Analysis of Missingness in UCR Crime Data

• NCJ Number: 215343
• Author(s): Michael D. Maltz
• Date Published: 2006
• Length: 21 pages
• Grant Number(s): 2004-IJ-CX-0083
• Agencies: NIJ-Sponsored
• Publication Type: Report (Study/Research)
• Award: Creating a UCR Utility

Annotation

This paper presents an analysis of the nature and type of the missing data problem in the Federal Bureau of Investigation’s (FBI’s) Uniform Crime Reports (UCR) data.

Abstract

The analysis indicated that the most populous law enforcement agencies and larger law enforcement agencies had the least amount of missing data, termed “missingness.” This is true for two reasons: (1) agencies that handle higher crime rates have greater statistical capabilities, and (2) agencies with populations of 100,000 or more are contacted by the FBI and urged to complete their data. The impact of missingness on crime statistics was analyzed at the population, State and county levels, which revealed that missingness, particularly at the county level, could lead to over half of the population being left out of crime statistics. The analysis suggests that UCR data cannot be used in an analysis without a consideration of the gaps in the data. This research was part of a larger grant project for the National Institute of Justice (NIJ) that focused on four main deliverables: (1) cleaned UCR data for the 50 States and the District of Columbia; (2) a UCR charting utility to allow users to plot monthly crime data; (3) a set of instructions to allow the addition of subsequent years of data; and (4) an analysis of the nature and extent of missingness in the UCR data. This paper provides the fourth deliverable by exploring the length of runs of missingness and how missingness varies by State, by year, and by size and type of agency. Figures, footnotes, references, appendix