Quality Data Is the First Step: The National Homicide Data Improvement Project

The goal of this study is to improve dramatically the quality of data on homicides for the United States since 1959. By augmenting the data gathered in recent years by the National Violent Death Reporting System, and by using statistical tools to estimate the number of homicides that occurred in various social, demographic, and geographical groups, the Project will for the first time give scholars and law enforcement officers a highly accurate long-term homicide series. Preliminary analyses document the extent to which reliance on extant datasets has distorted knowledge of homicide in the United States. For some categories of homicide, the researchers have found that merging police and vital statistics homicide records at the state level can increase the tally of homicides by as much as 30 percent over the numbers available from police reports alone. By combining (a) data reported by police to the FBI, with (b) data reported by health officials to their state vital statistics agencies and to the National Center for Health Statistics and the National Child Abuse and Neglect Data System, augmented when necessary by (c) internal police records and newspaper accounts, the project will create homicide data series that are as complete as possible. The research will be conducted initially in Ohio and then in four more 'open record' states (California, Connecticut, Massachusetts, and North Carolina), so named because their death records are available for examination and chosen because they represent four different regions of the country.

This research may improve the development and implementation of policies concerning child abuse, domestic violence, and homicides among unrelated adults, and will permit the testing of hypotheses concerning the impact of homicide rates on clearance rates and other forms of deterrence.