Community, opportunity, and crime on school property: A quantitative case study

School violence is not a uniquely American concern. Instead, it is a high-profile public policy issue in a broad range of nations, including Japan, Australia, Canada, Ireland, Scotland, Wales, England, Germany, and France (Griffiths 1995; Mellor, 1995; Takashi & Wataru, 1995: MacDonald & DaCosta, 1996; Clowes, 1999; MacDonald, 1999).

Similarly, school violence is not a new issue. At least since the 1890s, it has received sporadic attention in the literature on educational policy and research (Greer, 1973; Nasaw, 1979; Reese, 1995; Nims & Wilson, 1998). More than two decades before the horror of school shootings in the United States, Willis' (1977) ethnographic classic Learning to Labour was replete with incidents of fighting, bullying, intimidation, harassment, and vandalism in a British secondary school. In addition,Maximum Security, Devine's (1996) award-winning account of pervasive, sometimes deadly violence in New York City high schools, was published two years before the killings at suburban Columbine.

Diversity with respect to the time and location of school violence is matched by variety in its forms. Homicides and suicides, massively lethal in-school shootings, have understandably generated a public policy obsession (Kiweli, 1997; West Virginia Criminal Justice Statistical Analysis Center, 2000a; Maeroff, 2000). Occurring far more frequently, however, are less physically destructive but nevertheless harmful forms of school violence and other school crimes. These include, most often, simple assault, fist-fighting, intimidation, taunting, and bullying (William Gladden Foundation, 1992; Feinstein, Roach, & Wood, 1996; Sandhu & Aspy, 2000). In addition, reports of school violence often include much more serious crimes, such as rape, as well as lesser offenses which may be criminal but need not entail physical violence, such as sexual harassment, use of street drugs, burglary, breaking and entering, vandalism, and petty theft (Texas Kids Count Project, 1999; North Carolina State Department of Public Instruction, 1999; West Virginia Statistical Analysis Center, 2000a).

Wherever it occurs and whatever its specific nature, school violence is destructive. In its extreme forms it means loss of life. In all its forms, school violence and other crimes committed in schools undercut achievement by diminishing students' sense of security in school settings where violence begets, often as a protective response, still more violence (Foley, 1990; Grady, 1996; Devine, 1996; Coleman, 1998).

An Alternative Perspective on School Violence

In the following, we offer an alternative perspective to usual accounts as to the provenance of school violence and other crimes on school property. Instead of placing the focus on individual students and their putative deficiencies, limitations, or other individual-level characteristics, we seek to understand school violence as, at least in part, a contextually determined social phenomenon.

Previous Research: Opportunity and Community

Conceptually, our school violence research follows from earlier work on teen pregnancy, early teen pregnancy, adolescent social distress, and early school achievement in poor, rural neighborhoods (Bickel, Lange, Weaver, & Williams, 1997; Bickel & McDonough, 1997; Bickel & McDonough, 1999; Bickel, Smith, & Eagle, 2001). In the earlier research we found that rates of teen pregnancy among females aged fifteen to nineteen varied inversely with levels of economic opportunity and with levels of in-school and out-of-school community (Bickel, Lange, Weaver, & Williams, 1997). As levels of economic opportunity increased, rates of teen pregnancy declined. As levels of in-school and out-of-school community increased, rates of teen pregnancy declined.

Research results were very similar with rates of early teen pregnancy among females aged ten to fourteen (Bickel & McDonough, 1999). Again, as levels of economic opportunity increased, early teen pregnancy rates declined; and as levels of in-school and out-of-school opportunity increased, early teen pregnancy rates declined.

Our findings were much the same, moreover, with regard to levels of social distress among adolescents living in poor rural areas (Bickel & McDonough, 1997). Using rates of violent death, homicide and suicide, among young people aged fifteen to nineteen as an indicator of adolescent social distress, we found that social distress levels varied inversely with the levels of economic opportunity and with levels of in-school and out-of-school community.

Finally, in combined ethnographic and quantitative research on the effects of community quality on early school achievement, we found that kindergarten students living in areas where measured levels of community quality were high had a substantial achievement advantage over their peers (Bickel, Smith, & Eagle, 2001). Neighborhood quality was most conspicuously manifest, moreover, as a group effect, occurring within elementary school classrooms.

In the present instance, we seek to build on this earlier work by trying to explain district to district variability in rates of school violence and other criminal offenses occurring in schools from one district to another. Specifically, we investigate the possibility that variability in levels of in-school and out-of-school community, and variability in levels of economic opportunity and educational opportunity (a variable not used in our earlier research) help to explain differences in rates of crime on school property. As with other kinds of seemingly irrational and self-defeating behavior, we seek to shift the focus from characteristics of individuals, and make an effort to understand crime on school property as a contextually determined phenomenon.