Homicide: Behavioral Aspects

Homicide data generally derive from either health or police agencies. There are two major sources of international data; one complied by the United Nations in World Health Statistics Annual and the other by the International Criminal Police Organization (Interpol), which was established in 1950. The national police agency of each country reports the number of that country's homicides for every two-year period. World Health Statistics Annual publishes the cause of death, including homicide, for each reporting country. These statistics, which have been collected since 1939, are the joint product of the health and statistical administration of many countries and the office of the United Nations and the World Health Organization. Problems in the use of these sources include lack of consistent definitions and interpretations across jurisdictions and lack of consistent reporting by all countries. Some countries, including most in Africa and many in Asia, do not routinely report (LaFree). Furthermore, there are few validation procedures to assess the accuracy of the data. For a summary of difficulties with these data sources, see LaFree.

Within the United States there are two major national sources of data on homicide: the National Center for Health Statistics (NCHS) and the Federal Bureau of Investigation Crime in the United States (known as the Uniform Crime Reports, which is published annually). The NCHS data derive from coroners and medical examiners, who forward death certificates to the center's Division of Vital Statistics. These data focus solely on the homicide victim and generally include information on the cause of death and the age, race, and sex of the victim. Data about offenders, victim offender relationships, and motives are not included. The various states entered this national reporting system at different times. Prior to the 1930s, when the system became fully national, the data available depended on which states and cities were included. Boston was the first entrant, and in general there were data from the East Coast cities very early. Boston had death data in 1880, Pennsylvania in 1906, and Washington, D.C., in 1880. Other states, such as Georgia and Texas, entered the registry much later—in 1922 and 1933, respectively. In establishing trends, then, there is difficulty in obtaining national data before 1930.

The Uniform Crime Reports, a voluntary national data-collection effort, began in 1930 and gradually accumulated reporting police districts. Homicide reports are detailed and include information on both victims and offenders and, since the 1970s, on victim-offender relationships. This system is the only national one with information on homicide offenders and includes information on crimes classified by size of population, state, county, and Standard Metropolitan Statistical Area. Although there are some problems with the use of the Uniform Crime Reports data, they are commonly used in studies of homicide.

In general, information on the number of homicides in the nation reported by NCHS and the UCR show relatively high agreement. However, there is variation, sometimes substantial, when comparing the two sources on such questions as age or ethnic background of the victim. A discussion of these differences are found in Riedel.

Although the UCR and NCHS are the most commonly used national data sources for studies of homicide, the National Incident Based Reporting System (NIBRS), which is under development, may become an important source in homicide research. This system originated as a result of the 1982 F.B.I. Bureau of Justice Statistics task force comprehensive evaluation and redesign of the UCR program. When fully implemented there will be more detailed data at the national level than are currently available (Reidel).

In addition to these national sources, researchers have records of specific homicide cases, available either from medical examiners' offices or from police departments. Such records are richer in detail than those at the national and international levels and provide more specific evidence on time and location of homicides, alcohol and drug involvement, sequence of events leading to victim-offender confrontations, and the like. Locally based data can be used to augment those compiled nationally and are useful for describing homicide events in detail. There is extensive long-term city level information available from police records in Chicago (Block), St. Louis (Decker, 1993; Rosenfeld), Philadelphia (Wolfgang; Zahn, 1997) as well as other cities. These data are available in analyzed form from the publications of the authors, and are also available from the Inter-University Consortium for Political and Social Research (ICPSR) at the University of Michigan and at their web site (http://www.umich.edu). Most of these studies rely on the model established by Marvin Wolfgang in his classic study Patterns in Criminal Homicide (1949).