A recent Justice Center examination of sexual assault case processing shows that close to 60 percent of cases referred by the Anchorage Police Department to the Alaska Department of Law result in a conviction on some charge—although not necessarily the original charge. The number of cases referred, however, is less than 20 percent of those reported.
Anchorage has a very high incidence of reported rape. Between 2000 and 2003, the rate of reported forcible rape in Anchorage was 163 percent higher than in the U.S. as a whole. Over the past twenty years, Anchorage has been consistently at or near the top of U.S. metropolitan statistical areas for rates of reported forcible rape.
The Justice Center examined the outcomes of all sexual assaults involving one suspect and one victim reported to the Anchorage Police Department (APD) between January 2000 and December 2003. In this article we summarize how the Alaska Department of Law disposed of these reported assaults, detail case and charge outcomes and discuss the reasons given by prosecutors for these outcomes.
Overall, 1,235 sexual assaults were reported to APD from January 2000 through December 2003. We sampled all 1,074 sexual assaults committed by one suspect against one victim (87.0% of all sexual assaults reported from January 2000 to December 2003). Data were collected on 1,052 (98%) of these reports. Of these 1,052, 188 (17.9%) were referred for prosecution and had reached final disposition prior to data collection. Offenses in these 1,052 reports included sexual assaults, forcible rapes, attempted sexual assaults, and attempted forcible rapes. As defined by the Federal Bureau of Investigation’s Uniform Crime Reports, forcible rape is “the carnal knowledge of a female forcibly and against her will.” Sexual assault is a less restrictive state-defined offense that does not consider the gender of the parties involved and does not require carnal knowledge.
For this article, we consider three stages of prosecution: referral, acceptance, and conviction. These stages can be thought of as formal decision points at which prosecutors record what has transpired with the case, and why that outcome resulted. Referral is the forwarding of charges by APD to the DOL. This is the initial stage of prosecution and the first stage at which prosecutors officially become aware of a report. Once a case has been referred to the DOL, it is screened for acceptance. More specifically, prosecutors will screen each charge within each case for acceptance. The prosecutor’s initial screening decision for each charge is recorded with a screening disposition code which indicates in what fashion the charge did or did not move forward. At this decision point, in addition to disposition codes, reason codes are given to charges that are not accepted as referred by law enforcement. The reason code indicates the formal reason that prosecutors gave for the disposition of the charge in that manner.
Acceptance can be considered the second stage of prosecution, or the second formal prosecutorial decision point. At this stage prosecutors formally agree to move forward with criminal prosecution. Once a final outcome has been established, a final disposition code and potentially a reason code are attached to each individual accepted charge within a case. The final disposition code indicates the final outcome of an accepted charge. The reason code associated with this final disposition reflects the prosecutor’s reason for the final disposition of the charge. A conviction is an accepted charge resulting in a finding of guilt. A finding of guilt can occur through plea bargaining or being found guilty beyond a reasonable doubt by either judge or jury.
Table 1 describes the number of cases reported, referred, accepted, and convicted. Of the 1,052 reported cases, 188 (17.9%) were referred to DOL and 127 (67.6%) were accepted. Finally, 111 (87.4%) of these 127 cases resulted in a conviction. Obviously, the point with the greatest attrition is from report to referral.
For the purposes of this article, two separate levels of analysis are used: a case-level analysis and a charge-level analysis. For case-level analysis, a case as a whole is the unit of analysis. With this level of analysis, we are concerned only with the most serious outcome at a particular stage. The second level of analysis, at the charge level, considers individual charges within a case, using the individual charge as the unit of analysis. Using this level of analysis, we are concerned with all charges at each decision point, not simply the single charge that received the most serious outcome. This distinction is important for understanding many of the results of this project.
From referral to acceptance, there is often cross-movement of charges within a case. While it is necessary that at least one charge be referred for any case to be accepted, the number of charges within a case need not be static from referral to acceptance or from acceptance to conviction. Law enforcement may refer many charges to prosecutors, but upon review of the case, prosecutors agree to move forward with prosecution on a subset of the referred charges (i.e., the number of accepted charges is less than the number referred). Similarly, it is possible that when law enforcement refers only a single charge to prosecutors, upon review, prosecutors feel that additional charges are appropriate (i.e., they create an accepted charge that was not referred).
As mentioned earlier, of the 1,052 reported sexual assault cases, 188 (17.9%) resulted in a referral to the DOL. Fewer than half (41.5%) of the referred cases included one charge; 76.5 percent had three or fewer charges. Overall, the 188 referred cases contained 434 referred charges. Table 2 shows the distribution of these 434 referred charges by charge type. One hundred and eighty (41.5%) of the referred charges were sexual assaults in the first degree. Further, 290 (66.8%) of the 434 referred charges were sexual assaults of some degree. Fifty-two (12.0%) of the referred charges were for sexual abuse of a minor; 38 (8.8%) were assault charges; 17 (3.9%) were kidnapping charges; and 37 (8.5%) were other charges. Examples of other charges included contributing to the delinquency of a minor, driving under the influence, and possession of child pornography.
Table 3 is a breakdown of referred charges by class. Over 90 percent (n = 394) of the referred charges were felonies. Almost half of the charges (48.6%) were referred as unclassified felonies, that is, the gravest crimes under Alaska statutes, carrying with them the harshest penalties. Under the statutes, sexual assault is taken very seriously. Further, when a report results in a referral, the charges referred for prosecution are at a high level of class severity and carry with them some of the most significant penalties exacted by the justice system.
All charges referred to the DOL receive screening disposition codes. Table 4 indicates the disposition codes that were attached to the 434 referred charges. Prosecutors agreed to prosecute 68.2 percent of charges as referred by law enforcement. “Prosecution declined” disposition codes were attached to 24.0 percent (n=104) of referred charges. Interestingly, 92.2 percent (n=400) of charges were either accepted as referred by law enforcement or declined.
Here we make claims about charges, not cases. As stated earlier, a majority of cases (58.5%) have more than one charge. It is therefore possible that at least one charge in a case progresses forward, while others are dismissed.
Only charges that were not accepted as referred receive reason codes. Thus, in our sample, the 138 charges not accepted as referred received reason codes. Reasons for not accepting charges as referred are shown in Table 5. What is important to note is the proportion of reasons, by type, given by prosecutors for a charge not being accepted as referred. At this stage, evidentiary reasons were the most typical reasons (45.7%) for not accepting a charge as referred. Discretionary reasons were the second most common reasons for not accepting a charge as referred—32.6 percent.
Transition between Referral and Acceptance
Of the 434 charges that were referred for prosecution, 330 (76.0%) were accepted, either as referred or as a different charge, while the remaining 104 (24.0%) were dropped. Also, 83 new charges were created by prosecutors from referral to acceptance. This means that there were a total of 517 charges at some stage of prosecution within these 188 cases (see Table 6).
We now briefly focus on the 104 charges that were declined for prosecution and the 83 new charges that were added by prosecutors. Half (n=52) of the 104 charges not accepted by the DOL were sexual assaults in the first degree and an additional 21.2 percent (n=22) were sexual assaults in the second degree. Over half of the charges that were not accepted by prosecutors were unclassified felonies and an additional 27.9 percent were class B felonies. Of the 83 charges added by prosecutors, 32 (38.6%) were sexual assault charges, 20 were sexual abuse of a minor charges, 6 were assault charges, 2 were kidnapping charges, and 23 were other charges. New charges added by prosecutors were most commonly class B felonies or unclassified felonies.
We now turn our attention to the 413 charges that moved forward from referral to acceptance. Within the case-level analysis, 127 cases of the original 188 moved forward from referral to acceptance (see Table 1). Stated differently, prosecutors agreed to move forward on 67.6 percent of the cases referred to them. Of the 127 accepted cases, 28.3 percent contained one charge and 70.8 percent contained three or fewer charges. Overall, 413 charges moved forward between referral and acceptance. Table 7 shows the charge type of these 413 charges. As with referral, the most common accepted charge was sexual assault in the first degree—36.6 percent (n=151) of the accepted charges. Further, 58.8 percent of all accepted charges were sexual assaults in some degree.
Table 8 shows exactly how charges moved from the referral stage to the accepted stage, for the 330 charges which were both referred and accepted. What is interesting here is the remarkable consistency with which charges fall on the table’s diagonal: A charge referred and accepted as the same charge type will fall on a diagonal line from the upper-left to lower-right corners. We see that the vast majority of charges (92.4%) fall on this diagonal.
Table 9 provides the charge class frequencies for the 413 charges that were accepted by the DOL. Most charges were accepted as unclassified felonies. Furthermore, 86.9 percent of the charges in our sample were accepted at the felony level. Knowing that the vast majority of charges are accepted as referred, we could reasonably expect charge class to remain relatively constant from referral to acceptance—as Table 10 shows. For charges that were both referred and accepted, there is a great degree of class congruency between referral and acceptance.
Once a final outcome has been determined for a charge, prosecutors attach a final disposition code. Table 11 shows the final disposition codes that were attached to the 413 accepted charges in our sample. What is immediately apparent is the large of proportion of charges that are dismissed by prosecutors. Of the 413 accepted charges, 243 (58.8%) were dismissed by prosecutors. Taken as a whole, 61.7 percent of accepted charges were later dismissed, including those for which no true bill was returned. Another point becomes clear from Table 11: For 124 of the 138 accepted charges that resulted in a finding of guilt, the conviction was a result of plea bargaining. This means that plea agreements were responsible for 89.9 percent of guilty findings in our sample. To state this differently: Only 10 percent of guilty findings were a result of court action.
Charges that were dismissed by prosecutors or charges for which prosecutors allowed a plea to an amended charge are given a final disposition reason code. These reasons are shown in Table 12. The most striking result in Table 12 is the degree to which discretionary reasons are cited by prosecutors. Of the reasons given by prosecutors for dismissing a charge or allowing a plea to an amended charge, 70.5 percent were discretionary in nature. Among discretionary reasons for charge dismissal can be the choice of another charge as more accurate, the consolidation of charges, or the suspect pleading to the essence of the offense.
Of the 127 cases that were accepted by DOL, 16 cases (12.6%) resulted in no conviction. Most cases (87.4%) resulted in a conviction on at least one charge. When examining the charge-level analysis, 138 (33.4%) of the 413 accepted charges resulted in a finding of guilt. Table 13 is a frequency table of convicted charge types. Table 14 presents the changes in the types of charges between acceptance and conviction. We see much more variation about the diagonal than we did in earlier crosstabulations. This is an indication that charge types are shifting between acceptance and conviction. Given that a large proportion of cases are disposed of by plea bargaining, this is not surprising.
There is a shift in charge class similar to the shift in charge type we observed between acceptance and conviction. Table 15 presents the convicted charge class, and Table 16 demonstrates the way charge class shifted between acceptance and conviction.
Sexual assaults in some degree formed 60.8 percent of charges that were accepted but failed to result in a conviction.
As we have seen, the picture of sexual assault case processing that emerges depends, in large part, on the level of analysis. Conclusions about the prosecution of sexual assaults by the DOL vary substantially depending on which level of analysis we consider. Figure 1 succinctly illustrates the points and rates of attrition for sexual assault cases and charges at the three formal prosecutorial decision points considered in this project. Figure 1 shows that a higher proportion of charges than offenders are being removed from the system at each decision point. Furthermore, results indicate that most offenders whose cases reach prosecutors are being held accountable in some degree; that is, they receive criminal sanctions. Another fact becomes clear when looking at Figure 1: The point of highest attrition is prior to prosecutorial involvement, with 82.5 percent of reported sexual assaults not referred for prosecution. This is by far the largest filtering that occurs at any formal decision point studied in this research. However, determining what precisely is happening between report and referral is beyond the scope of the present project.
In addition to the rates of attrition shown in Figure 1, there is a substantial change in the types of charges that flow through the three decision points examined in this research (Tables 8, 10, 14, and 16). The distribution of charges by type is relatively constant from referral to acceptance, but there is a demonstrable shift in the distribution between acceptance and conviction. The most dramatic shift is observed for the sexual assault in the first degree charges. At referral, charges for sexual assault in the first degree constitute 41.5 percent of all charges. At acceptance, this proportion is relatively unchanged at 36.6 percent. However, sexual assault in the first degree constitutes only 8.7 percent of convicted charges.
Upon first inspection, this result may be startling, but it becomes readily explainable when viewed in light of previous results. We stated earlier that the vast majority (89.9%) of convictions in our sample are a result of plea bargaining, which commonly results in convictions on less serious charges. We could therefore reasonably expect that many charges in our sample would result in convictions on charges that are less severe than those originally referred or accepted.
While the reason codes may not capture all possible reasons for charge dispositions, they offer useful insight into prosecutorial decision making. A comparison of reason codes between those given from referral to acceptance (Table 7) and those given from acceptance to conviction (Table 12) shows that there is a perceptible shift in the reasons attached to case dispositions at the respective decision points. From referral to acceptance, evidentiary reasons account for 47.5 percent of reasons given for a referred charge not being accepted as referred. From acceptance to conviction, this proportion drops substantially to 10.8 percent of reasons given by prosecutors. The opposite happens for the proportion of total reasons that are discretionary in nature. Between referral and acceptance, discretionary reasons account for 32.6 percent of the reasons given by prosecutors. This proportion jumps to 70.5 percent of reasons given for a dismissal or plea agreement to an amended charge.
Overall, we began with a total 1,074 sexual assaults reported to the Anchorage Police Department between January 2000 and December 2003 involving one suspect and one victim. Data were collected on 1,052 (98%) of these 1,074 reports. These 1,052 reports to APD resulted in the referral of 188 cases and 434 charges to DOL. These 188 referred cases resulted in the acceptance of 127 cases that included 413 charges. Prosecutors obtained convictions in 111 cases on 138 charges. Thus, the majority of cases resulted in a conviction on at least one charge. For referred cases, a majority of respondents were held accountable for their actions to some degree.
It remains true, however, that a heavy majority of reported sexual assaults are not referred for prosecution. More examination of this point in case processing—between reporting and referral—is urgently needed.
Since the greatest source of case and charge attrition is from report to referral, enhancing offender accountability may depend on increasing the proportion of reported sexual assaults that are referred to the Department of Law.
G. Matthew Snodgrass is a researcher in an ongoing study of sexual assaults in Anchorage. The study is being directed by André Rosay of the Justice Center. The complete report on which this article is based is available at http://justice.uaa.alaska.edu/research/2000/0611sexassault/0611sxcaseprocess.html. This project was supported by Grant No. 2004-WG-BX-0003 awarded by the National Institute of Justice, Office of Justice Programs, U.S. Department of Justice. Points of view in this document are those of the authors and do not necessarily represent the official position or policies of the U.S. Department of Justice.