The statutory requirement that arrests be made for all crimes involving domestic violence (Alaska Statute 18.65.530) is an important dimension of the state's efforts to reduce violent victimization occurring among intimate partners. This policy is rooted in research that suggests mandatory arrest has a specific deterrent effect in cases of misdemeanor assault committed by males against their female intimate partners. Although there is reason to believe mandatory arrest may reduce re-victimization in its own right, the impact of the policy upon further criminal justice system intervention is largely unknown. On the one hand, it is possible that caseload pressures make it difficult for prosecutors to continue with zero-tolerance in intimate partner violence (IPV) cases. But it is also possible that mandatory arrest sets the tone for continual rigorous enforcement of IPV assault cases throughout the remainder of the criminal justice process. In this article we consider police, prosecutor, and court decision-making about IPV cases initially dealt with by the Alaska State Troopers (AST or Troopers) in 2004.
Information gathered from AST and Alaska Department of Law (DOL) case files was analyzed for this article to examine IPV assault case processing decisions. Two basic types of analyses were conducted. We first charted the processing of IPV assault cases as they made their way from the Troopers through the court system. We also used commonly employed categories (i.e., Alaska Native vs. Non-Native and isolated location vs. non-isolated location) to determine if there were variations in the processing of IPV cases relative to the victims' race or geographic location. Together, these analyses allow us (1) to understand how likely it is that IPV assault cases from the Troopers' area of responsibility are fully prosecuted, and (2) to begin to determine if particular aspects of the criminal justice environment have an impact upon the likelihood that IPV assault cases are dealt with fully.
The results presented in this article allow us to consider the attrition of cases of IPV assault reported to the Troopers and prosecuted by the DOL. For our purposes here, case attrition refers to the manner in which criminal cases are screened out of the criminal justice process as they are considered at successive decision making points. By definition, for a specific offense type, the number of cases resulting in conviction is smaller than the number of prosecutions which, in turn, is smaller than the number of arrests. Put another way, the greater the proportion of cases screened out of the process, the higher the rate of case attrition.
This article focuses upon IPV rather than domestic violence because the former is more indicative of the particularly pernicious power-based relationship violence that police and prosecutors have perpetually struggled against and that has served as the impetus for mandatory arrest policies. Under Alaska state law (Â§18.66.990), domestic violence is inclusive of all violent offenses committed by one household member against another household member where household members are broadly defined to include individuals who are related to one-another through common ancestry or marriage, unrelated individuals who are roommates or live-in caretakers, and individuals who are or were intimate partners (i.e., suspects and victims involved together in marital, dating, and/or sexual relationships). From a conceptual standpoint, it makes sense to consider IPV assault separately because it is qualitatively different compared to other types of violence that fall under the umbrella of domestic violence (e.g., child abuse, elder abuse, teenage brothers engaged in fisticuffs) in terms of its etiology and in terms of the nature and severity of official responses.
Likewise, the results presented below focus only on assault cases involving female victims because of the dissimilarities in the IPV assault committed against men compared to that committed against women. (IPV is not a purely heterosexual phenomenon. The analyses presented below include the cases of three women who reported being assaulted by their female partners.) Even though both women and men are violent toward one another when in intimate relationships, the character of women's violence against their male partners is substantially different than men's violence against their female partners in terms of its purpose and its potential for physical harm. Generally speaking, men are much more likely than women to employ physical violence in an instrumental fashion, using it as one method among many to maintain power and control over an intimate partner. Furthermore, the likelihood of physical harm is substantially higher in IPV assault against female victims relative to that of male victims because of the physiological differences between the sexes in terms of size and strength. Given these differences we would expect the victim's sex to have a differential impact on the probability that cases would be screened out during the criminal justice process.
Overall Attrition of IPV Cases
In Figure 1 we consider IPV assault cases reported to AST in 2004 as they made their way through the criminal justice process. First, we see that IPV assault cases made up a majority of the assaults classified as domestic violence. Roughly two-thirds of reported domestic violence cases (65%) involved an assault by one member of an intimate relationship against another member. The remainder of the domestic violence cases included victims and suspects that were connected by some other form of household membership. Of the IPV assault cases reported to police, the large majority (80%) involved a female victim.
Nearly all of the IPV assault cases against female victims that came to the Troopers' attention were founded (i.e., deemed to have actually occurred and classified as "crimes known to the police"). Less than two percent of the time was an IPV assault case against a female victim classified as unfounded after further investigation by AST. Some comparison figures are found in Alaska Department of Public Safety data, Crime Reported in Alaska, 2004. When compared to the founding rates for all assault cases reported to AST in 2004 regardless of victim-suspect relationship, IPV assault cases against female victims were more likely than simple assaults to be founded (87%) and just as likely as aggravated assaults to be founded (97%).
After establishing that an assault did occur, the next two steps in the process involve decisions about carrying a case forward for prosecution. An investigating Trooper first has to decide whether to refer the case to DOL and then, once referred, a DOL attorney has to choose to accept the case and proceed with prosecution. We see in Figure 1 that a large majority (86%) of IPV assault cases that were known to AST were referred to DOL for prosecution and a slightly larger proportion of those referred cases were accepted by DOL attorneys for further adjudication. Ultimately, of the cases that DOL attorneys accepted for prosecution, most (80%) resulted in a conviction.
To put these rates of prosecution and conviction into context, comparisons can be drawn with similar rates found for other U.S. jurisdictions as published in a recent meta-analysis by Joel Garner and Christopher Maxwell in Criminal Justice Review 34(1) in 2009. By any measure, it appears that there is substantially less case attrition for IPV assaults reported to AST than any other place in the U.S. where it has been studied. For example, across the studies considered by Garner and Maxwell, the proportion of IPV assault cases known to the police that were not accepted for prosecution was 72 percent. In comparison, only 23 percent of the IPV assault cases known to AST in 2004 were not accepted for prosecution. When the basis of comparison is acquittals per prosecutions, we see that DOL attorneys were much less likely to lose IPV assault cases than what has been reported elsewhere; only 20 percent of the IPV assault cases prosecuted by DOL resulted in acquittal while, on average, half of the cases prosecuted elsewhere in the U.S. did not result in conviction. Overall, of the IPV assaults known to the police, the cases dealt with by AST and DOL were much more likely to result in conviction than what has been found in other U.S. jurisdictions. In 2004, 61 percent of all IPV assault cases known to AST ultimately resulted in a conviction whereas the average conviction rate for founded cases in Garner and Maxwell's meta-analysis was 14 percent. Comparatively speaking, the AST/DOL conviction rate in 2004 was four times greater than is typical in the U.S. With such an extreme difference, it is no wonder that the prosecution and conviction rates shown in Figure 1 rank among or above the highest rates found in individual studies reviewed by Garner and Maxwell. While the DOL rate of prosecuted cases resulting in conviction was surpassed in a few jurisdictions, the AST/DOL rate of prosecutions per founded case and the AST/DOL rate of convictions per founded case were higher than what was reported in any of the American studies considered by Garner and Maxwell. Based upon these comparisons, it is fair to say that the case attrition rate of IPV assaults dealt with by AST and DOL in 2004 was among the lowest anywhere in the U.S.
Comparisons of Attrition by Victim Race and Location
To consider disparities in prosecution rates, we made comparisons of case attrition in terms of the victims' race and in terms of the geographic location where the IPV assault took place. Given the scrutiny placed on the State of Alaska's response to violence against Alaska Native women in rural villages (e.g., AITC v. Alaska, 110 P.3d 947, 2005 or Amnesty International's Maze of Injustice), these comparisons were made between the cases of Alaska Native and non-Native victims and between the cases that occurred in isolated villages versus cases that occurred in places that are less remote. These results have been used to argue that there is unequal enforcement by the State of Alaska in cases of IPV assault against Alaska Native women.
To test this assertion, we first examined the differences in case attrition of cases of Alaska Native victims compared to cases of non-Native victims. The first two rows of figures in Table 1 allow comparisons between the cases of female Alaska Native and non-Native IPV assault victims in terms of the proportion of all cases reported that were founded, that were referred for prosecution, that were accepted for prosecution, and that resulted in a conviction. With a lack of statistically significant differences between the two groups, it appears that Alaska Native victims' reports of IPV assault are just as likely as non-Native victims' reports of IPV assault to be founded and referred for prosecution by AST, to be accepted for prosecution by DOL, and to result in conviction in state court. In the third and fourth rows of Table 1 the two groups are compared in terms of the proportion of founded cases that resulted in referral for prosecution, acceptance for prosecution, and conviction. Although there was little difference between the referral and acceptance rates for founded cases reported by the two groups of victims, founded IPV assault cases against Alaska Native victims were actually more likely to result in a conviction relative to founded IPV assault cases against non-Native victims. The fifth and sixth rows of figures in Table 1 let us compare the rates of acceptance for prosecution and conviction for those cases that were referred by AST to DOL for prosecution. Once again, the difference in the acceptance rates of referred cases was not statistically significant whereas the conviction rates of the referred cases of Alaska Native victims was higher than the referred cases of non-Native victims. A similar result is found in the bottom two rows of Table 1 in which a greater proportion of cases of Alaska Native victims that were accepted for prosecution resulted in conviction when compared to the corresponding proportion for non-Native victims. Overall, the results of Table 1 indicate that Alaska Native and non-Native IPV assault victims' cases are processed at similar rates and that the only exception is that IPV assault cases are more likely to result in a conviction when the victim is Alaska Native.
To consider the effect of geographic isolation on case attrition, we made comparisons in terms of whether the IPV assault occurred in a village that was difficult for AST to reach to conduct an investigation. Villages were considered isolated if they did not have a local AST post or if they could not be reached by the Troopers from a post via automobile. The results of these comparisons are made in Table 2. In many ways, the results for an isolated/non-isolated comparison mirror the results presented in Table 1 regarding the differences in the processing of Alaska Native and non-Native victims' cases. First, the results indicate that isolation did not have an effect on the proportion of cases that were founded or that were referred for prosecution. In other words, it appears that the difficulty of reaching a village to make an investigation did not make it less likely that an IPV assault case would be carried forward by AST to DOL for prosecution. Next, there is some indication that DOL is more likely to accept IPV assault cases that occurred in isolated villages as compared to those cases that occurred in non-isolated locales; we see in Table 2 that there was a statistically significant difference in the proportion of cases referred by AST that were accepted for prosecution and that it was the cases from the isolated villages that were least likely to be screened out at that point. Finally, the results in Table 2 indicate that the conviction rate for cases that occurred in isolated villages was higher than for cases from non-isolated villages for all possible bases of calculation (i.e., convictions per reported case, convictions per founded case, convictions per case referred for prosecution, and convictions per case accepted for prosecution). Taken as a whole, it appears that the isolation of a village where an IPV assault occurs does not increase the likelihood of case attrition. If anything, these results indicate that IPV assault cases from isolated villages are not neglected by the State of Alaska but are instead more likely to be fully prosecuted.
Although Alaska's mandatory arrest law is the only policy specifically requiring official response to IPV cases, it appears that the law's spirit of full-enforcement guided other decisions regarding IPV cases as they continued through the criminal justice process in 2004. And while attrition still occurred in the IPV cases handled by AST and the DOL, it happened at a rate that was substantially less than what has been found outside of Alaska. By most standards, the fact that at any given decision point there was less than a one-in-five chance that a case would be screened out of the criminal justice process is an indication that IPV is taken seriously in rural Alaska and that AST and the DOL are willing and able to deal with the cases brought to their attention. Furthermore, this appears to be as true for Alaska Native victims as it is for non-Native victims and it does not seem to be affected by the difficulties of travel to conduct investigations.
Darryl Wood is an Assistant Professor at Washington State University Vancouver. André Rosay is the Director of the Justice Center. Marny Rivera is an Assistant Professor of the Justice Center. Katherine TePas is a Program Coordinator with the Alaska State Troopers. This project was supported by Grant No. 2005-WG-BX-0011 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.