U.S. flag

An official website of the United States government, Department of Justice.

Analysis of the United Nations Data Set on Crime Trends and the Operations of Criminal Justice Systems: Part I

NCJ Number
R. W. Burnham
Date Published
November 1998
32 pages
This analysis of the data collected by the United Nations on crime and criminal justice since the mid-1970’s focuses on the nature and quality of the data set, issues related to analyzing data collected in different countries, interpretation issues, and the uses and limitations of the data.
The data set includes a separate file for each year. Each file contains all the variables and cases for that year. The United Nations collected the data in five different sweeps, which covered 1970-74, 1975-80, 1980-86, 1986-90, and 1990-94. Issues related to reliability and underlying validity appear to vary from both case to case and variable to variable. Thus, the data are useful for suggesting and refining questions for further study, but not with any authority for activities hypothesis testing, which demand statistical rigor. Difficulties related to analyzing data across countries, cultures, or jurisdictions relate to definitions, recording practices, operating practices, factual inequalities, and crime recording. However, the use of ratios within countries is not vulnerable within countries. Data from different countries on attrition as a case progresses from recording through case processing through to imprisonment demonstrates the use of the data set. Using the data to analyze the precise amount of change in any variable from year to year is inappropriate. Appropriate uses include: (1) the development of descriptions of the dynamics of the main components in the criminal justice process; and (2) the generation of hypotheses. Organizations should use the data set for research and particularly to develop more precise questions. In addition, the National Institute of Justice and Bureau of Justice Statistics should urge the United Nations to improve data quality, even if the trade-off is less data quantity. Figures and appended technical details