Official crime rates and victim surveys: An empirical comparison

Abstract

This paper compares the Uniform Crime Reports (UCRs) to the results of the National Crime Survey (NCS). Six Part I offenses are used to form the basis for comparison across twenty-six large American cities. The two data sources are conceptually comparable. Empirically, the NCS reveals significantly higher rates for all offenses but auto theft. The two sources are compatible in that the UCR revealed the same distributional pattern as the NCS. Thus it was concluded that the UCR provides an accurate indicator of the relative occurrence of criminal acts.