This chapter discusses outcome measures grouped by the goals commonly sought by sexual violence response systems. These goals include:

• Reducing the number of sexual violence crimes in the community.
• Increasing the number and percentage of cases that are prosecuted.
• Defining and categorizing successful case resolutions.
  • Successful case resolution is where: (a) case complexity has not improperly resulted in declination, discharge, or acquittal; (b) promising prosecutions practices have been used; (c) the perpetrator has been held accountable; and (d) the victim’s experience with criminal justice professionals has been positive.
• Protecting victims and witnesses.
• Helping victims feel respected by criminal justice professionals.
• Delaying fewer cases.
• Enhancing public trust in case management (e., the prosecution’s decision-making and practice is research-informed and data-driven, and outcomes are discussed transparently).

Included below is an explanation for how each outcome measure helps capture the overall progress made toward each desired goal. Likely data sources for each measure also are presented. A list of all of measures categorized by goal is provided in Exhibit 2-1.

Please note that the availability of data to support these measures may be significantly affected by many factors, including the data collection practices of each partner agency, the sophistication of data management systems, and the level of collaboration between agencies.

Many jurisdictions will likely find it very difficult to track all the measures listed below, at least initially. It is not necessary to collect data for all of the below outcome measures. The measures can be implemented in phases. To start, stakeholders should work jointly to prioritize a set of measures they would like to begin tracking. They could then focus on (1) implementing systems to track these data, and (2) training staff to record them. Other measures could then be added over time.

While the below measures provide a solid foundation for sexual violence response systems and prosecution performance management, the measures may be adapted or tailored by a jurisdiction to fit its unique needs.⁴ It is not necessary for a particular jurisdiction’s identified measures to fit squarely within any of the measures described below. If you are in interested in adapting these measures to suit your own jurisdiction’s practice, please contact the authors of this volume for assistance.

GOAL 1: The number of sexual violence crimes is reduced.

1. Total number and percentage of sexual violence crimes, both reported and unreported.

This is a very powerful outcome measure to track. However, assessing this with any degree of accuracy will most likely require use of either a community-wide survey or national survey providing reliable data at the community level, conducted on a regular basis, and including questions specific to sexual violence.

Unfortunately, few communities are likely to have such data. If such data become available, prevalence rates (i.e., cases per 1,000 residents or percentage of cases reported) could be derived from the survey. The number of sexual violence crimes would be estimated by multiplying the percent derived from the survey by the estimated population of the community.

For example, if 24 percent of survey respondents reported that they had been sexually assaulted in that particular jurisdiction during a particular timeframe, and there are 10,000 people in the adult population, the estimated total number of sexual violence crimes in the jurisdiction, both reported and unreported, would be 2,400. (.24 x 10,000 = 2,400).

Although it is unlikely that most agencies will immediately be able to collect the data necessary to measure the total number of reported and unreported sexual violence crimes, this measure is included in the RSVP Model because it provides critical information to guide our criminal justice response. It is well-known that prevalence of sexual violence in a jurisdiction is greater than the number of sexual violence incidents reported to law enforcement. Nevertheless, capturing a number, even a rough estimate, makes the theoretical concrete.

If we understand the extent of sexual violence in our communities, we will be able to better guide prevention efforts and more accurately understand the percentage of sexual violence cases that go unreported. We will also be able to assess changes in sexual violence incidence that may occur over time following the introduction of new practices or policies addressing sexual violence response.

While tracking the total number of sexual violence crimes may seem impossible initially, additional sources of data may become available in the future, eventually making this measure feasible. Findings from national surveys and special one-time local surveys may provide rough indicators, but will be less useful for assessing changes over time.

Likely data source(s): This outcome would likely require surveying a large sample of adults, with carefully crafted questions and a very specific definition of sexual violence crimes. The most feasible approach might be to add a small number of questions to a community survey already in use, which would greatly reduce the cost of data collection. Another cost-effective option would be to collect such information every other year, or even every third year, though the data would be less timely. With successive surveys, respondents would be asked about sexual violence occurring within the period since the most recent, previous survey. This activity will be most successful when undertaken by researchers, and, therefore, can be a catalyst for expanding collaboration in the community.

2. Total number of known victims.

This is the sum of the number of victims who reported to law enforcement (Measure 3 below) and the number who had reported the assault to another agency but NOT to law enforcement (Measure 4 below). This number may include some double-counting attributable to victims who initially decline to report to law enforcement but later decide to report. When such cases are known to law enforcement, that case should be deducted from the unreported cases. When Measure 1 (above) is not available – as it most likely will not be for many jurisdictions — this may be the best available estimate of the overall prevalence of sexual violence in the community. (Note that this measure does not include those incidents for which victims did not report to any of the partner organizations).

Likely data source: See data sources under Measures 3 and 4 below.

GOAL 2: The number and percentage of sexual violence cases prosecuted is increased.

3. Number and percentage of sexual violence cases reported to law enforcement, including police departments, sheriff’s agencies, and campus/school police.

These numbers represent the measure that has commonly been used, in the absence of the above-mentioned population survey data, to indicate the prevalence of reported sexual violence in a community. The percentage refers to the total number of reports, divided by the most current estimate of the community’s adult population (yielding a per capita reported sexual violence rate).

Likely data source: The number of sexual violence crimes reported to police is available from police records. For the percentage measure, the most current community adult population estimate is likely to be available from census surveys or from the local planning department.

4. Number and percentage of sexual violence cases reported by victims to a health or victim service agency, public or private, but not to law enforcement.

Such cases include those in which the victim requested a sexual assault medical forensic exam or sought help from a community-based advocate, but declined to report to law enforcement. It would be helpful to understand the reasons why victims declined to report the assault to law enforcement (if recorded by the medical forensic examiner or the advocate), because knowing this information could assist in targeting areas for improvement. See Chapter 3 for more on how to collect this information. The percentage is calculated by dividing the number of cases in which victims indicated they would not report the assault to the police, by the sum of this number plus the number of cases that were reported to law enforcement (Measure 3 above).

Likely data source: The number of sexual violence incidents and number of assaults for which the victim declines to report to the police are not commonly tracked by health or victim service organizations. To utilize this measure, the site will need to set up a process for routinely obtaining this information from partner agencies in a way that does not compromise victims’ identities. The only data needed for each reporting period here are the number of assaults reported only to medical or victim service agencies and subtotals of the reasons for declining to report to law enforcement. Personally identifying information is not needed to track this performance metric.

a. Number and percentage of sexual violence cases reported by victims to a health or victim service agency, public or private, but not to law enforcement, categorized by reason for not reporting to law enforcement. As mentioned above, it would be helpful to understand the reasons victim choose not to report sexual violence to law enforcement. By identifying the most prominent reasons, criminal justice professionals can take steps to address them. See Chapter 3 for more on how to collect this information.

5. Number and percentage of sexual violence cases reported to law enforcement that are not referred by law enforcement to the prosecutor’s office.

This measure may indicate the need to provide training and/or technical assistance to law enforcement or for better communication and coordination among partners. The percentage is calculated by dividing the number of cases not referred to the prosecutor’s office by the number of sexual violence cases reported to law enforcement.

Likely data source: Law enforcement records.

a. Number and percentage of sexual violence cases reported to law enforcement that are not referred by law enforcement to the prosecutor’s office, categorized by reason for not referring to the prosecutor’s office. Measure 5 may be more helpful if it is categorized by reasons given for not referring a case, such as insufficient evidence, inability to disprove a claim of consent, victim declined to speak with police, etc. See Chapter 3 for more on how to collect this information.

6. Number and percentage of cases declined by the prosecutor.

This percentage is calculated by dividing the number of sexual violence cases declined by the prosecutor by the number of cases referred to the prosecutor from law enforcement.

Likely data source: Prosecutor’s office records.

a. Number and percentage of cases declined by prosecutor, categorized by reason for declining to prosecute. The above measure may be more helpful if it is categorized by reasons given for declining to prosecute, such as insufficient evidence, outside the statute of limitations, inability to locate the victim or witnesses, etc. See Chapter 3 for more on how to collect this information.

7. Total number and percentage of cases declined, whether by law enforcement or the prosecutor.

The percentage is calculated by dividing the total number of reported cases declined by either law enforcement or the prosecutor by the total number of cases reported to law enforcement.

Likely data source: Law enforcement and prosecutor’s office records. For this number to be accurate, offices must ensure that they have access to complete and reliable data related to the number of cases referred – formally or informally – by law enforcement.

GOAL 3: Prosecutors acknowledge and utilize more comprehensive metrics of case success.

8. Number of persons charged with sexual violence crimes and percentage convicted of that charge (the “conviction rate”).

This is the outcome measure most often used as the primary measure for assessing sexual violence case outcomes. A considerably more accurate and fairer assessment can be obtained by factoring the complexity of cases into conviction rates. See Chapter 5 for more information. In so doing, the overall conviction rate would be supplemented by tabulations of the number and percentage of convictions for each level of complexity (that is, the percent of cases convicted that are rated as highly complex, the percent of cases convicted that are rated as low complexity, etc.).

Likely data source: Prosecutor’s office records.

9. Number and percentage of cases rated as: (a) fully successful; (b) partially successful; and (c) fully unsuccessful based on multiple measures of case success.

Conviction rates, of course, are of major importance in determining success. However, many other case resolutions are possible that indicate varying degrees of success. The conviction rate outcome does not account for cases resulting in dispositions to other charges or for lesser sentences. The conviction might be for a less serious charge (g., attempted rape rather than forcible rape). Pleas may be accepted that involve concessions of various kinds. The success level also will likely be affected by case complexity and prosecutors’ expectations of what is possible to achieve. As such, three measures of success should be combined to understand the success of case outcomes: Measure 9a and Intermediate Measure 9b (below), and Measure 10 (below).

a. The number, percentage, and disposition of cases accepted for prosecution by case resolution. Chapter 6 provides a comprehensive discussion of how to consider case disposition as an important (though not sufficient) case outcome to consider. Though not the only measure of case success to consider, case resolution remains a primary way that prosecutors examine cases for success.

Likely data source: Law enforcement and prosecution data for case resolution, as well as data to be created by prosecution for measures of implementation of best practice.

b. An intermediate measure based on the use of prosecution practices that are research-based and trauma-informed (“best practices”). The process of prosecuting a sexual assault case is arguably as important as its outcome. Even if a case results in a resolution that falls short of the charges or sentence pursued by the prosecutor, the implementation of best practices throughout the life of the case, from initial evaluation and charging through resolution, can generate a high quality of procedural justice for the victim and the public. A comprehensive definition of case success should thus account for the prosecution’s efforts to bring about justice. This can be captured as an intermediate outcome measure.

A “best practices” checklist can function as a tool to measure the level of procedural justice in our cases. It can also be used as the basis for regular meetings between unit chiefs and line-level prosecutors to examine the actions taken throughout the case and identify areas for improvement.

A comprehensive explanation of best case-level practices for prosecuting sexual violence crimes can be found in Chapter 4 of RSVP Volume 1. Exhibit 7-1 provides a sample checklist based on some of these critical practices, with a sample rating scale for each. Both the checklist and the rating scale can be modified based upon a jurisdiction’s needs.

Likely data source: Prosecutor self-report and review and modification by supervisor during evaluation.

10. Percentage of victims who felt that justice was served in their case.

This measure captures a victim’s overall perception of justice — based on both case process and case outcome. One victim may find a conviction on the highest possible charge as just, while another may find a conviction on a lesser charge, or even an acquittal, as just. These perceptions vary based on a number of factors, including someone’s cultural background, their relationship with the person who committed the assault, their belief system, etc. It is important to not only capture the criminal justice system’s perception of case success, as defined above, but victim perceptions of justice as well.

Likely data source: Surveys of victims conducted soon after disposition. See Chapter 9 for more details on suggested survey content and process.

GOAL 4: Victims and witnesses are protected.

11. Percentage of cases in which the victim was threatened, by the offender or the offender’s allies, broken down as follows:

a. Threat before the sexual violence crime.
b. Threat during the sexual violence crime.
c. Threat after the sexual violence crime.

Measures could be further categorized by the timing of the report of the threat to law enforcement:

a. Threat reported before the conclusion of the case.
b. Threat not reported until after conclusion of the case (e., first reported during follow-up survey).

Likely data sources: [a] Victim statements reported to law enforcement or to any of the other partners; and [b] surveys of victims conducted soon after disposition. See Chapter 9 for more details on suggested survey content and process.

12. Percentage of cases in which the victim reported being threatened by the offender or the offender’s allies, after case disposition.

It is important for prosecutors’ offices to consider victim safety and to work with victim advocates as needed to ensure victim needs are being met after the case has been resolved.

Likely data source: Surveys of victims conducted soon after disposition. See Chapter 9 for  more details on suggested survey content and process.

GOAL 5: Victims are satisfied with how they were treated by criminal justice professionals.

13. Percentage of victims who rated their overall experience with the sexual violence case handling as good or excellent.

This measure includes the input from victims themselves, allowing us to understand their perspectives about case handling processes and the resolution of the case.

a. Percentage of victims who rated the performance of the advocacy agency they dealt with as good or excellent (rather than fair or poor).
b. Percentage of victims who rated the performance of the SANE program/medical forensic examiner/medical personnel as good or excellent (rather than fair or poor).
c. Percentage of victims who rated the performance of the police (including investigators) they dealt with as good or excellent (rather than fair or poor).
d. Percentage of victims who rated the performance of the prosecutor’s office as good or excellent (rather than fair or poor).

Likely data source: Surveys of victims conducted soon after disposition. See Chapter 9 for more details on suggested survey content and process.

GOAL 6: Fewer cases are affected by avoidable delay.

14. Average case processing time from initial report to law enforcement to arrest to case resolution/disposition; and/or number and percentage of cases with delays.

Delays should be quantified (g., “by X days”) and preferably broken down by step where the delay occurred (e.g., submission of SAK evidence to the crime lab for testing, forensic testing of kit evidence and DNA analysis, receipt of lab reports, referral to prosecutor, trial delay).

Likely data source: Law enforcement and/or prosecutor’s office records.

15. Number and percentage of cases in which: findings from sexual assault kits were not available in time to be useful for investigatory or prosecutorial purposes.

Kits were either not appropriately stored or chain of custody was not fully documented, or sexual assault kits were misplaced. This measure focuses on the processing time for sexual assault kits, on the assumption that timeliness of sexual assault kit submission is a major source of case delays. The measure also focuses on processing quality by tracking cases in which the handling of sexual assault kits or other evidence did not follow proper protocol.

Likely data source: Law enforcement and/or prosecutor’s office records.

16. Number and percentage of cases in which the forensic lab required over “Y” days to provide its findings; and/or average lab processing time.

Some states may have set a legal limit on the amount of time a lab has to test a kit and upload it its findings into the Combined DNA Index System (CODIS). The value of “Y” could then be that amount of time. For jurisdictions without such statutory requirements, the stakeholders involved in the response system could jointly decide what threshold they want to set for this requirement.

Likely data source: Law enforcement and/or prosecutor’s office records.

GOAL 7: Public has trust in handling of cases (i.e., the prosecution handles cases competently, the prosecution explains their practices as based upon research and data collection/analysis, and the process is perceived as transparent).

17. Percentage of the population reporting trust in the way cases are handled.

Likely data source: The data for this outcome can be measured by including a question in the same community survey conducted to obtain data for Measure 1.

18. Ratings by allied professionals of the overall performance of the prosecution of sexual violence cases as either good or excellent.

Judicial, law enforcement, sexual assault forensic examiners/healthcare providers and advocate evaluations, while enlightening, should be reviewed with an awareness that these and other allied professionals, as well as the public, may have gaps in their understanding of sexual violence and the principles underlying the RSVP Model. Negative evaluations should be discussed to determine whether they reflect problems in the prosecution response or whether they suggest a need for partner, judicial, or perhaps even community education or training. In any case, collecting data for this outcome measure provides a good basis for increased transparency and enhanced interagency dialogue.

a. Number and percentage of judges who rated the overall performance of the prosecution of sexual violence cases as good or excellent.
b. Number and percentage of law enforcement who rated the overall performance of the prosecution of sexual violence cases as good or excellent.
c. Number and percentage of advocates who rated the overall performance of the prosecution of sexual violence cases as good or excellent.

Likely data source: Surveys of judges, law enforcement, and advocates responsible for criminal sexual violence cases during the reporting period. The feedback information should be confidential. These surveys would likely be done on an annual basis. Respondents would be asked about the prosecution’s handling of sexual violence cases collectively, not about individual cases or prosecutors. The survey would also prompt them to evaluate specific aspects of case processing, possibly specialized to expertise – such as victim treatment and notification, prosecutor preparation, direct examination, cross-examination, knowledge of the law, and knowledge of the rules of evidence – as well as to provide suggestions for improving the handling of sexual violence cases.

⁴ One of the Sexual Assault Justice Initiative (SAJI) pilot sites, for example, adopted several new measures to meet its jurisdictional needs, including “number and percentage of sexual battery cases that resulted in dismissal by the court due to the defendant’s mental health.”

* This measure is labeled IOM rather than OM because it reflects a intermediate outcome measure, as opposed to a measure representing the ultimate outcomes of such efforts.