Statistics: Reporting Systems and Methods

Returning to the two primary purposes of crime statistics, to measure the "amount" and "distribution" of crime, it is clear that there has been, and will probably continue to be, discrepancy among the estimates of the amount of crime that are generated by the variety of crime statistics. The dark figure of crime may never be completely illuminated, the reporting practices of victims will probably remain erratic, and the recording of crimes by authorities will continue to be less than uniform.

However, the ultimately more important purpose of crime statistics is the measurement of the distribution of crime by a variety of social, demographic, and geographic characteristics. Fortunately, the major sources of crime data—crimes known to the police, victimization surveys, and self-report surveys—generate similar distributions and correlates of crime, pointing to convergence rather than discrepancy among the measures of the basic characteristics of crime and criminals. The problems associated with each of the data sources remain, but they diminish in significance because these imperfect measures produce similar perspectives of crime. As Gwynn Nettler concluded, "Fortunately, despite the repeatedly discovered fact that more crime is committed than is recorded, when crimes are ranked by the frequency of their occurrence, the ordering is very much the same no matter which measure is used" (p. 97).

Comparisons of data from the UCR and the NCVS program show that they produce similar patterns of crime (Hindelang and Maltz). There is substantial agreement between the two measures in the ordering of the relative frequencies of each of the index crimes. Comparisons of self-reports of delinquency with crimes known to the police show that each provides a complementary rather than a contradictory perspective on juvenile crime (Hindelang, Hirschi, and Weis, 1981; Belson). Self-reports do not generate results on the distribution and correlates of delinquency that are contrary to those generated by police statistics or, for that matter, by victimization surveys. The youngsters who are more likely to appear in official police and court record data—boys, nonwhites, low achievers, youths with friends in trouble, urban residents, and youths with family problems—are also more likely to self-report higher rates of involvement in crime.

This message should be of some comfort to a variety of people interested in crime and delinquency, from researchers and theorists to policy-makers, planners, program implementers, and evaluators. The basic facts of crime are more consistent than many scholars and authorities in the past would lead one to believe. In fact, the major sources of official and unofficial crime statistics are not typically inconsistent in their representations of the general features of crime but rather provide a convergent perspective on crime. The characteristics, distribution, and correlates of crime and, therefore, the implications for theory, policy, and programs are not discrepant by crime measure, but convergent. The data generated by a variety of measures are compatible and confirming sources of information on crime. The study and control of crime can best be informed by these complementary sources of crime statistics.