent Health, one of the most comprehensive and rigorous studies of the behavior of American high school students, it finds that suburban public high school students have sex, drink, smoke, use illegal drugs, and engage in delinquent behavior as often as urban public high school students. Students also engage in these behaviors more often than most people realize.
This report finds that:
- Urban and suburban high schools are virtually identical in terms of widespread sexual activity. Two thirds of all suburban and urban 12th graders have had sex; 43% of suburban 12th graders and 39% of urban 12th graders have had sex with a person with whom they did not have a romantic relationship.
- Pregnancy rates are high in both suburban and urban schools, although they are higher in urban schools; 14% of suburban 12th grade girls and 20% of urban 12th grade girls have been pregnant.
- Over 60% of suburban 12th graders have tried cigarette smoking, compared to 54% of urban 12th graders; 37% of suburban 12th graders have smoked at least once a day for at least 30 days, compared to 30% of urban 12th graders.
- Alcohol use followed a similar pattern; 74% of suburban 12th graders and 71% of urban 12th graders have tried alcohol more than two or three times; 63% of suburban 12th graders and 57% of urban 12th graders drink without family members present; 22% of suburban 12th graders and 16% of urban 12th graders have driven while drunk.
- About four out of ten 12th graders in both urban and suburban schools have used illegal drugs; 20% of suburban 12th graders and 13% of urban 12th graders have driven while high on drugs.
- Urban and suburban students are about equally likely to engage in other delinquent behaviors such as fighting and stealing.
The data show that fleeing from the city to the suburbs doesn't produce much difference in the levels of sex, substance use, and delinquency one finds at the local public high school. The comforting outward signs of order and decency in suburban public schools don't seem to be associated with substantial differences in student behavior.
About the Authors
Jay P. Greene is a Senior Fellow at the Manhattan Institute's Education Research Office, where he conducts research and writes about education policy. He has conducted evaluations of school choice and accountability programs in Florida, Charlotte, Milwaukee, Cleveland, and San Antonio. He has also recently published research on high school graduation rates, charter schools, and special education.
His research was cited four times in the Supreme Court's opinions in the landmark Zelman v. Simmons-Harris case on school vouchers. His articles have appeared in policy journals, such as The Public Interest, City Journal, and Education Next, in academic journals, such as The Georgetown Public Policy Review, Education and Urban Society, and The British Journal of Political Science, as well as in major newspapers, such as the Wall Street Journal and the Washington Post.
Greene has been a professor of government at the University of Texas at Austin and the University of Houston. He received his B.A. in history from Tufts University in 1988 and his Ph.D. from the Government Department at Harvard University in 1995. He lives with his wife and three children in Weston, Florida.
Greg Forster is a Senior Research Associate at the Manhattan Institute's Education Research Office. He is the co-author of studies evaluating vouchers, charter schools, high-stakes testing, special education funding, and other education issues. He has published op-ed articles in the Los Angeles Times, the Philadelphia Inquirer, the New York Post, and other newspapers. He received a Ph.D. with Distinction in Political Science from Yale University in May 2002, and his B.A. from the University of Virginia, where he double-majored in Political and Social Thought and Rhetoric and Communications Studies, in 1995.
We would like to thank the National Longitudinal Survey of Adolescent Health at the University of North Carolina for making their data available for analysis. We would also like to thank Robert Enlow, Marcus Winters, and the Rose and Milton Friedman Foundation for their advice and support of this project.
About Education Working Papers
A working paper is a common way for academic researchers to make the results of their studies available to others as early as possible. This allows other academics and the public to benefit from having the research available without unnecessary delay. Working papers are often submitted to peer-reviewed academic journals for later publication.
For decades, middle-class Americans have been fleeing to the suburbs in order to escape what they perceive as a rising tide of disorder and decadence in American cities. In particular, parents have moved to the suburbs not just in hopes of providing their children with schools that are higher in academic quality, but also to provide them with schools that are safer, more orderly, and more wholesome than those available to city children. In movies, on television, and in the imagination of the average American, the urban public school conjures images of students doing drugs, having sex, stealing, and getting into fights, while the suburban public school appears as a safe haven from urban chaos.
But hard data on high school students taken from one of the biggest and most rigorous studies of American students paint a very different picture. However different urban and suburban public schools may be in terms of academics or facilities, there isn’t much difference between them when it comes to sex, drugs, violence, and delinquency. This isn’t because these problems are less serious at urban schools than people think; it’s because they’re just as serious in suburban schools as they are in urban schools.
The National Longitudinal Study of Adolescent Health finds the same results in suburban high schools as it finds in urban high schools. Two thirds of all twelfth graders have had sex, and four out of ten have had a non-romantic sexual relationship. Eight out of ten twelfth graders have touched another person’s genitals sexually. Tobacco and alcohol use are somewhat more frequent in suburban schools than in urban schools: six out of ten suburban twelfth graders have smoked, and well over a third have smoked regularly (at least once a day for at least 30 days), while more than six out of ten drink alcohol without adult family members present. In both urban and suburban schools, four out of ten twelfth graders have used illegal drugs. Among all students, about one in three have been in a serious physical fight in the past year, and about one in five have been in a fight between groups of students.
The parents who send their teenagers off to those freshly-painted, wholesome-looking suburban public schools every morning would probably be shocked if they realized those schools are virtually indistinguishable from urban schools on most measurements of sex, drugs, violence, and delinquency. There may have been a time when suburban schools really were a safe haven from the rise of these so-called “urban” problems. But if there ever was such a time, it’s gone.
The association between poor urban neighborhoods and delinquent teenage behavior is so widely accepted that it’s more often simply taken for granted rather than formally studied. The Manhattan Institute’s Kay Hymowitz has pointed out that “when Americans think about public education, they tend to see a stark divide” between the “blackboard jungles” of urban schools, “where drugs abound” and “gangs rule the hallways,” and the “shining, achievement-oriented public schools of the suburbs.” She contrasts this widespread perception with relatively recent “concerns about the often poisonous social and moral environment of the high schools in more prosperous communities” (see Hymowitz 2002; see also Hymowitz 1998; Hymowitz 2000a; and Hymowitz 2000b).
Managing to Make It: Urban Families and Adolescent Success, a recent study of how teenagers and their families respond to hostile urban environments, notes that the public “has been inundated with statistics about children in poverty, disadvantaged minorities, and the ‘urban underclass.’ “ Such statistics are typically accompanied by compelling anecdotes designed to convey “the corrosive features of growing up in persistent poverty. These images have become part of our conventional wisdom about life in the inner city. In both professional and popular writings, we are constantly being reminded that a crisis exists in our families, our schools, our legal system, and our social services” in the inner city (see Furstenberg, Cook, Eccles, Elder, and Sameroff 1999).
Managing to Make It goes on to argue that this widespread presumption of a link between urban environments and teenage delinquency has tended to blind both scholars and the public to the presence of many teens in urban neighborhoods who succeed despite their difficult neighborhoods. But the same widespread presumption also tends to distract attention from the occurrence of sex, drug use, and delinquent behavior among suburban teenagers. When the media present statistics describing the levels of such behavior among teenagers, most Americans simply assume that the teenagers engaging in this behavior are mostly, if not overwhelmingly, urban teenagers.
One of the arguments of Managing to Make It is that urban neighborhoods have less of the “social capital” that helps suburban families raise better-behaved teenagers. “Parents with the means to select their communities prefer physical settings that are safe, contain schools and services for youth, and include neighbors who support one another,” write the authors. “Parents are looking for contexts in which their children are co-socialized according to their values and expectations, contexts in which other members of the community will look out for their children and let them know if they are misbehaving.” To document this deficiency in social capital, the authors cite the work of William Julius Wilson and scholars following him (see Wilson 1987; Gans 1990; Wilson 1991; Huston 1991; Massey and Denton 1993; Duncan and Brooks-Gunn 1994; Huston, McLoyd, and Garcial Coll 1994; Wilson 1996; and Gephart 1997).
Another recent study, Consequences of Growing Up Poor, sets out to test the popular theory that poverty causes undesirable childhood outcomes. While the study is mostly concerned with academic performance and later life outcomes, it does include some analyses of behavior problems. Contributors Linda Pagani, Bernard Boulerice, and Richard E. Tremblay cite a body of research in which the general consensus is that “family socioeconomic status was inversely associated with behavior problems and school failure in children” (see Robins 1966; Magnusson 1988; Rodgers 1990; Pulkkinen 1990; Caspi, Elder, and Herbener 1990; Power, Manor, and Fox 1991; Pulkkinen and Tremblay 1992; and Farrington 1994). However, the editors of the study conclude that while income has a significant effect on children’s academic ability and economic achievement, it has little effect on their health or behavior (see Duncan and Brooks-Gunn, 1997).
All data were taken from the National Longitudinal Study of Adolescent Health. The study—which is sponsored by the National Institute of Child Health and Human Development and other federal agencies—is one of the nation’s largest and most rigorous studies of adolescent behavior.
Student data were taken from the Wave 1, Wave 2, and Wave 3 In-Home sections of the study. In 1995, during Wave 1 of the study, over 20,000 students in grades seven through twelve were selected to complete a highly detailed survey, administered in their homes, on health-related topics. For sections of the survey covering sensitive topics such as sex, drugs, and delinquency, students listened to recorded questions through headphones, entering their answers directly into a laptop computer. Students were assured of complete anonymity. Further in-home interviews were conducted with the same set of students during Wave 2 of the study in 1996 and Wave 3 in 2001-02.
School data were taken from the study’s Wave 1 Administrator survey. Each school whose students were selected to participate in the study was given a survey to be completed by a school administrator. These surveys collected a variety of basic information about the school, including whether the school is urban, suburban, or rural.
Finally, population weight data were provided as part of the study. By applying these weights to the student data, we were able to analyze the data in a way that is more accurately representative of the U.S. population as a whole.
Since every school in the survey was assigned a unique identification number, we were able to match school information from the administrator survey with student information from the in-home surveys. This allowed us to identify students attending urban public schools and suburban public schools. In Waves 1 and 2 we disregarded data from students currently in seventh and eighth grade, leaving us with a data sets containing over 11,000 urban and suburban public high school students in Wave 1 and almost 8,000 in Wave 2. In Wave 3 we followed a slightly different method, for reasons discussed below.
We then chose a set of basic indicators, drawn from the survey questions, to measure the behavior of urban and suburban students regarding sex, drugs, and delinquency. The sexual and substance use indicators drawn from Wave 1 were: whether the student has had sex, whether the student has had a non-romantic sexual relationship, whether the student has ever been diagnosed with a venereal disease, whether the student has ever been pregnant, whether the student has ever smoked, whether the student has ever smoked regularly (at least once a day for at least a month), whether the student has drunk more than two or three times in his or her life, whether the student ever drinks without adult family members present, whether the student has ever used illegal drugs, whether the student was ever drunk or high while at school, and whether the student has ever driven while drunk or high. The survey also provided a large number of delinquency indicators; we include all indicators from Wave 1’s sections on “delinquency” and “fighting and violence” that show whether the student had committed a delinquent act.
We drew as many indicators as possible from Wave 1 for the sake of consistency. However, there were two topics that were not addressed by any questions in Wave 1; in those cases we drew our indicators from Wave 2. The indicators drawn from Wave 2 were whether the student had ever touched another person’s genitals or had another person touch his or her genitals in a sexual way, and whether the student had ever had anal sex.
Also, Wave 1 did not provide usable data on abortion due to an error in the implementation of the survey. This information was collected again in Wave 3. Using data on respondent sex from Wave 1 and school data from the administrator survey, we identified female respondents who had attended urban or suburban public schools. This left us with 6,000 female respondents. Then, since the Wave 3 survey was conducted long after all the survey respondents had left high school and it collected outcome information for all pregnancies each respondent had ever had, we had to separate abortions that occurred during high school from abortions that occurred after the respondents had left high school. Unfortunately, the survey did not directly record whether each abortion occurred while the respondent was in high school, but it did record the year when each pregnancy ended. Using data on respondents’ birth years from Wave 1, we approximated abortions that occurred during high school by counting all abortions each respondent had within 18 years of her birth year. We then calculated the percentage of urban and suburban girls who had ever had an abortion.
For each of these indicators, we compared the responses of students in urban and suburban public schools. We disregarded all responses that were not specific answers, such as when students refused to respond, or responded that they didn’t know the answer. In addition to comparing urban and suburban students overall, we broke down student responses by grade level for 9th, 10th, 11th, and 12th grade students.
Some questions were only asked of students who had given particular answers to previous questions. Students who had never had sex were not asked whether they had ever been pregnant or been diagnosed with a venereal disease. Students who had never smoked were not asked whether they had ever smoked regularly. Students who had not drunk more than two or three times in their lives were not asked whether they had ever drunk without an adult family member present, been drunk at school, or driven while drunk. Students who had never used illegal drugs were not asked whether they had ever been high at school or driven while high. In Wave 2, students who had never engaged in genital touching with others were not asked whether they had ever had anal sex. For each of these questions we count students who were “skipped” as not having engaged in the activity in question. For example, we counted students who had never smoked as never having smoked regularly. Also, in Wave 2, students who reported that they had ever had sex were also counted as having engaged in genital touching with others.
The results for the sexual indicators are summarized in Tables 1-7. For the sexual intercourse indicator, urban and suburban schools are virtually identical; half of all students, and two thirds of all twelfth graders, have had sex. The broader indicator of sexual genital touching with others finds that sexual activity is more widespread than what is found by the intercourse indicator alone, and this indicator finds somewhat more activity in suburban schools than in urban schools. Seven out of ten suburban students have engaged in sexual genital touching with others, compared to two thirds of urban students, and eight out of ten suburban twelfth graders have done so, compared to three quarters of urban students. Suburban students are also somewhat more likely to have non-romantic sexual relationships, with a third of all suburban students having had such relationships, compared to about three out of ten urban students, and just over four out of ten suburban twelfth graders, compared to just under four out of ten urban twelfth graders. Urban and suburban students are very similar in terms of anal sex (about one of every twelve students) and venereal disease (about one in 25). Overall, about one in ten female students has been pregnant in both types of schools, but urban twelfth graders are somewhat more likely to have been pregnant than suburban twelfth graders. One in five urban female twelfth graders has been pregnant, as opposed to just under one in seven suburban female twelfth graders. Urban and suburban girls are about equally likely to have had an abortion.
The results for the substance use indicators are summarized in Tables 8-16. On every indicator, suburban students were either more likely than urban students to engage in substance-related behaviors, or about as likely to do so. For smoking, suburban students were more likely than urban students to engage in both of the measured behaviors: six out of ten suburban students have smoked, compared to just over half of urban students, and three out of ten suburban students have smoked regularly (at least once a day for at least 30 days), compared to about a quarter of urban students. Twelfth graders were not much more likely to have smoked than students overall, but they were significantly more likely to have smoked regularly: almost four out of ten suburban twelfth graders, and three out of ten urban twelfth graders, have smoked regularly.
Suburban students were slightly more active drinkers than urban students on the broad alcohol indicators: about two thirds of suburban students have drunk, compared to just over six out of ten urban students, and just over half of suburban students drink without adult family members present, compared to just under half of urban students. Almost three quarters of suburban twelfth graders, and about seven out of ten urban twelfth graders, have drunk; over six out of ten suburban twelfth graders drink without adult family members present, compared to just under six out of ten urban twelfth graders. Suburban students were about as likely as urban students to have been drunk at school (about one in twelve students overall, and over one in ten twelfth graders) or have driven drunk (about one in ten students overall, and about one in five twelfth graders).
For illegal drugs, suburban and urban students were virtually identical on the broad indicator—in both cases, well over a third of all students and over four of ten twelfth graders have used illegal drugs. Almost one out of seven students in both urban and suburban schools, and about one out of six twelfth graders, have been high on drugs at school. About one in ten suburban students, and about one in fourteen urban students, have driven while high, and about one in five suburban twelfth graders have done so, compared to about one in eight urban twelfth graders.
The results for delinquency are summarized in Tables 17-34. Here suburban students are generally the same as urban students. In a few cases they are slightly less likely to engage in the measured behaviors, but differences between urban and suburban students are small in every case. Over half of urban and suburban students have lied to their parents in the past year about where they’ve been or who they were with. Suburban students are slightly less likely than urban students to have been in a physical fight or a serious physical fight in the past year; about three out of ten suburban students have been in a fight or serious fight, compared to about a third of urban students. However, urban and suburban students are about equally likely to have been in a fight between groups of students or to have seriously injured someone; about one in five students has participated in a group fight or injured someone seriously enough to require medical attention in the past year. About a quarter of urban and suburban students have shoplifted in the past year, and a little less than two out of ten students have deliberately damaged someone else’s property in the past year. About one in eight students has stolen a car in the past year. About one in ten students has run away from home in the past year, and about one in eleven has sold drugs or painted graffiti. About one in fifteen students has carried a weapon to school in the past year. In almost all cases, older urban and suburban students are less likely to have engaged in these behaviors.
Parental concern about the rising influence of sex, drugs, and delinquency in urban schools has long been recognized as a significant factor in the last few decades’ population flight from the cities to the suburbs. Parents are fleeing urban schools not just because of low academic performance but also because they believe suburban schools are safer and more wholesome. But the results from the National Longitudinal Study of Adolescent Health suggest that fleeing from city to suburb doesn’t produce much difference in the level of these problems one finds at the local school. The desks may be newer, the paint may be fresher, and the faces may be whiter, but the students are just as likely to have sex, use controlled substances, and break the law. The comforting outward signs of order and decency—shiny new schools armed with expensive textbooks and staffed by teachers who have mastered the latest educational fads—don’t seem to be associated with substantial differences in student behavior.