Criminal Justice Section  


Criminal Justice Magazine
Spring 2001
Volume 16, Issue 1

Are Girls Closing the Gender Gap in Violence?

By Meda Chesney-Lind

Girls in the juvenile justice system were once dubbed the "forgotten few." That concept has rapidly faded as the increase in the number of girls arrested has dramatically outstripped that of boys for most of the last decade. Girls now account for one out of four arrests, and statistics show the greatest increase in arrests is for violent offenses. This shift highlights both the need to better understand the dynamics of female delinquency and the need to tailor the criminal justice system's response.

Changes in arrests

Between 1989 and 1998, arrests of girls increased 50.3 percent, compared to only 16.5 percent for boys, according to the FBI's 1999 report, Crime in the United States 1998. During that same period, arrests of girls for serious violent offenses increased by 64.3 percent and arrests of girls for "other assaults" increased an astonishing 125.4 percent. In 1999, the Office of Juvenile Justice and Delinquency Prevention reported that the female violent crime rate for 1997 was 103 percent above the 1981 rate, compared to a 27 percent increase for males, prompting the statement that "increasing juvenile female arrests and the involvement of girls in at-risk and delinquent behavior has been a pervasive trend across the United States." Concomitant with the rising number of arrests are increases in girls' referral to juvenile courts. Between 1988 and 1997, the number of delinquency cases involving girls increased by 83 percent compared to a 39 percent increase for males. (Anne L. Stahl, 2000 Delinquency Cases in Juvenile Courts 1997, Off. Juv. Just. & Delinq. Prevention, p. 1).)

Aggression and violence

This apparent spike in what might be called girls' "nontraditional" delinquency has to be understood in context: Girls' capacity for aggression and violence has historically been ignored, trivialized, or denied. Perhaps because of this, self-report data, particularly from the 1970s and 1980s, have always shown higher involvement of girls in aggressive behavior than is reflected in official statistics. For example, in a 1976 survey, adolescents reporting on their own behavior showed a male to female ratio of 3.5:1 for serious assault, and a ratio of 3.4:1 for minor assault. Meanwhile, the FBI's 1980 arrest statistics showed a much greater male to female ratio of participation in aggravated assault-5.6:1. Another study reported the ratio of male to female for simple assault at 3.8:1. (R. J. Canter, Sex Differences in Self-Report Delinquency, 20 Criminology 373-93 (1982).) FBI 1999 arrest statistics show a 3.54:1 ratio for "aggravated assault" and a 2.25:1 ratio for "other assaults." Taken together, these numbers suggest that what we're seeing is not an increase in violence among girls so much as a closing of the gap between what girls have always done (and self-reported) and arrest statistics.

Detailed comparisons drawn from supplemental homicide reports from unpublished FBI data also hint at the way that gender has colored and differentiated girls' and boys' violence. (Meda Chesney-Lind & S. Okamoto, Gender Matters: Patterns in Girl's Delinquency and Gender Responsive Programming, J. Forensic Psychol. Prac. (forthcoming 2001). In a study of FBI data on the characteristics of girls' and boys' homicides between 1984 and 1993, girls accounted for "proportionately fewer homicides in 1993 (6 percent) than in 1984 (14 percent)." In comparison to homicides committed by boys, girls who committed homicides were more likely to use a knife than a gun and to murder someone as a result of conflict rather than in committing a crime. Girls were also more likely than boys to murder family members (32 percent) and very young victims (24 percent of their victims were under the age of three compared to 1 percent of the boys' victims). When involved in a peer homicide, girls were more likely than boys to have killed as a result of an interpersonal conflict and were more likely to kill alone, while boys were more likely to kill with an accomplice. The researchers concluded that the stereotype of girls becoming gun-toting robbers was not supported. "The dramatic increase in gun-related homicides . . . applies to boys but not girls." (A.B. Loper and D.G. Cornell, Homicide by Girls, 5 J. of Child & Family Studies 321-33 (1996).)

Other studies using self-report methods also fail to show the dramatic changes found in official statistics. A matched sample of high-risk youth (ages 13-17) surveyed in the 1977 National Youth Study and the 1989 Denver Youth Survey revealed significant decreases in hard drug use, and felony and minor assaults committed by girls, and no change in a wide range of other delinquent behaviors, including felony theft and minor theft. Further, a summary of two recent studies on self-reported aggression (see Table 1) reflects that only about a third of girls reported having been in a physical fight in the last year while more than half the boys have. Girls are far more likely than boys to fight with a parent or sibling (34 percent compared to 9 percent), whereas boys are more likely to fight with friends or strangers. Finally, boys are two to three times more likely to report carrying a weapon. ( Prevention and Parity: Girls in Juvenile Justice, Girls Incorporated National Resource Center (1996).)

Table 1.

Actual and Potential Involvement in Physical Violence






Involved In:

Physical fight in past year



Adams et al.



Kann et al.

Four or more physical fights in the past year



Adams et al.

Fought With:




Adams et al.




Adams et al.

Date/romantic partner



Adams et al.




Adams et al.




Adams et al.

Several of the above



Adams et al.

Carried a Weapon:

In the past month



Adams et al.



Kann et al.

Adams et al. (1995: ages 14-17, 1992 data) and Kann et al. (1995: grades 9-12, 1993 data) in Girls, Inc. 1996.

The psychology literature on aggression, which considers forms of aggression other than physical aggression or violence, is also relevant. Taken together, this literature generally shows that boys and men are more likely to be physically aggressive than females, but the differences begin to even out when verbal aggression is considered (yelling, insulting, teasing). Further, adolescent girls may be more likely than boys to use "indirect aggression" such as gossip, telling bad or false stories, or revealing secrets. When this broad definition of "aggression" is utilized, only about 5 percent of the variance in aggression is explained by gender. (K. Bjorkqvist & P. Niemela, New Trends in the Study of Female Aggression, in Of Mice and Women: Aspects of Female Aggression 1-16 (1992).)

Those who study aggression in children and young adults also note that girls' aggression is usually within the home or "intrafemale" and, thus, likely to be less often reported to authorities. The fact that these forms of aggression have been largely ignored by scholars as well as the general public also means that there is substantial room for girls' aggression to be "discovered" at a time where concern about youth violence is heightened.

Finally, girls' behavior, including violence, needs to be put in its patriarchal context. In an analysis of self-reported violence in girls in Canada, Sibylle Artz did precisely that, and the results were striking. First, she noted that violent girls reported significantly greater rates of victimization and abuse than their nonviolent counterparts, and that girls who were violent reported great fear of sexual assault, especially from their boyfriends. Specifically, 20 percent of violent girls stated they were physically abused at home compared to 10 percent of violent males and 6.3 percent of nonviolent girls. Patterns for sexual abuse were even starker; roughly one out of four violent girls had been sexually abused compared to one in 10 of nonviolent girls. Follow-up interviews with a small group of violent girls found that they had learned at home that "might makes right" and engaged in "horizontal violence" directed at other powerless girls (often with boys as the audience). Certainly, these findings provide little ammunition for those who would contend that the "new" violent girl is a product of any form of "emancipation." (S. Artz, Sex, Power and the Violent School Girl (Trifolium Books 1998.)

Indeed, what needs to be understood about girls' delinquency, particularly from a programmatic and policy standpoint, is the clear link between victimization, trauma, and girls' delinquency. The other major theme is that this trauma most often does not produce violent offenses, rather what have long been regarded as "trivial" or unimportant offenses such as running away from home.

Relabeling status offenses

What about dramatic increases in arrests of girls for "other assaults"? It cannot be ruled out that relabeling as violent what were once called status offenses-"running away" or "requires supervision"-has had an impact on reported crime rates as has the change in police practices towards domestic violence. A review of more than 2,000 cases of girls referred to Maryland's juvenile justice system for "person-to-person" offenses revealed that virtually all (97.9 percent) involved assault. A closer examination of the records revealed that about half were family centered and involved such activities as a girl hitting her mother and her mother subsequently pressing charges.

More recently, Leslie Acoca's study of nearly 1,000 girls' files from four California counties found that while a high percentage of these girls were charged with "person offenses," a majority of these involved assault. Further, "a close reading of the case files of girls charged with assault revealed that most of these charges were the result of nonserious, mutual combat situations with parents." Acoca details cases that she regards as typical, including: "father lunged at her while she was calling the police about a domestic dispute. She (girl) hit him." Finally, she reports that some cases were quite trivial in nature, including a girl arrested "for throwing cookies at her mother." (L. Acoca, Investing in Girls: A 21st Century Challenge, Juv. Just. 6(1) 7-8 (1999).)

In essence, when exploring the dramatic increases in the arrests of girls for "other assault," it is likely that changes in enforcement practices have dramatically narrowed the gender gap, including an increase in arrests for domestic violence. A recent California study found that the female share of these arrests increased from 6 percent in 1988 to 16.5 percent in 1998. African-American females had arrest rates roughly three times that of white females. (State of California Criminal Justice Statistics Center Bureau of Criminal Information and Analysis, Report on Arrests for Domestic Violence in California, 1998 (1999).)

Relabeling girls' arguments with parents from status offenses to assault is a form of "bootstrapping." It facilitates the incarceration of girls, especially African-Americans, in detention facilities and training schools-something that would not be possible if the girls were arrested for a noncriminal status offense.

Arrests of juveniles for minor or "other" assaults can range from schoolyard tussles to relatively serious assaults. Researchers have noted that an increasing number of arrests of girls for "other assaults" are relatively nonserious in nature and tend to consist of being bystanders or companions to males involved in skirmishes and fights. (R.H. Steffenmeier & D.J. Steffenmeier, Trends in Female Delinquency: An Examination of Arrest, Juvenile Court, Self-report, and Field Data, Criminology 18, 62-85 (1980).) Often simple assaults without injury are attempted or threatened or not completed. At a time when official concern about youth violence is almost unparalleled and school principals are increasingly likely to call police to their campuses, it should come as no surprise that youthful arrests in this area are up.

This observation is supported by recent research on the dynamics of juvenile robbery (another violent offense where girls' arrests showed sharp increases) in the City of Honolulu. In the last decade, Hawaii, like the rest of the nation, saw an increase in juvenile arrests for violent crimes followed by a recent decline. Murder, rape, robbery, and aggravated assault increased 60 percent from 1987 to 1996, but declined 8.6 percent between 1996 and 1997. Most of the increase resulted from increases in the number of youth arrested for two offenses: aggravated assault and robbery. Between 1994 and 1996, for example, the number of juveniles arrested for robbery doubled in Honolulu. (Meda Chesney-Lind & V. Paramore, Are Girls Getting More Violent?: Exploring Juvenile Robbery Trends, J. Contemp. Crim. Just. (forthcoming).)

These increases prompted my colleagues and me to study the actual dimensions of juvenile robbery in Honolulu. We worked with police files of robbery arrests in 1991 and 1997. According to the data, in 1991 only 5 percent of those arrested for robbery were female. In 1997, however, 16.7 percent were female. In six years, the proportion of robbery arrests involving girls more than tripled. ( Id.)

Taken alone, these increases, along with a few sensational anecdotes, are precisely why the apparent surge in girls' violence has made news. But the data suggest that no major shift in the pattern of juvenile robbery occurred. Rather, it appears that less serious offenses, including a number committed by girls, are being swept up into the system, perhaps as a result of changes in school policy and parental attitudes (many of the robberies occurred as juveniles went to and from school). Consistent with this explanation are the following observable patterns in our data: during the two time periods under review, the age of offenders shifts downward, as does the value of items taken. In 1991, the median value of the items stolen was $10; by 1997, the median value had dropped to $1.25. Most significantly, the proportion of adult victims declined sharply while the number of juvenile victims increased. Finally, although more of the robberies involved weapons in 1997, those weapons were less likely to be firearms, and the incidents were less likely to result in injury to the victim. In short, the data suggest that the problem of juvenile robbery in Honolulu is largely characterized by slightly older youth bullying and hijacking younger youth for small amounts of cash and jewelry, and that these arrests accounted for virtually all of the increase observed. ( Id.)

Nonaggressive offenses and drug use

Examining the types of offenses for which girls are arrested, it is clear that most are arrested for the less serious criminal acts and status offenses (noncriminal juvenile offenses such as running away or curfew violation). In 1998, roughly half of the arrests of girls were for either larceny theft, mostly shoplifting, (21.5 percent), or status offenses (22.1 percent.)

Status offenses have always played a significant role in bringing girls into contact with the criminal justice system. They accounted for about a quarter of all arrests of girls in 1998, but only 10 percent of arrests of boys-figures that remained relatively stable during the last decade. In 1998, more than half of those arrested for running away from home were girls. Running away and prostitution remain the only two categories where more girls than boys are arrested. Arrest rates for these activities have remained stable or climbed in recent years, despite passage of the Juvenile Justice and Delinquency Prevention Act in 1974, which encouraged state and local jurisdictions to divert and deinstitutionalize young people charged with status offenses. Between 1989 and 1998, according to 1999 FBI statistics, the number of girls arrested nationally for running away remained about the same, but arrests of girls for curfew violations increased by an astonishing 238.5 percent.

Why are girls more likely to be arrested than boys for running away from home? There are no simple answers. Studies of delinquency-not simply arrests-show that girls and boys run away from home in about equal numbers. Rachelle J. Canter found in a national youth survey that there was no evidence of greater female involvement in any category of delinquent behavior. Indeed, in this sample, males were significantly more likely than females to self-report status offenses. There is some evidence to suggest that parents and police may respond differently to the same behavior. Parents may call the police more frequently when their daughters do not come home, and police may be more likely to arrest a female than a male runaway. (R.J. Canter, supra 373-93.)

Finally, research on the characteristics of girls in the California Youth Authority (CYA) system reveals that while these girls cannot be incarcerated for status offenses, nearly half (45 percent) had been charged with status offenses prior to their incarceration in the CYA for more serious offenses. Focus groups working in California found that girls in that state were chiefly involved in the juvenile justice system for offenses such as "petty theft, shoplifting, assault and battery, drug violations, gang activity and truancy, lying to a police officer, and running away." (B. Bloom, & R. Campbell, Literature and Policy Review, in Modeling Gender-Specific Services in Juvenile Justice: Policy and Program Recommendations (B. Owen & B. Bloom, eds., 1998).)

Sexual and physical abuse

Research shows that boys and girls often have different reasons for running away. Girls are much more likely to be victims of childhood sexual abuse-some experts estimate that roughly 70 percent of the victims of child sexual abuse are girls. (D. Finkelhor & L. Baron, Risk Factors for Child Sexual Abuse, 1 J. Interpersonal Violence, 43-71 (1986).) Not surprisingly, the evidence also suggests a link between this problem and female delinquency-particularly running away from home.

Studies of girls on the streets and in court populations show high rates of both sexual and physical abuse. A study of a runaway shelter in Toronto found that 73 percent of girls and 38 percent of boys had been sexually abused. The same study found that sexually abused female runaways were more likely than their nonabused counterparts to engage in criminal activities such as theft and prostitution. No such pattern was found in abused boys. (A. McCormack, M.D. Janus & A.W. Burgess, Runaway Youths and Sexual Victimization: Gender Differences in an Adolescent Runaway Population, Child Abuse & Neglect, 387-95 (1986).)

Detailed research on young people entering the juvenile justice system in Florida has compared the constellations of problems presented by girls and boys entering detention. These studies found that female adolescents were more likely to have abuse histories and contact with the juvenile justice system for status offenses, while male youth had higher rates of involvement with various delinquent offenses. Further research on a larger group admitted to an assessment center in Tampa concluded that "girls' problem behavior commonly relates to an abusive and traumatizing home life, whereas boys' law violating behavior reflects their involvement in a delinquent lifestyle." (R. Dembo, S.C. Sue, P. Borden & D. Manning, Gender Differences in Service Needs among Youths Entering a Juvenile Assessment Center: A Replication Study, Address at the Annual Meeting of the Society of Social Problems (August 1995).)

More recent research confirms these observations. In a study of the backgrounds of 96 girls in the custody of the California Youth Authority, researchers compared these results with those garnered from a comparison sample of male youth. They found that boys were more likely to be traumatized as observers of violence, but "girls were more likely to be traumatized as direct victims." As a result, it's possible that girls are more likely than boys to suffer post-traumatic stress disorder; the levels of the disorder found in the researched population were "significantly higher than among the general adolescent female population"-65 percent compared to 11 percent. About two-thirds of the girls were serving time for a violent offense (murder, assault, and robbery) and 43 percent were identified as gang members. (E. Cauffman, S.S. Feldman, J. Waterman & H. Steiner, Post-traumatic Stress Disorder among Female Juvenile Offenders, 31(11) J. Am. Acad. of Child & Adolescent Psychiatry 1209-16 (1998).)

Program as if girls matter

National data indicate that between 1988 and 1997 detentions involving girls increased by 65 percent compared to a 30 percent increase for boys. (Gillian Porter, Detention in Delinquency Cases 1998-1997, Off. Juv. Just. & Delinq. Prevention (2000).) San Francisco researchers examined the situation of girls in their juvenile justice system and concluded they frequently languished in detention centers waiting for placement, while the boys were released or put in placement. ( A.D. Shorter, L. Schaffner, S. Shick & N.S. Frappier, Out of Sight, Out of Mind: The Plight of Girls in the San Francisco Juvenile Justice System, San Francisco Center for Juv. & Crim. Just. (1996).) As a result, 60 percent of the girls were detained for more than seven days, compared to only 6 percent of the boys. These figures reflect a system that has failed to develop programs shaped by girls' unique situations or to address the special problems girls have in a sexist society.

Part of the challenge is that girls remain all but invisible in programs for youth and in the literature available to those who work with youth. For example, a 1993 study of the San Francisco Chapter of the National Organization for Women found that only 8.7 percent of the programs funded by the major city organization funding children and youth programs "specifically addressed the needs of girls." Not surprisingly, then, a 1995 study of youth participation in San Francisco after-school or summer sports programs found only 26 percent of the participants were girls. (Nina Siegal, Where the Girls Are, S.F. Bay Guardian, Oct. 4, 1995 at 18.)

In addition, people who work in the juvenile justice system typically prefer working with boys and routinely stress the "difficulty" of working with girls. According to one study, juvenile justice professionals who work with both male and female adolescents talk almost exclusively about their male clients. Likewise, in 1997, Christine Alder noted that "willful" girls produce problems for a system devised to handle boys: professionals in these systems often conceptualize girls as "hysterical," "manipulative," "verbally aggressive," and "untrusting" while boys are "honest," "open," and "less complex." Clearly, the juvenile justice system has its work cut out for it if it hopes to deal fairly with girls, to say nothing of creating programs and services tailored to girls' problems and needs.

Alder also notes that serving girls effectively will require different and innovative strategies since young men tend to be more noticeable and noticed than young women. When girls go out, they tend to move in smaller groups, they face greater proscriptions against "hanging out," and they may justifiably fear being on the streets at night. Finally, girls have many more domestic expectations than do boys, and these may keep them confined to their homes. Alder notes that this may be a particular issue for immigrant girls. (Christine Alder, Passionate and Willful Girls: Confronting Practices, Address at the Annual Meeting of Academy of Criminal Justices Sciences (1997); Delinquency Prevention with Young Women, Delinquency Prevention Conference in New South Wales (1995); Unemployed Women Have Got It Heaps Worse: Exploring the Implications of Female Youth Unemployment, Australian & New Zealand Soc. of Criminology 19, 210-24 (1986).)

Given what we know about girls' problems, including girls' aggression and violence, what should effective programs for troubled girls look like? They should address the following:

o dealing with the physical and sexual violence in their lives (from parents, boyfriends, pimps, and others);

o confronting the risk of HIV/AIDS;

o handling pregnancy and motherhood;

o coping with drug and alcohol dependency, facing family problems;

o dealing with employment training and unemployment;

o finding safe housing;

o managing stress; and

o developing a sense of efficacy and empowerment.

Many of these needs are universal and should be part of programs for all youth. However, most are particularly important for young women. Juvenile justice professionals should also scrutinize programs to ensure that they are culturally-specific as well as gender-specific. Since girls of color experience their gender differently than do their white counterparts, programs to divert and deinstitutionalize must respond to the unique developmental issues confronting minority girls. They must also build in the specific cultural resources available in ethnic communities.

Programs, particularly those that are issue-specific, also need to provide transition and after-care services that support the progress young women make. Girls' programs also need to create separate time and space for girls, apart from boys, so that concerns about sexism will not be overshadowed by boys' more disruptive behavior. Programs, particularly prevention programs for girls, need to begin at earlier ages. Many at-risk girls may engage in delinquent behavior simply because there is little else to occupy their free time. Structured recreation that gets past the girls-watching-boys-play-sports approach should be vigorously explored. Girls in the juvenile justice system often say that they probably would have avoided arrest had they had opportunities to engage in meaningful, interesting activities.

Finally, programs should work to empower girls and advocate for change that will benefit girls. This entails not only building on girls' innate strengths, skills, and creativity to develop their voices and their abilities to assert themselves, but also identifying and challenging barriers that girls, particularly marginalized girls, face in our society.

Meda Chesney-Lind is a professor in the Women's Studies Program at the University of Hawaii at Manoa.

Return to Table of Contents - Spring 2001

Return to Criminal Justice magazine home page