Measuring Violent Crime

Abstract

There are a variety of considerations related to the measurement of violence. Concerns regarding collection of primary data include the fact that violence is rare, and typical samples would yield very low variability. Self-report data may provide rich detail, but asking questions about violent behavior may result in social desirability biases. Techniques such as life history calendars have enhanced data collection in some studies. Children may be too young to respond to surveys; measuring their physical aggression can be done with ratings. Aggregate measures of violence are typically derived from government reports. The primary source of official crime data in the United States has been the summary Uniform Crime Reports (UCR), collected from police departments by the Federal Bureau of Investigation for many decades, but it has been replaced by the National Incident-Based Reporting System (NIBRS), a change made nationwide as of 2021. Considerable gaps in the data are expected to be narrowed as local jurisdictions and states adapt to the new system. To complement UCR and NIBRS official estimates of reported crime, the Bureau of Justice Statistics also fields the annual National Crime Victimization Survey (NCVS) to understand victimization in more depth. Researchers wishing to use these sources of data should be aware of the many published discussions of limitations and caveats for each data set. Those interested in studying specific forms of violent crime, including homicide, rape, robbery, and assault, should be aware of idiosyncrasies in the measurement of those crime types. For homicide, there are a variety of additional sources of official data, such as the Supplemental Homicide Reports, the National Vital Statistics System, and the National Death Index. Changes in the collection of assault data, including a move away from the “aggravated assault” category commonly used in the past by UCR and NCVS, deserve particular attention. Cross-national data are limited, and most investigators rely on World Health Organization data for homicide or the International Crime Victimization Survey. These have limited reach and scope compared to Interpol reports that were available in the past. Researchers interested in individual-level data should also be aware that there are many large-scale, secondary data sets available for analysis. These are sometimes restricted and require approval of an application to access sensitive data. Data include those derived from longitudinal studies of children and teenagers such as AddHealth and the National Longitudinal Survey of Youth, the Social Development Project, the Cambridge Study in Delinquent Development, the Dunedin Multidisciplinary Health and Development Study, and the Pathways to Desistance study. There are several other categories of violence that deserve special attention. Because family violence is underreported to police, and official sources of data are thought to be inadequate, researchers have been systematically working to measure and track domestic violence for a long time. The data and reports generally come from noncriminal justice agencies. Many studies use the Conflict Tactics Scale and the Revised Conflict Tactics Scale(CTS2) to gauge intimate partner violence. This index includes subscales that account for severity of violence and context (e.g., fighting in self-defense). The CTS2 has been adapted for a variety of family and dating relationships. The Centers for Disease Control and Prevention fields the National Intimate Partner and Sexual Violence Survey, and researchers may access those data. In the early 21st century, the public and researchers have expressed growing interest in violence by and against the police, and there are now several sources of data for researchers to use. These include the Law Enforcement Officers Killed and Assaulted data collection program and nongovernment sources such as The Washington Post’s police shootings database. Similarly, for those interested in gun violence, government sources are limited, and many investigators turn to nongovernment sources such as the Gun Violence Archive, which has no government or advocacy affiliation. Finally, those interested specifically in mass violence will also find the National Mass Violence Resource Center. The magazine Mother Jones publishes data on mass shooting events, by year, dating back to 1982, and The Washington Post and USA Today also publish an archive of mass violence events.