A cautionary note on using county-level crime and homicide data

There has been a surge in recent years in the use of counties as the unit of analysis in homicide research. Employing county-level data, however, can create several problems. In some cases, these problems can be overcome with various methodological procedures. Yet these procedures are not magic, and they assume the use of valid data. Recent measurement and methodological research examining county-level data, however reveals serious reporting errors in the Supplementary Homicide Reports (SHR) data, that agreement between reporting systems at the county level is not strong, and that the amount of person time at risk in many counties is simply too small to generate a reliable estimate of the homicide rate. Despite these limitations, homicide studies using these small counties continue be carried out without heeding these previous warnings. This article is thus meant to provide a review of the limitations and to urge caution when employing county-level homicide data.