With the encouragement of the federal Office of Management and Budget, the Federal Bureau of Investigation (FBI) and the Bureau of Justice Statistics (BJS) are jointly funding a multiyear, comprehensive study of the way crime statistics are generated in America. A panel convened by the National Academy of Sciences, Modernizing the Nation's Crime Statistics, is tasked with examining the current tools and methods used to generate information about crime in America, and also assessing the gaps and limitations in our knowledge about the extent of criminal behavior. The panel has three main charges:
The panel is composed of 17 members, including a number of well-known academic researchers, federal employees who are involved in other major data collection efforts (census, health), a few representatives of police agencies, and one SAC director. As a result, a wide variety of expertise and views are represented in discussions about the issues raised. These differing viewpoints are necessary, as "crime" is a somewhat slippery concept, and defining and measuring it is difficult.
Much of what we characterize as crime is based on criminal law definitions. These are used, for example, in both the FBI's Uniform Crime Reporting (UCR) system (and in the National Incident Based Reporting System - "NIBRS") as well as in BJS's National Crime Victimization Survey (NCVS). In many cases, the categories and definitions of crime have not changed substantially over the last 50 years. However, what is considered "criminal" has changed based on social changes or technology. For example, the UCR system is not fully capable of dealing with homosexual spouses as victims of same-sex rape; and domestic violence, cybercrime, and theft of intellectual property are only now being tentatively addressed by the FBI.
The FBI's UCR system collects data from the nation's 18,000 police agencies, which are charged with enforcing state laws-and these data may not translate accurately to the FBI's definitions. The quality and completeness of these data are issues that the panel will be considering.
Victimization data, as collected in the NCVS, contain information on crime that may not be reported to the police. These data are based on a sample of the population and only include responses from persons 12 years of age and older. Due to the limitation of its sample size, the NCVS is not able to provide state-level estimates of victimization.
Another area of concern for the panel is the near total lack of comprehensive data on regulatory violations. These include many areas covered by federal agencies, such as the Security and Exchange Commission, the Federal Trade Commission, the Department of Agriculture, and other agencies. (Many states have similar agencies.) The "violations" that these agencies uncover may have a much greater financial impact on the nation than other "crimes" such as bank robbery. However, no current system collects and publishes these data.
Finally, there is the issue of unreported crimes that victimize corporate America. The panel heard from one major retailer who reported hundreds of millions of dollars of "inventory shrinkage" over the course of a year. None of these cases are reported to the local police but instead are handled internally.
As part of its efforts to examine the cur- rent system of crime classification, the panel has considered international systems as well. The United Nations is currently developing an attribute-based system that may provide useful insights into alternative methods of crime classification. The panel has also heard presentations from subject matter experts and held day-long sessions with representatives of local, state, and national "parties of interest" (agencies that collect, analyze, or use crime and violation data).
An interim report is expected to be released later this year. For more information on the panel, visit the National Academies website.