A Review: An Evaluation of Lifetime Likelihood of Victimization

Abstract

Popularizing findings from complex surveys is a difficult balancing act. On the one hand, the integrity of the data must be preserved and professional standards of accuracy observed. On the other hand, the data must be used to inform issues of interest to a lay audience and presented in a way that is appealing. Lifetime Likelihood of Victimization (Koppel, 1987), a Bureau of Justice Statistics (BJS) technical report, attempts to strike this balance. In many ways it is successful, but in the end it tips slightly to the sensational. The basic idea is good. The author takes issue with the exclusive use of annual rate estimates as a method for assessing risk of victimization. He argues that people are more interested in their risk of ever being victimized than their risk in a given year. The bulletin presents estimates of the probability of being a victim over a respondent's lifetime based upon age-specific victimization rates from the National Crime Survey (NCS). There are two customary approaches to estimating the lifetime likelihood of victimization. One involves asking respondents if they had ever been victimized. This approach is not desirable, however, because we know that respondents will fail to report many crimes that occurred even a few weeks prior to the interview. This recency bias will result in a substantial underestimate of the lifetime prevalence of victimization. A second, and more acceptable approach, involves the collection of panel data on the victimization experience of persons over a long period of time. Since neither type of data were available to BJS, they turned the NCS data to a purpose that the data were never meant to serve. The NCS surveys approximately 60,000 households annually. All members of the household 12 years of age and older are asked to report their victimization experience in the previous six months. Interviewers return to the housing unit, but not necessarily to the same respondents, every six months for three years. The resulting reports are used to