Comparing the UCR and the NCVS
For victimologists, the greater variety of statistics published in the NCVS offer many more possibili- ties for analysis and interpretation than the much more limited data in the UCR. But both official sources have their advantages, and the two can be considered to complement each other.
The UCR, not the NCVS, is the source to turn to for information about murder victims because questions about homicide don’t appear on the BJS’s survey. (However, two other valuable, detailed, and accurate databases for studying homi- cide victims are death certificates as well as public health records maintained by local coroners’ and medical examiners’ offices. These files may contain information about the slain person’s sex, age, race/
ethnicity, ancestry, birthplace, occupation, educa- tional attainment, and zip code of last known address [for examples of how this non-UCR data can be analyzed, see Karmen, 2006]).
The UCR is also the publication that presents information about officers slain in the line of duty, college students harmed on campuses, and hate crimes directed against various groups. The UCR is the place to go for geographically based statistics; it provides data about the crimes reported to law enforcement agencies in different towns and cities, entire metropolitan areas, and counties, states, and regions of the country. NCVS figures are calculated for the whole country, four geographic regions, and urban/suburban/rural areas, but are not available for specific cities, counties, or states (because the subsamples would be too small to analyze). The UCR, but not the NCVS, calculates the overall proportion of reported crimes that are solved by police departments. Incidents counted in the UCR can be considered as having passed through two sets of authenticity filters: Victims felt what happened was serious enough to notify the author- ities shortly afterward, and officers who filled out the reports believed that the complainants were telling the truth as supported by some evidence. Although limited information about arrestees is provided in the UCR, this annual report doesn’t provide any descriptions of the persons harmed by those accused rapists, robbers, assailants, burglars, and other thieves (until the NIBRS replaces current record-keeping formats).
NCVS interviewers collect a great deal of information about the respondents who claim they were harmed by street crimes. The NCVS is the source to turn to for a more inclusive account- ing of what happened during a given year because it contains information about incidents that were but also were not reported to the police. The yearly surveys are not affected by any changes in the degree of cooperation—or level of tension—between com- munity residents and their local police, by improve- ments in record-keeping by law enforcement agencies, or by temporary crackdowns in which all incidents are taken more seriously. But the NCVS interviewers must accept at face value the
76 CH APT ER 3
Copyright 2016 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-203
accounts respondents describe. Also, the NCVS annual report has nothing to offer about murders, line-of-duty assaults and deaths of police officers, offenses committed against children under 12, robberies and burglaries directed at commercial establishments, and injuries and losses from inten- tionally set fires.
Even when both of these official sources col- lect data about the same crimes, the findings might not be strictly comparable. First of all, the defini- tions of certain offenses (such as rape) can vary, so the numerators may not count the same incidents. The UCR kept track only of rapes of women and girls (sexual assaults against boys and men were considered as “forcible sodomy”) until a gender- neutral definition that explicitly described all forms of intrusive bodily invasions was adopted in 2011. The NCVS counts sexual assaults against males as well as females. Similarly, the definitions of rob- bery and burglary are not the same in the two official record-keeping sources. For example, the UCR includes robberies of commercial establish- ments and burglaries of offices, but the NCVS does not.
In addition, the denominators differ. While the FBI computes incidents of violence “per 100,000 people,” the BJS calculates incidents “per 1,000 people age 12 or older.” For property crimes, the NCVS denominator is “per 1,000 households,” not individuals (the average household has between two and three people living in it).
Therefore, it is difficult to make direct compar- isons between the findings of the UCR and the NCVS. The best way to take full advantage of these two official sources of data from the federal government is to focus on the unique information provided by each data collection system.
Even when the new and improved NIBRS data storage system is taken into account, several important variables are not tracked by any of these government monitoring systems. For exam- ple, information about the victims’ education, occu- pation, ancestry, birthplace, and rap sheet is not collected by the UCR, the NIBRS, or the NCVS. Yet these background variables could be crucial to investigate certain issues (like robbers preying on
recent immigrants or murders of persons known to be involved in the drug scene).