Defining and Measuring School Crime
In the past decade, parents, policy makers, and school officials have paid increasing attention to school crime. Because of extreme incidents of school violence, the problem of school crime has assumed national importance. Barely a day goes by when some school crime incident, great or small, is not reported in the mass media or the subject of a government report or investigation. With increased focus on school crime comes the need for accurate statistics.
There are many different ways to define school crime. Schools appear to be safe or dangerous, depending on what one counts as school crime. Definitions of school crime range from considering any threat or theft as a crime to considering only violent attacks that are reported to police as crimes. They differ, too, depending on whether or not crimes committed against children on their way to school or on school playgrounds as well as in school buildings are counted. They also differ in whether crimes are counted only during school hours or also before and after school. Rates also differ because some surveys count crimes only by or against school personnel and students, whereas others count any victim on school property. In addition, the amount of school crime reported differs in relation to who gives the information and whether it is acquired by personal interview, telephone interview, or questionnaire or is from official records.
Without a standard definition of school crime, tracking incidents of crime is problematic. Behaviors and offenses included as school crime, assertions made about its incidence and prevalence, and estimates of students’ risk of victimization are unreliable (Hanke, 1996). Lack of defi-
nitional precision may also contribute to the mistaken assumption that extreme incidents of school violence, like the tragic shootings in Littleton, Colorado, are representative of school crime in general. While there are advantages to having a single, widely accepted definition of school crime, circumstances may arise in which more fine-tuned measures of school crime are preferred (e.g., when data are collected for a single school, school district, or jurisdiction).
Attempts to standardize definitions have been made. While there are no guidelines at the national level for data collection on school crime, widely accepted definitions of school crime do currently exist. The Crime, Violence and Discipline Task Force created by the National Forum on Education Statistics in 1995 has developed definitions and protocol for collecting data on school crime and violence (Minogue et al., 1999). It recommended that school crime be inclusive of: incidents that occur on school grounds, on school transportation, or at off-campus school-sponsored events; incidents involving alcohol, drugs, or weapons; incidents involving a gang; hate-crime motivated incidents; and all incidents reported to law enforcement agencies.
There is a scarcity of reliable data on school crime. Moreover, there are limitations to the data that are collected. Several workshop participants noted the inconsistencies across data collection efforts in who is sampled, how data are collected (e.g., personal interview, telephone interview, questionnaire), which incidents are included as school crime, and how estimates are derived. Several workshop participants also noted that little attention has been given to ascertaining the accuracy of data used for reporting school crime.
In the last several years one-time and on-going data collection efforts have been initiated by government agencies to provide better data on school-related crime (see Chandler et al., 1998; Kaufman et al., 1998; National School Safety Center, 1998; National Center for Education Statistics, 1998).
In 1974 Congress mandated the first national study of school safety. Researchers from the Research Triangle Institute asked public school students and teachers in grades 7 through 12 to report school-related victimization experiences. School principals, too, supplied information on such crimes as vandalism in their schools. According to the information collected, in a typical month during the prior year an estimated 128,000 junior and senior high school teachers had something worth more than $1 stolen from them. In a typical month an estimated 5,200 teachers were physically assaulted (National Institute of Education, 1986). Students reported considerably more crime on the questionnaires than they did in personal interviews (Gottfredson and Gottfredson, 1985). In addition to actual victimization incidents, students and teachers reported being threatened at a rate several times greater than actual thefts or attacks.
In 1989 the National Crime Victimization Survey added a School Crime Supplement to measure victimization of youth ages 12 to 19. During the first 6 months of 1988-1989, according to this supplement, 7 percent of students surveyed reported being victims of at least one property crime and 2 percent reported their victimization in a violent crime to the police. A larger proportion, 18 percent, reported being afraid sometimes of being attacked (Bastian and Taylor, 1991). In an analysis of National Crime Victimization Survey Incident Reports and interviewers’ narrative data, Garofalo et al. (1987) found that 54 percent of victimizations reported by adolescents were school related (i.e., occurred while attending school); 41 percent of aggravated assaults, 44 percent of robberies, and 59 percent of simple assaults were school related. Robbery and aggravated assault were more likely to occur while students were traveling to or from school than on school grounds or on a school bus; simple assault was most likely to occur in a school building (Garofalo et al., 1987).
The 1993 National Crime Victimization Survey indicated that about half the students in grades 6 through 12 witnessed a victimization at school and about an eighth had personally been victimized (Nolan et al., 1996). Surveying 11,000 students in grades 8 through 10, the American School Health Association (1989) found that during the prior year 40 percent had been in a physical fight at school or on the school bus; 34 percent reported having been threatened; and 22 percent reported carrying a knife, gun, or other weapon. A survey by the National School Board Association (1993) found that 78 percent of the responding school districts reported incidents of assault; in more than 80 percent of the districts, school violence reportedly had increased. A major limitation of the National Crime Victimization Survey School Crime Supplement is that it limits participation to individuals 12 and older.
While data on school crime suffer from limitations—representativeness in terms of student age groups, differences in how school crime is operationalized across studies, missing/unreported incidents, types of incidents reported—existing data converge on the finding that most school crime (like crime generally) involves minor personal and property offenses. According to data from the National Crime Victimization Survey School Crime Supplement, most schools do not experience incidents of serious violence (i.e., murder, rape, robbery, aggravated assault) (Chandler et al., 1998; Kaufman et al., 1998). Of all school crime reported, the percentage of serious crime and violence is quite small. In the 1996-1997 school year, 10 percent of public schools reported a serious incident of violence to the police (Kaufman et al., 1998).
Other sources of data are available on school crime and violence, including surveys of self-reported victimization and offending. For example, the Monitoring the Future Study (University of Michigan, Insti-
tute for Social Research) is an ongoing survey (started in 1975) that collects information from high school seniors on their behavior, attitudes, and victimization experiences. The National School-Based Youth Risk Behavior Surveillance System (Centers for Disease Control and Prevention) is part of a larger monitoring system that focuses on behaviors, beliefs, and experiences that influence young people ’s health, including weapon carrying, involvement in physical fights, and drug and alcohol use. The 1997 National Longitudinal Survey of Youth (Bureau of Labor Statistics) follows a cohort of youth ages 12 to 16 to collect data annually on their educational and work experiences and attitudes as they transition into adulthood. These data include information on drug and alcohol use and self-reported offending. The National Longitudinal Study of Adolescent Health (Carolina Population Center, University of North Carolina at Chapel Hill) is a school-based survey of youth in grades 7 through 12 that collects data on the health-related behavior of adolescents in school, family, peer, and neighborhood contexts.
Surveys on the conditions of school environments also exist, including the National Household Education Survey (National Center for Education Statistics [NCES]), which looks at school safety and discipline; the Schools and Staffing Survey (NCES), which collects data on teacher victimization; and the Principal/School Disciplinarian Survey on School Violence (NCES), which is part of the NCES Fast Response Survey System designed to gather information on educational issues of interest.
Some states have instituted systems for collecting data across school districts. For example, Kansas and Louisiana have conducted censuses of all 6th, 8th, 10th, and 12th grade students in participating secondary schools. These surveys provide data on the prevalence of self-reported crime, substance abuse, and risk and protective factors on a school-by-school basis.
Although standardized definitions of school crime would permit comparisons across time and place and the establishment of an official repository for data on incidents, such standardization is unlikely to overcome pressures to resist recognition of crime in their own schools by some principals, teachers, and students. Neither would standardization provide an antidote for the tendency to overlook extremely common incidents of minor theft and assault, threats, and extortion that go on in the name of “normal childhood behavior.”
Kenneth Trump, of National School Safety and Security Services, in Cleveland, Ohio, noted that there are institutional factors that exert a great deal of influence on the quality of school crime data. Measurement of school crime may be complicated by the reporting practices of schools. Schools may fail to report incidents to local law enforcement because they do not distinguish crime from disruptive behavior, or schools may simply
choose to handle incidents internally instead of reporting them to the police.
Estimates of school crime may also be affected when schools differentially report incidents to local law enforcement. The perceived seriousness of an offense influences whether it is reported to the police and varies depending on the context in which the offense occurred, including details specific to the incident, and whether similar offenses have occurred in the past. For example, incidents involving students and teachers may be more routinely reported to the police than are minor, student-on-student attacks. Likewise, incidents in large urban schools may be more likely to be reported than those that occur in rural school districts. Several state legislatures are addressing this problem by implementing guidelines for mandatory school crime reporting. How effective these guidelines will be remains to be seen.
Underreporting and differential reporting illustrate the importance of closely scrutinizing study findings, especially those based solely on official police arrest statistics. The use of student self-reported victimization data may serve as a counterbalance to official estimates (Chandler et al., 1998; Kaufman et al., 1998).
Although rare, serious incidents of school crime and violence raise pressing policy issues that school administrators must address. At the workshop, Trump explained that in order to grasp the meaning that school administrators attach to school crime and violence, one must understand that schools operate in a highly politicized environment and that school crime and violence are intensely political issues. His presentation emphasized the importance of understanding how schools behave as organizations with their own sets of contingencies. Because schools are concerned about their image in the community, school administrators may be hesitant to make public a whole range of student misbehaviors.
