hospitalization following crime. Fully half of older injured victims, compared to about 25 percent of younger injured victims, required hospitalization. Moreover, elderly victims were more likely than younger victims to be assaulted or robbed by a stranger and were more likely to be victimized in or around their home. Half of elderly victims, compared to 26 percent of those under 65, experienced violence in or around their homes and were more likely than younger adults to face offenders armed with a gun (Bachman, 1992). Elderly men were at greater risk of violent crime than elderly women. Low income, minority racial status, and geography also contributed to increased risk of assault (Bachman, 1992). For example, African American older adults were victimized at twice the rate of Caucasian elderly, and older adults living in urban settings were three times as likely to experience crime.

McCabe and Gregory (1998) used the FBI’s NIBRS to assess crime against the elderly. This system differs from the UCR in that each incident, not only the worst incident, of crime is recorded. Moreover, like the NCVS, the NIBRS includes information on the perpetrator’s relationship to the victim, permitting assessment of abuse versus assault rates. The NIBRS also includes demographic and gender information, providing some ability to conduct risk-factor research. Finally, the NIBRS differs from the UCR in that additional, nonindex crimes are also covered. Unfortunately, only crime reported to police is included in these records.

An advantage of using criminal justice system (CJS) statistics is its nationwide data collection frame. That is, many CJS studies are actual population derivations, not sample estimates. In addition, information on reported (to police) crime includes data regarding gender, race, and perpetrator status. Moreover, older adults are more likely to report some forms of crime to police than younger adults, increasing the relative validity of published rates of reported crime. However, crimes of abuse and neglect are less likely to be reported, mitigating this advantage somewhat.

In contrast to these strengths, CJS data generally have very poor sensitivity (excepting the NCVS). Furthermore, CJS data collection requires criminal justice system interaction for case identification (excepting NCVS), an activity that may be specifically avoided by older adults. Another weakness is that UCR and NIBRS data are entirely record-based and are removed from direct reports of victims. As a result, they are affected by subjective interpretations by police officers of (1) whether an event actually occurred and (2) classification of the event by police departments across the country. Overall, these forms of assessment methodology represent preliminary, as opposed to comprehensive, epidemiological data regarding elder mistreatment.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement