EMPIRICAL ANALYSIS OF THE IMPACT OF MEASUREMENT ERROR IN UCR (UNIFORM CRIME REPORTS) DATA (FROM CRIMINAL JUSTICE STATISTICS ASSOCIATION, INC - PROCEEDINGS..., 1978 SEE NCJ-63086)

• NCJ Number: 63096
• Author(s): R W GILLESPIE
• Date Published: 1978
• Length: 10 pages
• Corporate Author: Justice Research and Statistics Association
• Sale Source: National Institute of Justice/
• Publication Type: Report (Study/Research)
• Language: English
• Country: United States of America

Annotation

THE EFFECT OF MEASUREMENT ERROR IN REPORTED CRIME ON ESTIMATIONS OF THE IMPACT OF POLICE ACTIVITY ON CRIME LEVELS IS ASSESSED.

Abstract

VICTIMIZATION SURVEYS HAVE CONFIRMED THAT OFFICIAL CRIME STATISTICS CONTAIN A SIGNIFICANT UNDERREPORTING ERROR. THIS UNDERREPORTING HAS THE POTENTIAL FOR PRODUCING BIASED ESTIMATES OF THE IMPACT OF ANY VARIABLE ON THE CRIME RATE, IF THIS VARIABLE IS ALSO CORRELATED WITH THE UNDERREPORTING ERROR. SUCH A BIAS MAY EXPLAIN, IN PART, THE DIVERGENT RESULTS PUBLISHED OF THE ESTIMATES OF THE LEVEL OF THE EFFECT OF POLICE ACTIVITY ON CRIME RATES. ANALYSIS OF VICTIMIZATION SURVEY DATA FOR LARGE AMERICAN CITIES SHOWS A POSITIVE RELATIONSHIP BETWEEN VICTIMIZATION REPORTING RATES AND MEASURES OF POLICE ACTIVITY AND EFFECTIVENESS. HENCE, ESTIMATES OF THE IMPACT OF THESE POLICE VARIABLES ON THE LEVEL OF CRIME WILL BE BIASED DOWNWARDS. THE RESULTS SUGGEST THAT IMPROVED POLICE EFFECTIVENESS, PARTICULARLY IN THE AREA OF POLICE COMMUNITY RELATIONS, INCREASES THE LIKELIHOOD THAT VICTIMS WILL REPORT THE CRIMES COMMITTED, THUS PRODUCING A POSITIVE CORRELATION BETWEEN POLICE EFFECTIVENESS AND AN INCREASE IN REPORTED CRIME. RESULTS SHOW THAT IT WOULD BE PREMATURE TO REJECT THE DETERRENCE HYPOTHESIS BECAUSE OF THE WEAK IMPACT OF POLICE ACTIVITY UPON CRIME RATES DERIVED FROM UNIFORM CRIME REPORTS CRIME RATE DATA. TABULAR AND GRAPHIC DATA AND REFERENCES ARE PROVIDED. (AUTHOR ABSTRACT MODIFIED--RCB)