Epidemiology of Crime the United States

Part 1. Measuring the Extent and Distribution of Crime

Sources of Evidence and the "Iceberg" of Crime

For years, criminologists who studied the rate and distribution of crime in the United States had to rely on official data originally collected by law-enforcement agencies for their own bureaucratic purposes. For instance, ecological researchers at the University of Chicago during the 1920s collected most of their data from police arrest records and records of the juvenile court as they attempted to describe the distribution of criminals and juvenile delinquents throughout the urban area. Official arrest data had, and continue to have, some obvious limitations. As a measure of the extent of crime in the city, even the Chicago Sociologists realized that arrest data only measured the "tip of the iceberg." For one thing, most crimes are never "cleared by arrest"--that is, offenders are never caught in the majority of common crimes such as theft and assault. Furthermore, offenders who are caught and arrested may be atypical in many respects as compared to offenders who manage to elude police and escape arrest. In other words, the great bulk of offenders and offenses which remain in the "hidden part of the iceberg" may quite different from those "known to police." Official data on arrests will tend to reflect the biases or political motives of law-enforcement agencies, in that police may tend to target certain kinds of people (e.g., "racial profiling") or certain areas (such as low-income communities) in their routine arrest practices. It is always important to remember that an arrest is an act of social control rather than an act of deviant behavior. Thus, the arrest data at the tip of the iceberg of crime may tell us more about the nature of social control than about the nature of crime and criminals.

Starting in the 1930s, the Federal Bureau of Investigation began collecting a second type of data--offenses known to police--which would provide a more extensive view of the total volume of crime in the United States. The FBI distributed questionnaires to local law-enforcement agences that asked them to report the number of violent offenses and property crimes that had come to their attention during a given period of time. This source of evidence was called the Uniform Crime Reports (UCR) because the questionnaires used standardized or "uniform" definitions of criminal offenses to insure that the responses of different law-enforcement agencies would be consistent and, therefore, comparable. The annual publication of national data from the UCR , which include data on arrests as well as offenses known to police, is greeted with a great deal of press coverage and political attention. Changes in national or regional rates of UCR offenses known to the police are often treated by the news media, politicians, and other claimsmaking groups as the primary measure of whether the crime problem is improving or getting worse. And, yet, UCR offenses known to police only shed light on a bit more of the "iceberg" of crime than do arrest data. The UCR data do not, of course, reveal anything about crimes "unknown" to police, where victims neglected to report the incident to the police or where the police failed to "officially record" an incident. In fact, changes in rates of crimes known to police may have as much or more to do with changes in the reporting practices of victims or the police than they do with changes in the underlying rate of offending.

As a consequence, social scientists involved in the study of crime and delinquency have developed other measurement techniques that reveal still more of the iceberg of crime in the United States or in smaller settings like school districts. In the pages that follow, we will take a brief look at two other sources of evidence on the extent and distribution of crime: victimization survey data and self-report survey data. Each of these survey techniques is capable of measuring criminal behavior that has never come to the attention of the police. We have already discussed two examples of the self-report survey, the Monitoring the Future study of self-reported drug use by high-school seniors and the national survey of domestic violence reported in Behind Closed Doors (Straus et al. 1980). Because the latter study also asked respondents to report incidents where they themselves were victims of violence, the research of Straus et al. serves as an example of the victimization survey as well. Since the early 1970s, the Bureau of Justice Statistics and the U.S. Census Bureau have been conducting the National Crime Victimization Survey (NCVS), which is designed to provide a nationally representative estimate of the total rate of criminal victimizations in the U.S. We will examine some of the findings of this important source of evidence on officially unreported crime. However, as shown in the diagram below, even the best designed self-report or victimization surveys will still be vulnerable to some missing data resulting from problems of sampling bias (e.g., kids not attending school when the survey was administered) or invalidity due to misreporting of deviant acts (e.g., failure to disclose spouse abuse). Thus, although we will be never be able to examine the entire iceberg of crime, we can now use self-report and victimization surveys to explore far beneath the superficial view of crime and criminals provided by data on arrests.

Official Data:

FBI Uniform Crime Reports (UCR)

  • Arrest Data
  • Offenses Known to Police
Victimization Survey Data: National Crime Victimization Survey (NCVS)
Self-Report Survey Data: Monitoring the Future Project (MTF)


  Index Page