Purpose: Over the past decade, concern has developed among juvenile justice practitioners, advocates and researchers over the apparent increase in violence among girls (Acoca, 1998; Chesney-Lind, 2004; Gaarder & Belknap, 2002, Heimer & De Coster, 1999). The evidence cited for this concern is the increased rates of girls arrested for violent crimes in the 1980s and 1990s (Juvenile Justice Digest, 2004; Snyder, 2004). This study counters the perception of girls' increased use violence based on trends in arrest data with self-report data that show no significant increase in girls' use of violence. Given these results, this study goes on to analyze the discourse on girls' use of violence in the media and literature. Findings suggest that the apparent concern over the increase in girls' use of violence may more accurately reflect societal tensions over shifting norms of gender and race.

Methods: For this study I analyzed discourse on girls' use of violence in a) major newspapers from Detroit, Washington DC, New York, and Los Angeles, b) national news magazines, c) professional juvenile and criminal justice journals, and d) peer-reviewed academic journals. I also analyzed longitudinal self-report data on a nationally representative sample of eighth and tenth graders from the Monitoring the Future data set.

Results: Findings from the Monitoring the Future data set show no significant differences in girls' use of violence during the 1980s and 1990s. This finding is supported by analysis of other nationally representative self-report data (Chesney-Lind, 2004; Snyder, 2004). Taken together, these quantitative findings suggest that concern over the increase in girls' use of violence may be rooted in something other than an actual increase.

Results from the discourse analysis show frequent use of sensationalized and extreme examples of girls' violence presented in raced and gendered terms, as well as uncritical acceptance of “official” crime and arrest data as reflective of the actual incidence of crime. These findings suggest that concern over the increase in girls' violence is a result of cultural tensions over changing norms and understandings of race and gender in the United States, and limited theorizing about violence that constructs our understandings of violence as solely an enactment of masculinity.

Implications: Discourse analysis is an important method for social work research that allows us to see the construction of social problems, and provides insight into the processes through which that construction occurs. Awareness of the ways in which media and other public discourses construct our understanding of social problems is essential to designing effective social work programs and policies. Without it we run the risk of allowing social work's greatest strength – action – to become a weakness. Additionally, this study shows a dire need for those involved in social work research, policy, and practice with girls and violence to examine existing assumptions about the connection between violence and masculinity. Exploring other possible meanings and motivations for violence has the potential to vastly improve policy and practice with all adolescents at risk for use of violence.