Can Social Disorganization or Case Characteristics Explain Sexual Assault Case Clearances?

Abstract

Understanding factors related to the clearing of criminal cases by law enforcement is an important, but understudied, issue in criminal justice. Through an examination of 2,437 sexual assault cases reported to the Orange County (Florida) Sheriff's Office and Orlando Police Department between 2004 and 2006, this study examines the ability of case and community characteristics to predict the outcome of investigations. Results show that ages of Victims & Offenders are significant predictors of cleared cases, but that measures of community social disorganization are not significantly related to sexual assault investigation outcomes.

Keywords:

• clearance rates
• sexual assault
• social disorganization

Notes

1. For the present study, however, since we are looking at correlates of whether or not sexual assault cases were solved, the biases associated with incomplete reporting are lessened substantially. Here we are only looking at cases that are reported and whether there are community or incident characteristics that are associated with the case disposition. It is unknowable if the same pattern of relationships would exist if all cases were reported to law enforcement.

2. The variable “case number” was present in all of the original datasets, and it was utilized as a common field from which other data could be merged or integrated. A problem with analyzing law enforcement datasets on criminal offenses is that they may include multiple case entries for a single incident because of the presence of multiple victims and/or offenders. There is a need to maintain access to these data for second and additional offenders and victims, however. In order to combine the suspect and victim datasets while maintaining the original case numbers, two additional variables were created for both suspects and victims. These newly created variables indicate the total number of suspects and victims involved in each incident and assign a unique number to each of the suspects and victims involved in the incident. These variables allow the maintenance of all information on all suspects and victims involved in an incident for analyses. If, for example, an incident contained five suspects and three victims, the first variable indicates that the total number of suspects is five, and the second variable indicates that the total number of victims is three. The other two variables contain a counting mechanism that assigns numbers from one to five for each suspect and from one to three for each victim, respectively. The incident number appears as many times as the highest number of either suspects or victims involved. The data were acquired from the OPD and the OCSO in June 2007, so there is a minimum period of five months between the reporting of a case and its designation as unsolved in the datasets.

3. The sexual assault law for Florida is found in Chapter 794 of the Florida State Statutes. “Sexual battery” is defined as oral, anal, or vaginal penetration of another's sexual organ or oral or anal penetration by any object without consent. Florida statutes are available online at http://www.leg.state.fl.us/Statutes/index.cfm?App_mode=Display_Statutes &URL=0700-0.

4. In cases with multiple victims or offenders, the incident address is the same. This is because it is the incident address that is used for geocoding, not the addresses of victims or suspects. All incidents in the dataset were assigned to a single location (i.e., address) by the OPD and the OCSO.

5. This is not the purpose of the present study, so this limitation is not important.

6. Sampson et al. (1997) used percent of female-headed households and percent of the population under 18 years old. Morenoff et al. (2001) combined these two variables with the percent of female-headed families with children. We use the latter index as it is more recent and includes more relevant measures.

7. Since all census tracts in Orange County had at least one sexual assault over the three-year period, this table provides descriptions of all sexual assault cases and all communities.

8. In addition to the 168 cases deleted because of missing information on case disposition (N = 30) and because the case was unfounded, 710 (or 31%) of the remaining cases were excluded from the analyses because of missing information on one or more variables. Not surprisingly, most of the missing information was for offenders' demographic characteristics, particularly for their age. To estimate the effect of the missing cases on the results, we first deleted offenders' age from the model. In a second step, we deleted offenders' sex and race from the model. There was no substantive effect on the size or significance of the coefficients for the remaining predictors in the model.

9. Specifically, all tolerance levels are above .6 (Menard, 1995).