Skip Navigation

Institution: CLOCKSS Sign In as Personal Subscriber

Brief Treatment and Crisis Intervention Advance Access originally published online on December 20, 2006
Brief Treatment and Crisis Intervention 2007 7(1):55-76; doi:10.1093/brief-treatment/mhl020
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
7/1/55    most recent
mhl020v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Buzawa, E.
Right arrow Articles by Byrne, J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Buzawa, E.
Right arrow Articles by Byrne, J.
Social Bookmarking
 Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2006. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org.

Understanding the Impact of Prior Abuse and Prior Victimization on the Decision to Forego Criminal Justice Assistance in Domestic Violence Incidents: A Life-Course Perspective

   Eve Buzawa, PhD
   Gerald T. Hotaling, PhD
   James Byrne, PhD

From the Department of Criminal Justice and Criminology, University of Massachusetts—Lowell

Contact author: James Byrne, Professor, and Eve Buzawa, Professor, Department of Criminal Justice and Criminology, University of Massachusetts, Lowell. E-mail: profbyrne{at}hotmail.com.

While there has been much attention focused on offenders, crime, and desistance from crime through the life course, researchers have largely ignored victimization and desistance from victimization through the life course. Focusing specifically on domestic violence among a cohort of offenders and victims in a single court setting (Quincy, Massachusetts), the authors explore the factors (including both offender and victim characteristics) related to revictimization and then examine the difference between those victims who do not report revictimization to the police and those victims who do report .The research reported here demonstrates that a childhood history of abuse may increase the probability that revictimization—if/when it occurs—will not be reported to police. The implications of this finding for criminal justice system's response to victims of domestic violence are discussed, focusing on (a) the development of new strategies for the identification and control of revictimization that are proactive and police- and/or probation-driven as well as (b) new approaches to treatment and crisis intervention that recognize the importance of understanding victimization—both at the individual and family level—throughout the life course.

KEY WORDS: abuse, victimization, life-course perspective, domestic violence, crisis intervention, formal and informal social control mechanisms, batterer treatment

This study includes the results of a secondary analysis of data originally collected for a recently completed National Institute of Justice (NIJ)-sponsored evaluation of a "model" domestic violence program located in Quincy, Massachusetts. We originally chose the Quincy, Massachusetts, site because we wanted to explore a setting in which a policy of "aggressive enforcement" was actually practiced. Our earlier research suggested that the police, district attorney's office, probation systems, and judges in Quincy have developed a truly integrated, system-wide strategy incorporating the "best practices" of full enforcement for a wide range of domestic violence incidents. However, Quincy's enforcement strategy relied on a crucial component: the decision by the victim of domestic violence to report subsequent victimizations to the police (or supervising probation officer). In our view, a system-wide response to domestic violence relying on the victim's decision to request criminal justice assistance is fatally flawed because it places the detection and control of domestic violence offenders with the victim rather than the criminal justice system and—ultimately—the community. In this article, we present research that supports this view and then we describe a new, proactive criminal justice response to the domestic violence problem that draws on three important advances in the field—life-course research, reentry partnership initiatives between police and community corrections, and the restorative justice movement in corrections.


    Background: The Quincy Court Domestic Violence Model
 TOP
 Background: The Quincy Court...
 Results
 Discussion: Implications for...
 Concluding Comments
 References
 
Beginning in l986, the Quincy District Court (QDC) initiated what has been described as one of the nation's first, and most comprehensive, proactive domestic violence programs. The court's aggressive, prointervention strategy has been recognized as a national model by the U.S. Violence Against Women Agency (VAWA) and the National Council of Juvenile and Family Court Judges. In recent years, the VAWA has designated the QDC as a national training site to be emulated by other jurisdictions searching for an (apparently) effective, integrated system-wide response to "domestic" incidents.

The purpose of the original project was to examine the impact of a rigorous intervention strategy upon a population of victims and perpetrators of domestic violence. That project, completed in February 1999, analyzed the actions of the police, prosecutors, and courts upon 353 domestic violence cases seen by the QDC in Massachusetts over a 7-month period. The QDC was chosen as the data collection site for this study because of its status as an acknowledged leader in implementing prointervention strategies in domestic violence cases, having been cited by the Department of Justice in the implementation of the Violence Against Women Act (VAWA) as a model jurisdiction.

One of the milestones of the original NIJ study was linking together eight separate data sets about the domestic violence incidents we used in that study. They included the original police incident report, prosecutorial charge data on each case, as well as initial and final case dispositions, data from batterer treatment programs on attendance and completion status offenders, data from the Registry of Civil Restraining Orders on the number and types of all prior civil orders taken out in Massachusetts against all defendants in our study, data from the Massachusetts Criminal Records System Board on all prior criminal charges accumulated by the defendants in our study, extensive interview data on 118 victims, and data from computerized files to track all 353 offenders for a 1-year period subsequent to the incident for any new criminal charges and civil restraining orders.

The original study included victim interviews containing considerable information that is not typically available, about what victims themselves have to say about the role of the police, prosecutors, victim advocates, and judges. We have always believed it critical to determine directly from victims their observations and experiences. The use of victim surveys in the study of domestic violence has certainly become more commonplace over the past 20 years. What is unusual is to have extensive information from official data sources and self-reports on the same individuals. This allows for a fuller understanding of certain issues concerning the reporting of revictimization. For example, some victims may not report revictimization because of their past treatment by the criminal justice system, because they preferred to use alternate responses, and because they are intimidated by dangerous males. From the perspective of domestic violence policy and practice, it is important to know not only what victims do when they are victimized but also why they do it. The analyses included here allow us to examine this issue directly.

The victim survey produced a revictimization rate substantially higher than that reported in official criminal justice data. Based on victim surveys, 49.2% (58 of 118) of respondents reported that they were either assaulted by the study offender, that the study offender had violated at least one condition of an existing restraining order, or they were forced to take out another restraining order on the study offender during the 1-year study period. Compared to the estimate of 22.1% reported from official data, estimates from victim accounts are 123% higher.

Therefore, this study will examine victim rereporting among those who reported a new incident 12 months post target incident. Our purpose is to determine if we could distinguish between those who recontacted the criminal justice system from those who did not report new incidents. Those reporting no new incidents (in our victim survey) are included as a control group to insure that variables distinguishing rereporters from nonreporters who reported a new incident are also distinguished between those with no new incident.

The QDC is a model court that aggressively processes domestic violence cases. Like most other jurisdictions, it does not use victim preferences as part of the decision-making process. However, empirical research rarely reports on the consequences of this practice on the future reporting and nonreporting behavior of domestic violence victims. This is an important issue to understand for several reasons. First, because victim reporting is needed to jump-start the "proactive" criminal justice response, it can be argued that at its core, even the most aggressive criminal justice strategy is essentially a reactive response, ultimately dependent on a victim who is willing to report abusive behavior to the police. When viewed in this context, it is not surprising that the lack of victim reporting has often been a source of frustration for many criminal justice practitioners. Second, an important question for public policy is whether victim reporting or nonreporting behavior is based on rational or irrational factors. If, for example, a victim reports an offense because of her fear of an offender, this suggests a rational choice. However, if a victim chooses not to contact the police because of her (irrational) fear of retaliation by the offender, this suggests a problem that necessitates a different intervention. Third, failing to address the reasons behind victim's failure to report may have serious and negative long-term consequences. If the overall societal goal is to stop domestic violence and protect victims, failure to report revictimization becomes a systemic failure.

The Failure to Report Revictimization in the QDC
From the initial QDC research, we knew that many women who reported a revictimization by the study offender did not report the criminal act to the police or criminal justice agents. Of the 58 victims who reported revictimization in this study, 26 reported the offense to the police and 32 did not. An obvious question was why over half of the women who reexperienced a domestic violence episode within 1 year of the original incident decided against reinvolvement with the criminal justice system. There are several possible explanations. If one assumes that the nonreporting of domestic violence incidents is similar to the nonreporting of crime in general, then there are at least five separate reasons for nonreporting. First, many victims might simply not want intervention by the criminal justice system or prefer some other form of assistance. A large number of victims did not initiate the previous call leading to criminal justice intervention. Nationally, research has found the range of victim-initiated calls ranging from one third to two thirds (Buzawa & Buzawa, 2002; Felson, Messner, & Deane, 2003). These victims may not have wanted any intervention or, alternatively, may have wanted assistance (and actually would have or did) to engage in alternative help-seeking behavior.

Second, it may be that victims may not see the particular reoffense as serious enough to warrant criminal justice system intervention. For example, victims might be more likely to report subsequent assaults, but less likely to report restraining order violations. Alternatively, victims may believe an injury is needed during an assault in order to justify police involvement.

Third, victims might not report subsequent victimizations due to fear of retaliation by the offender. This hypothesis would gain support if it was found that women who were revictimized by offenders with extensive criminal histories or about whom the victim's fear has increased over time were those who failed to report.

Fourth, victims might be skeptical about the effectiveness of criminal justice involvement for their situation. Like other crime victims who do not report law violations, victims in this study might have felt that because they had already been involved with the criminal justice system, it was unsuccessful at providing a remedy for the situation. Some victims might have felt that criminal justice intervention previously had or might now exacerbate an already bad situation.

Fifth, the criminal justice system might not have followed victim preferences and, therefore, this group of victims might have sought alternative sources of help or simply not seek any further assistance.

Although there are certainly a number of possible explanations for the nonreporting of revictimization, a significant number of women do report subsequent incidents of victimization to the police. In these cases, a victim's need for immediate assistance and the seriousness of the situation may subsume any other considerations in the mind of the victim. She may prefer alternate sources of help and/or fear the consequences of intervention but call the police because she needs their assistance to ensure her safety and/or the safety of her children. The research question is obvious: Can the differential response to revictimization be examined in cases of domestic violence?

To date, empirical research has not fully examined this basic research question. Although there is some research on what victims say they will do, this does not necessarily mean that victims will act accordingly. Further, this research does not control for the actual case processing and outcome of the case in terms of revictimization and reoffending. Victim reactions may be partially based on her prior experiences with criminal justice intervention during her last assault victimization and its subsequent consequences. This research will examine and control for the impact of incident, victim, offender, and case processing characteristics on a victim's decision to report new incidents of abuse.

Factors Related to the Nonreporting of Revictimizations: A Review of the Evidence
It has long been known that a relatively high percentage of victims of domestic violence never report such incidents to police (Buzawa & Buzawa, 2003; Felson, Messner, Hoskin, & Deane, 2002; Felson & Paré, 2005; National Research Council, 2004). In the past, many victims realistically feared that police would simply fail to appropriately act and, at best, respond perfunctorily or even exacerbate the situation by demonstrating that the offender's conduct was tacitly condoned when no police action was taken. (Berk, Berk, Newton, & Loseke, 1984; Ferraro, 1989; Radford, 1989). Others have cited the concept of "relational distance" wherein victims and offenders bound by close relational ties rarely reported incidents to police (Black, 1980).

This dynamic, however, may be changing in a manner that makes victim calls to the police even less predictable than in the past. As is well known, the police response to reported domestic violence has undergone major transformation in many, if not most, jurisdictions (Buzawa & Buzawa, 2003; Felson & Paré, 2005). Today, it simply cannot be said that police en masse trivialize domestic violence. At times, the critique of police action is that it is too uniform, for example, that arrests are made regardless of whether most would consider it appropriate for the situation due to overly rigid policies or even statutes mandating this action. As a result, some have surmised that a number of victims may be prospectively deterred from calling the police due to fear of further losing control of the situation (Buzawa, Hotaling, Klein, & Byrne, 1999; Mills, 1998, 1999, 2003).

Reporting further abuse or revictimization to the police becomes an extremely interesting subset of the interaction between police and victims. In such cases, by definition, the victims have already encountered the actions of the police and, at least in this jurisdiction, those of the district attorney and the courts. For this reason, our analysis of victim rereporting behavior either reinforces or undermines an illustration of the impact of an aggressive, system-wide intervention response.

A number of studies have looked at the impact of mandatory arrest on subsequent offender behavior (see, e.g., Berk, Campbell, Klap, & Western, 1992; Dunford, Huizinga, & Elliott, 1989; Garner, Fagan, & Maxwell, 1995; Garner & Maxwell, 1999; Hirschel & Hutchison, 1992; Maxwell, Garner, & Fagan, 2001; Pate & Hamilton, 1992; Sherman et al., 1992). As an aggregate, it appears that mandatory arrest has not had the desired effect on either offenders or victims. In terms of offenders, it was anticipated that the use of mandatory arrest strategies would act as a deterrent to subsequent violence by offenders. This has not proven to be the case. In terms of victims, it has long been assumed that the initiation of aggressive intervention strategies such as mandatory arrest would result in increased victim participation in the criminal justice process. Specifically, victims would be empowered through intervention and therefore be more likely to report subsequent problems with the offender. One criticism of the existing research on the effectiveness of mandatory arrest policies is that researchers have not addressed the factors that distinguish victims reporting revictimization from those victims who do not. In addition, previous research has failed to disentangle the effects of other criminal justice components, for example, prosecution, sentencing, and corrections, on the subsequent behavior of victims.

Sample
Data used in this report are based upon domestic violence cases that resulted in arrest and arraignment before the QDC during a 7-month study period. All consecutive arrests for domestic violence involving male defendants and female victims that occurred between June 1995 and February 1996, were initially examined for inclusion in our final sample. From that pool, we eliminated all cases involving defendants and primary victims who were under the age of 17, cases involving same-sex relationships, and cases involving male victims and female defendants. The final sample was composed of 353 cases of male-to-female domestic violence. It can also be described as a population because it included every case in this category of incidents.

Even so, the representativeness of this sample of all male-to-female domestic violence cases cannot be fully determined. First, although we have reason to believe most cases resulted in arrest, little is known about cases that may not have resulted in arrest. Second, even though cases in this sample show little variation in numbers or on key characteristics on a month-to-month basis, there may still be seasonal variations in the nature of cases. Third, and perhaps most importantly, our sample size does vary from analysis to analysis due to the availability of data from the primary sources used in this study, that is, official records and self-report surveys. Consequently, we are often reporting results from a subsample of offenders and victims, raising questions about the generalizability of the study findings.

An additional concern was the generalizability of the 118 victims who were surveyed. Because we only interviewed 118 of 353 study victims, there is a distinct possibility of a skewed sample. However, we compared the characteristics of our interview sample to the total population of victims included in the original study and found no statistically significant differences between these two groups (see Buzawa et al., 1999). In addition, we found that the sample rate for the entire 353 victims in our study was the same as for our 118 respondents using only criminal justice data—22.0%. In other words, only 26 out of the 58 victims who reported an instance of revictimization to us reported the revictimization to the police or another agent of the criminal justice system. Thus, our estimate of revictimization of surveyed victims would have been 22.0% had we relied exclusively on official data, which is exactly the same reoffending rate as for the complete sample of 353 offenders.

In addition, such a differential between responding victims and those not available is, in theory, unlikely. Many victims who could be located left the community because they could not successfully escape abuse (or feared retaliation). We would not expect these victims to have lower rates of reabuse than those who stayed, apart from the fact that they were successful in flight and therefore may be less likely to be found by their abuser.

This finding adds to our confidence that these estimates are reliable. A revictimization rate based upon criminal justice reporting indicates that one in every four or five victims is revictimized in the year following coming to the attention of the criminal justice system. An estimate based on the accounts of the same individuals unmediated by reporting behavior would put the revictimization estimate at one of every two victims.

Study Design
As stated, the purpose of this research is to better understand victim revictimization and why it is or is not reported to the police. To facilitate this design, information was needed from multiple sources and perspectives covering data from significant periods of time both before and after the occurrence of the incident that led to its inclusion in our sample. In addition to procuring these data, an additional challenge was to link together information from several sources into one coherent data file. Sources of data used in this study are first described below. They include offender criminal history data, records of civil restraining orders, probation department data on prosecutorial charges, case disposition and risk assessment, data on offender treatment program participation, police incident reports, and self-report victim survey data.

Data Sources
Offender's Criminal History Data.
The QDC Department of Probation provided criminal biographies for all 353 defendants in the sample. For this research, each defendant's criminal activity was analyzed both prior to the study incident and for 1 year subsequent to that incident.i

Civil Restraining Order Data.
In September 1992, the State of Massachusetts implemented the Registry of Civil Restraining Orders: the first statewide, centrally computerized record keeping system on restraining orders. This registry is primarily designed to provide the police and courts with accurate and up-to-date information on the existence of active orders. The QDC Department of Probation provided information from this registry on the number and type of civil restraining orders taken out in Massachusetts against all 353 defendants both before the occurrence of the study incident and for a 1-year period following the study incident.ii

Prosecutor's Office/DCD.
The QDC Department of Probation also provided us with information on all 353 defendants concerning prosecutorial charges. For each defendant in our study, information was provided on up to three domestic violence-related charges for our study incidents and any additional non-domestic violence-related charges. This information enabled us to compare police charges to prosecutor charges on their number, severity, and type and to understand the link between prosecution charges and court handling of cases.

Data from the QDC on initial and final dispositions (and their dates) enabled a determination of the amount of elapsed time between arraignments and dispositions as well as the number of defendants who violated the conditions of their initial dispositions.

Data on Study Defendants and Batterer Treatment Programs.
Many study defendants had to enroll in a batterer treatment program as a condition of probation. We contacted the Directors of the two batterer treatment programs that serve the QDC and received data on offenders' treatment completion status at the end of our study period.

Police Incident Reports.
A key data source used in this study was the police reports for the study incidents from the seven departments served by the QDC. These reports were used to measure the officer's perspective and actions taken about the incident, what the call for service involved, characteristics of the incident, sociodemographics of the participants, and their narrative description of the incidents and their stated response.

The Victim Survey.
In addition to official criminal justice system data concerning our study incidents, we needed to capture the perspective of the victims on the study incidents and their handling.

The interviews had three primary goals: (a) to obtain the victim's point of view about what she wanted from the criminal justice system and how the criminal justice system responded to the domestic violence incident in which she was involved, (b) to get details about the study incidents and the context of the victim–offender relationship that are not typically available in official statistics, and (c) to hear directly from victims about the defendant's reoffending behavior.

Because one of the chief aims of the survey was to tap into the victim's perspective about experiences with the criminal justice system, victim interviews did not take place until approximately 12 months after the occurrence of the study incident. Our use of a 1-year time frame was dictated to us by the fact that we had to wait until victims passed through contact with the prosecutor's office and court and our interest in self-reports about reoffending behavior 1 year after the study incident.

This delay clearly had a severe effect on response rates, and we were able to complete usable surveys with 118 victims in this study, 35% of eligible study respondents. However, a more important question is the extent to which those who completed the survey are different from both "refusals" and those women we were unable to locate. On the basis of official record information (police incident reports and criminal history information), we compared those who completed the survey to refusals and to those we could not locate on the basis of victim, offender, and study incident characteristics. For most comparisons, there were no major differences between victims according to their status on our survey. We were originally concerned that those victims we did not interview were involved with more dangerous men or in more serious domestic violence incidents. This concern was not borne out. Those who completed the survey were, in fact, more likely to have been in incidents involving severe violence and the use or threat of guns and knives and were abused by men whose criminal histories were as extensive as offenders whose victims who did not.

We sought direct data from victims as a check on the accuracy of "official data." Respondents were asked about events that occurred at different points over the past year (i.e., police involvement, talking to a victim advocate, going to the prosecutor's office, going to court). The most distant event in that time span was the incident that led to their inclusion in the sample. If memory problems did affect the quality of the information gathered, we would expect that this problem would be most apparent for that event. There was a very high level of agreement between victims and the police on a number of details concerning the study incident. Victims accurately recalled specific details about the incident in terms of participants, location, dynamics of the incident, and police actions.

Statistical Procedures
For all dependent variables, bivariate analysis (chi-square, ANOVA, odds ratios) was used to examine variations by incident, victim, offender, and case processing characteristics. In addition, we have included the results of our multivariate analysis of the characteristics that distinguish the three groups at the outset of this review: Group 1 (victims with no incident during the 1-year follow-up period), Group 2 (victims who did report a new incident during the follow-up period), and Group 3 (victims who stated there was a new incident but they did not report it to the police). Finally, the analyses utilized logistic regression techniques to identify the factors that distinguish victims of new incidents who report these incidents to police and victims who decided not to report.


    Results
 TOP
 Background: The Quincy Court...
 Results
 Discussion: Implications for...
 Concluding Comments
 References
 
Types of New Offenses Reported
Over half of the victims in this sample who disclosed that they were revictimized during the 1-year period following the target incident did not report the new incident to the criminal justice system. As can be seen in Table 1, only 42.5% of victims contacted the criminal justice system again about the defendant's abusive behavior. Contrary to the assumption that more serious offenses get reported to the police, while less serious incidents do not, victims in this sample were more likely to rereport violations of restraining orders and recontact the court for new restraining orders than report actual physical assaults. Although the difference in rereporting of restraining order violations and assaults is not statistically significant ({chi}2 = 3.47; df = 1; p = .06), it should be noted that half of restraining order violations were reported (50.0%) compared to only 23.5% of physical assaults. (Note: Given the small sample size for these two groups, it is more appropriate to discuss the percentage differences than to focus on statistical significance).


View this table:
[in this window]
[in a new window]

 
TABLE 1. Percentage of Cases in Which a New Incident of Abusea Was Reported in Victim Interview During 1-Year Study Period and Percentage of Those Cases Reported to the Criminal Justice System and Involving Either a Restraining Order Violation or a New Physical Assault (N = 119)

 
Of course, we are assuming that physical assaults are, in fact, more serious incidents than violations of protective orders. But this may not be the case from a victim's perspective for whom either incident can be frightening and upsetting. Additionally, from the point of view of the criminal justice system, at least in the jurisdiction from which the sample was drawn, violations of court orders are handled as more serious incidents. In Massachusetts, domestic violence statutes mandate the arrest of an offender who violates one or more of the conditions of a protective order but use arrest as a preferred response to allegations of threats and violence. Victims are typically aware of this distinction because they are informed of what will happen if a protective order is violated and several victims told us that, in practice, the criminal justice system took violations of court orders more seriously. As one victim explained why she reported a restraining order violation to the police, " I figured the police and court would be more likely to go after him if he violated a court order."

The greater tendency to recontact the criminal justice system for violations of protective orders may also reflect the fact that women who were no longer in a relationship with the offender were much more likely to have taken out a restraining order following the target incident (90% vs. 10%) and to report those violations than women still in a relationship with the offender (57.1% vs. 33.3%).

Still, some victims did report new violations, whether new assaults or of conditions of restraining orders, and others did not. Data in Tables 2 through 7 attempt to explore on the bivariate level some of the factors associated with nonreporting.


View this table:
[in this window]
[in a new window]

 
TABLE 2. Criminal Justice System Actions Taken in Target Incident and Victim Appraisal of Actions by Rereporting Status (N = 119)

 
Prior Criminal Justice Action, Victim Appraisal of Criminal Justice Action, and Nonreporting of New Incidents
It has been hypothesized earlier that victims may decide not to report new offenses to the criminal justice based on their experiences with that system (see, e.g., Buzawa et al., 1999; Fleury, 2002). There are at least three ways in which prior contact with the criminal justice system might discourage a person from turning to it again for assistance. First, it may be due to the failure of that system to take action against the offender in prior domestic violence incidents. For example, women who do not report new offenses may have been treated differently than others in terms of how their cases were handled or how the offender was dealt with. Second, some victims may have found the criminal justice process too burdensome to return. That is, this form of help may have been found to be confusing in terms of the legal language used and steps to be taken and too time consuming. Third, some victims may have found the system to be arbitrary and out of step with their wishes and preferences in terms of case handling. Data in Tables 2 and 3 address each of these issues.


View this table:
[in this window]
[in a new window]

 
TABLE 3. Victim Preferences About and Evaluation of Criminal Justice Response in Target Incident, by Rereporting Status

 
Data in Table 2 show that actions taken by the criminal justice system in the target incident failed to differentiate between the three victim groups. Offenders across the groups were roughly equally likely to have been charged with a serious crime, to have been successfully prosecuted, and to have been placed under criminal justice surveillance. In fact, the offenders of victims who did not report a new incident to the criminal justice system were somewhat more likely to have been mandated to attend batterer treatment and to have been incarcerated.

It also does not appear that victims who did not report new offenses were concerned about the time-consuming or confusing nature of the criminal justice system response. Data in Table 3 show that victims in Group 3 were actually the least likely to have found the system troublesome in this sense. In fact, those victims who did report new offenses were the most likely to depict the criminal justice response as confusing and time consuming, but obviously this did not discourage them from rereporting.

Data in Table 3 does show, however, that victims who did not report new offenses (Group 3) were significantly more likely than victims in the other groups (Group 2) to depict the criminal justice system as unresponsive to their preferences. Over 55% of victims who did not report new offenses during the 1-year study period said they had no voice or rights with criminal justice officials compared to only 12% of victims who did rereport new offenses (Group 2) and 18.3% of victims who reported no subsequent revictimizations (Group 1). Group 3 (failure to report) victims also differed significantly from Group 2 victims in overall levels of dissatisfaction with the criminal justice system (25% vs. 7.4%) and in feeling that a "more therapeutic approach should have been taken toward offenders" (58.8% vs. 36.0%).

Group 3 (failure to report) victim appraisals of their contact with various components of the criminal justice system in the target incident were diverse and did not follow a consistent pattern. For example, victims who did not report new offenses were the least likely group of victims to resist the arrest of the offender in the target incident and the least likely to be dissatisfied with the police handling of the incident. But at the same time, they were the most likely to feel that the actions of the police negatively affected their safety. They were less likely than victims in other groups to have wanted the offender prosecuted but most likely to report that they wanted the prosecutor to make charges against the offender more severe and just as likely as others to have wanted charges lowered or dropped. Additionally, Group 3 victims did not differ from the other groups on the extent to which the threat of prosecution angered or frightened the offender.

Overall, Group 3 (failure to report) victims were dissatisfied with the actions taken by the prosecutor (compared to Group 2) and felt the prosecutor's actions decreased their safety (compared to Group 1 victims). Victims across all three groups did not significantly differ from one another in their appraisal of the impact of court actions on their safety or on their level of satisfaction with the court.

Data in Table 3 implies that regardless of the specific preferences Group 3 (failure to report) victims may have had about actions taken by criminal justice personnel, the perception that their preferences were ignored most distinguished Group 3 victims from victims in Groups 1 and 2. The perception of having no voice or no rights may have been most acute in their dealings with the prosecutor's office with whom nonreporters were most dissatisfied.

Offender Dangerousness and Nonreporting of New Incidents
An alternate hypothesis to account for the decision to not report new incidents of abuse to the criminal justice system may have to do with the fear of reprisal from dangerous offenders. There are at least two ways to assess offender dangerousness in this regard. Fist, through the seriousness and danger posed to the victim in the target incident itself and, second, through the past criminal histories of offenders.

Data in Tables 4 and 5 examine these facets of offender dangerousness. Table 4 displays characteristics of the target incident by reporting status. As would be expected from a court sample, these are serious domestic abuse incidents, especially for victims who disclosed subsequent incidents to us. However, there are a number of important differences across groups and most are between those victims who experienced a new incident and those who did not (Group 1 vs. Groups 2 and 3). There are very few factors that distinguish those who reported a new incident from those who did not. One exception is the escalation in the pattern of violence exhibited by the offender. Group 2 (new incident reported) victims were more likely than victims in the other groups to report that the offender's abuse was becoming more serious prior to the target incident (72% Group 2 vs. 58% Group 3 and 45% Group 1). We suspect that this may be an important factor in accounting for why this group reported new incidents to the police.


View this table:
[in this window]
[in a new window]

 
TABLE 4. Target Incident Characteristics, by Rereporting Status

 

View this table:
[in this window]
[in a new window]

 
TABLE 5. Offender Criminal Histories, by Rereporting Status

 
It appears that the dangerousness of the offender in terms of the severity of violence used and the fear that is generated by his actions is a common element across the sample, especially among victims who experience new incidents of abuse. However, an escalation in offender dangerousness does not appear to discourage the reporting of new incidents but rather may be a factor promoting the reporting of new incidents to the authorities.

Data in Table 5 also addresses offender dangerousness by examining a number of indicators of prior criminal involvement. Overall, this sample of offenders was very criminally active, with 81% having at least one prior criminal charge of any kind and 54.6% having at least one prior criminal charge involving a crime against a person. Even though the sample as a whole was quite criminally active, there were some differences across groups. Offenders of Group 3 (failure to report) victims had a significantly greater average number of restraining orders taken out against them by others (0.88) than Group 1 victims (0.25), but not a significantly greater number than those in Group 2 (0.52). Offenders of women who were revictimized and reported the new incident (Group 2) had a greater number of prior crimes of violence (4.80) as compared to Group 1 offenders (1.80) but not significantly different from offenders of women who did not report a new incident (3.47).

Only one variable differentiated between all groups and that was the proportion of offenders with at least one prior controlled substance abuse criminal charge. Almost a quarter of offenders of women who did not report new incidents (Group 3) had prior drug charges compared to less than 7% of Group 1 (no new incident) and only 4% of Group 2 (new incident) offenders.

Victim Characteristics and the Nonreporting of New Incidents
It is certainly possible that women forego additional criminal justice assistance because of their past life experiences or victimizations. Women who are still with their offenders may not want to jeopardize their relationship and so not call the police when new incidents occur. It has been hypothesized that a subgroup of victims may forego additional criminal justice involvement because of their current life situation or past victimization experiences (see Buzawa et al., 1999). Variations in reporting behavior may also be due to the type and extent of past victimization experiences. For example, women with multiple victimizations may grow increasingly skeptical about the system's effectiveness in protecting them and forego additional assistance.

Data in Table 6 examines group differences by demographic characteristics of victims and their assailants. There are no group differences by the race or age of victim or offender, educational attainment, or household income. There were also no differences across groups in the percentage of women who were living with the offender at the time of the incident.


View this table:
[in this window]
[in a new window]

 
TABLE 6. Demographic Characteristics at Time of Target Incident, by New Incident Reporting Status

 
Groups varied by employment status (Group 1 56.7%, Group 2 80%, Group 3 79.4%) and whether women were still with offenders (Group 1 36.7%, Group 2 92%, Group 3 81.8%. However, these factors differentiated those who experienced a new incident from those who did not and nonreporting behavior. Women who reported a new incident of abuse were more likely to have been in the labor force and were not in a relationship with the offender at the time of the new incident, whether they reported the new incident or not. Of course, it is certainly possible that the reason women fail to disclose a new incident to us is directly related to their ongoing relationship with the offender.

Table 7 shows an extensive pattern of past victimizations. For many women in this sample, especially those who were revictimized, revictimized women (Groups 2 and 3) were significantly more likely than Group 1 victims to have taken out three or more restraining orders against the study offender and to have made multiple prior calls to police about his abusiveness. Equivalent proportions of each group reported multiple adult violent relationships and severe parental violence while growing up.


View this table:
[in this window]
[in a new window]

 
TABLE 7. Patterns of Violence in Victim's Adult Relationships, by Rereporting Status

 
Women in all groups reported a high involvement of child sexual abuse. Over 38% of all women in the sample reported sexual assault, a rate almost twice as high as the 20% to 27% estimates derived from general populations of women (Finkelhor, 1994, 2000). Almost 56% of women who did not report a new incident of domestic violence reported child sexual abuse, a proportion significantly higher than found for any woman who reported new incidents to the police (28.0%) or for the group of women reporting no new incidents (32.8%). It is certainly possible that there is a link between the abuse history of domestic violence victims and their likelihood of reporting revictimization to police. For an individual who has experienced abuse throughout the "life course," reporting this latest incident to the police may be viewed as useless ritualism; for these victims, revictimization may be viewed as an inevitable and unavoidable continuation of a process of ongoing victimization throughout their life course.

Predictors of Nonreporting
An attempt to separate out the indirect effects of important factors in cases of nonreporting is shown in Table 8. A logistic regression comparing women who experienced a new incident and either reported it or not to the criminal justice system includes the predictors that differentiated groups at the bivariate level.iii


View this table:
[in this window]
[in a new window]

 
TABLE 8. Logistic Regression of Reporting versus Nonreporting of New Incidents of Abuse to Criminal Justice System (N = 59)

 
The model can be interpreted to point to the importance of indicators of frustration with the criminal justice system. The offenders of women who do not report are no more violent but have had drug problems and a history of harassment of the victim. It is certainly possible that the offenders represent a subgroup of "multiple-problem" domestic violence offenders who may require a different criminal justice intervention strategy.

In addition, a victim's past history of victimization also influences her decision to report, with sexual abuse prior to the age of 18 significantly influencing the decision to report. In addition, the significance of the offender's prior criminal history and prior restraining orders may also contribute to a victim's perception of the inability of the criminal justice system to effectively intervene and therefore discourage her from further efforts to involve the criminal justice system. At some point, it appears that for some victims, domestic violence becomes an "unshareable" problem.


    Discussion: Implications for Brief Treatment and Crisis Intervention
 TOP
 Background: The Quincy Court...
 Results
 Discussion: Implications for...
 Concluding Comments
 References
 
This research presented here suggests that a latent outcome of aggressive law enforcement and court response that includes the dismissal of victim preferences may be to discourage the future utilization of the system by both victims who wanted the system to do more (those who wanted more severe criminal charges brought against the offender) as well as those who wanted it to do less (those who felt taking the case forward would decrease their safety). Our study focused not only what happened to the victims of domestic violence in Quincy, Massachusetts (new incident/no new incident during a 1-year follow-up period), but also why victims responded to revictimization as they did (report to police, not report to police).

To begin, it is apparent from our victim interviews that official records of revictimization significantly underreport this behavior. Based on official records, only 22.1% of victims in our sample were revictimized. By comparison, the results of our victim interviews indicated that over 49.2% were revictimized. If prevention of revictimization is the goal of policy makers, new strategies must be developed simultaneously in two areas: (a) new methods need to be utilized to persuade victims to report subsequent victimizations to police and (b) new strategies need to be developed which make the identification of subsequent revictimization a community concern rather than a problem of victim notification. In terms of this latter point, it certainly seems possible that the strategies employed vis-a-vis the federal government's recent reentry partnership initiatives (see, e.g., Taxman, Young, & Byrne, 2002) could be applied directly to the problem of domestic violence. Specifically, one avenue to explore is the utilization of local community police agents, in conjunction with the courts and corrections system, as part of a proactive system-wide revictimization detection strategy. In this new system-wide strategy, the police will respond proactively to the problem of potential revictimization by identifying offenders of domestic violence, talking directly with offenders about the implications of reoffending on their liberty and lifestyle prior to sentencing, and then directly monitoring offender behavior toward the victim in particular and the community in general. By redefining domestic violence as a community problem, the new domestic violence partnership would effectively take the primary responsibility to detect and report revictimization out of the hands of the victim, thereby reducing the potential control of the offender over the victim.

Given the findings reported here about the prior substance abuse and/or criminal histories of these offenders (including the serial domestic abuser subgroup who appear to continue their abuse with new partners), it certainly makes sense to view this group of offenders as a threat to the community. We envision a control strategy that would begin at sentencing, where specific conditions of community release (e.g., curfews, participation in drug/alcohol treatment and/or mental health counseling, and restrictions on movements within the community and on contact with the victim) would be established, perhaps in the form of the type of individual offender responsibility plan currently used in Vermont with prisoners about to reenter the community (see, e.g., Byrne, Taxman, & Young, 2002). Compliance with individual offender responsibility plans would be monitored by members of the partnership working together in the community. We anticipate that local community police officers, in conjunction with community corrections, would conduct joint home visits and curfew checks, regularly interview both the offender and victim, and check for potential violations in the area of substance abuse and/or participation in treatment.

Judicial involvement and support will be a critical element of this type of system-wide domestic violence control strategy because judges will be expected to send a clear message to offenders (at revocation) that violations of the conditions of their community release will not be tolerated. If we are serious about developing a proactive strategy to address the problem of domestic violence, we then need to find a way to make the criminal justice "community" and not just the individual "victim" the lynchpin in the detection/reporting process.

Finally, our research suggests the significance of indicators of victim frustration with the criminal justice system. Revictimized victims were far more likely to have taken out prior restraining orders and contacted the police about the offender in the past. It is certainly understandable that many of these victims become discouraged with the efficacy of current intervention strategies and decided not to report subsequent victimizations to the police.

It is possible that victim frustration is not just related to the criminal justice system's response to this incident but to a series of prior victimizations. Revictimized women who failed to report their revictimization to the police were far more likely to have endured childhood sexual abuse and have been subjected to serial victimizations throughout their life. One avenue of research that has been particularly fruitful over the past decade is the discussion and an exploration of crimes through a life course. Equally compelling is the notion of victims through a life course. Although we have given considerable attention to offender careers in crime, less attention has been spent on studying victimization careers, not only in terms of individual victimization but also in terms of family victimization through the life course. It seems difficult to expect victims who have often endured a lifetime of abuse to enter into a criminal justice process that, up to this point at least, has failed to help them and/or their family.

It is within this broader context of victimization experiences through the life course that the new innovations in specialized domestic violence courts come into play (Center for Court Innovation, 2006). A number of states have developed specialized domestic violence courts in an attempt to coordinate the justice system's response to individual and family victimization, which all too often was focused on a court-specific response to specific, static, incidents or problems (court orders restricting contact by offenders, child custody/visitation/support, and criminal court processing of initial and subsequent victimization complaints). New York's misdemeanor and felony domestic violence courts offer an excellent recent example of this type of coordinated, system-wide initiative:

IDV Courts respond to a historic problem in the court system, which requires domestic violence victims and their families to appear in different courts before multiple judges, often located in different parts of a county, to address their legal issues. Dedicated to the "one family–one judge" model, IDV Courts allow a single judge to hear multiple case types—criminal, family and matrimonial—which relate to one family where the underlying issue is domestic violence.

IDV Courts aim to provide more informed judicial decision-making and greater consistency in court orders, while reducing the number of court appearances. In addition, these courts offer enhanced services to victims and ensure offender accountability.

Since their inception, IDV Courts have handled over 33,000 cases and served over 7,000 families. Currently, there are 28 IDV Courts in operation and 9 IDV Courts in planning*. Currently more than three quarters of the residents of New York State live in counties served by these courts (Kluger, 2006: http://www.nycourts.gov/ip/domesticviolence/index.shtml).

We certainly view the development of such specialized domestic violence courts—in New York State and at several other jurisdictions across the country—as an important first step toward developing a new approach to domestic violence that recognizes the need for a coordinated, system-wide response to the problem of victimization through the life course. However, recent evaluations of the effectiveness of batterer programs generally (and judicial monitoring in particular) highlight a simple fact: "coordination may improve efficiency but it does not guarantee effectiveness." According to a recent randomized field experiment on the effectiveness of the Bronx Misdemeanor Domestic Violence Court conducted by the Center for Court Innovation (Labriola, Rempel, & Davis, 2005), "the preponderance of evidence now accumulated in the field calls into question the efficacy of batterer programs based on the most prevalent national models. Indeed, the main findings from our randomized trial are consistent with other recent trials, of which none found that mandating offenders to a batter program for groups for men produced lower rates of re-abuse ... Contrary to our expectations, we did not find that judicial monitoring leads to lower re-arrest rates; nor did we find that the specific monitoring schedule (monthly or graduated) affects the recidivism outcomes ... Regrettably, our study suggests that some of the most prevalent responses to domestic violence may be ineffective; but perhaps these findings can liberate the justice system to innovate in as-yet unexplored ways and promote a new period of reflection and experimentation" (p. vii–ix).

One recommendation included in the Center for Court Innovation study was that program developers need to refocus their attention on the dual tasks of (a) developing and evaluating new strategies for offender monitoring/compliance while simultaneously (b) addressing the unique treatment needs of the victims of domestic violence (Labriola et al., 2005). On this latter recommendation, Roberts and Everly (2006, p. 1) present evidence from a recently completed meta-analysis of 36 crisis intervention studies that bear careful consideration here. According to the authors, "a number of crisis intervention studies focus on the extent to which psychiatric morbidity (e.g. depressive disorders, suicide ideation, and posttraumatic stress disorder) was reduced as a result of individual or group crisis interventions or multi-component critical incident stress management" (Roberts and Everly, 2006, p. 1). Although the research base in this area—like many areas of criminal justice—is noticeably weak, the results of the review of Roberts and Everly were promising and suggest that in-home, intensive, family crisis intervention programs (with at least 8 hours of in-home "treatment" over a 3-month period) may offer an effective alternative to traditional individual victim counseling programs now operating in domestic violence courts (see also Roberts & Roberts, 2005).


    Concluding Comments
 TOP
 Background: The Quincy Court...
 Results
 Discussion: Implications for...
 Concluding Comments
 References
 
Although we certainly need to develop and test new domestic violence offender monitoring and control strategies, we also need to recognize the inherent limitations of formal social control strategies (Byrne & Taxman, 2006). Clearly, victims with lifelong histories of victimization are at increased risk for future victimization; there is little we can do within the traditional criminal justice system to alter this life course trajectory of victimization. As we have highlighted in this review, research on the effectiveness of current domestic violence prevention and control strategies—including specialized domestic violence courts—points to a depressing reality: the criminal justice system can only provide limited protection and assistance for victims of domestic violence (Labriola et al., 2005).

Many, if not most jurisdictions, now coordinate victim services with police, the prosecutor's office, and the courts. However, these services are typically limited to addressing the victim's immediate needs for safety and needed resources, at the time of the call for service. Given what we now know about repeat victimization throughout the life course, program developers are beginning to recognize that many victims have needs that transcend the types of traditional individual counseling services currently available. Whether it is family crisis intervention, a muticomponent critical incident stress management program, or some other program targeting victimization through the life course, it is apparent that new approaches to the treatment of domestic violence victims need to be implemented and rigorously evaluated (Roberts and Everly, 2006). Although there is a great need for a personalized victim counseling that addresses the diverse range of reasons that places some victims at greater risk than others, it appears that we also need to incorporate the notion that individual domestic violence incidents may necessitate family crisis intervention.

In recent years, there has been a reluctance to provide psychological counseling for victims out of fear that it would be construed as "victim blaming." However, it could instead be part of a broader framework that examines victim entrapment and includes psychological needs along with safety needs, resource issues, and child welfare. Evan Stark, in his forthcoming book, Coercive Control: The Entrapment of Women in Personal Life (Oxford University Press), eloquently discusses the need to improve autonomy in the life of abused women and addresses the need to empower women to attain personal freedom in their lives. The need to individualize and personalize victims whose identities have already been challenged is critical, but obviously we need to first understand the unique life course experiences that lead toward—and hopefully way from—various forms of victimization.

In sum, we must now recognize that (a) our efforts to improve existing domestic violence interventions using formal, criminal justice system-based control strategies have not proven effective and that (b) we need to develop and test new approaches to the seemingly intractable problem of domestic violence that recognize the importance of informal social control mechanisms and the need for individual and family-based treatment interventions. Paradoxically, we appear to know much more about the characteristics of domestic violence offenders (Roberts, in press) than we know about the unique problems and needs of domestic violence victims. By studying victimization through the life course, we can begin to develop interventions designed to interrupt ongoing patterns (or careers) of victimization.


    Acknowledgments
 
This project was supported by Grant # 2000-WT-VX-0019 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. Conflict of interest: None declared.


    Footnotes
 
Gerald T. Hotaling passed away unexpectedly on June 16, 2002. Although he worked closely with us developing the conceptual framework and analyzing the data presented in this research study, he was unable to review the final manuscript. Therefore, his coauthors take sole responsibility for any interpretations or conclusions drawn from the research presented here while also emphasizing that Hotaling was the lead investigator on this project.

i These records contain all criminal charges filed against a defendant by any Massachusetts court during his lifetime, the dates of occurrence and court locations of each charged offense, as well as the defendant's age at time of first offense. These data were coded into several categories including the age of the defendant at time of first criminal charge, the overall number of prior criminal charges, and the total number of prior criminal charges for crimes against a person, property crimes, public order offenses, sex offenses, motor vehicle offenses, and alcohol and drug charges. Back

ii From this data source, we were able to construct measures on (a) the number of restraining orders taken out on the study defendant prior to the study incident, (b) the number of different female victims who have taken out restraining orders against the study defendant, (c) whether a restraining order was in effect at the time of the study incident and, (d) whether a new restraining order was taken out against the study defendant subsequent to the study incident by the same woman as in the study incident and/or by another person. Back

iii Several variables that differentiated the three groups at the bivariate level were highly correlated with one another. In these instances, the variable with the strongest bivariate correlation was selected for inclusion in the models used here. Back


    References
 TOP
 Background: The Quincy Court...
 Results
 Discussion: Implications for...
 Concluding Comments
 References
 

    Berk RA, Berk SF, Newton PJ, Loseke DR. (1984) Cops on call: Summoning the police to the scene of spousal violence. Law and Society Review 18:479–498.[CrossRef][find it @ Stanford]

    Berk SF, Campbell A, Klap R, Western B. (1992) Bayesian analysis of the Colorado Springs spouse abuse experiment. Criminal Law and Criminology 83:170–200.[CrossRef][find it @ Stanford]

    Black D. (1980) The manners and customs of the police(Academic Press, New York).

    Buzawa E, Hotaling G, Klein A, Byrne J. (1999, July) Response to domestic violence in a pro-active court setting: Final report(National Institute of Justice, Washington, DC).

    Buzawa ES and Buzawa CG. (2003) The criminal justice response. Domestic violence 3rd ed. (Sage, Thousand Oaks, CA).

    Byrne J and Taxman F. (2005) Crime control is a choice: Divergent perspectives on the role of treatment in the adult corrections system. Criminology and Public Policy 4:291–310.[CrossRef][find it @ Stanford]

    Byrne J and Taxman F. (2006) Crime control strategies and community change. Federal Probation June3–12.

    Byrne J, Taxman F, Young D. (2002) Emerging roles and responsibilities in the reentry partnership initiative: New ways of doing business(Office of Justice Programs, National Institute of Justice, Washington, DC).

    Center for Court Innovtation. (2006) Responding to domestic violence: The role of the court system. Briefing paper(Center for Court Innovation, New York, New York).

    Dunford FW, Huizinga D, Elliott D. (1989) The Omaha domestic violence police experiment: Final report to the National Institute of Justice and the City of Omaha(Institute of Behavioral Science, Boulder, CO).

    Felson R and Paré PP. (2005) The reporting of domestic violence and sexual assault by nonstrangers to the police. Journal of Marriage and Family August 67:3.

    Felson RB and Ackerman J. (2001) Arrest for domestic and other assaults. Criminology 39:655–675.[CrossRef][ISI][find it @ Stanford]

    Felson RB, Messner SF, Deane G. (2002) Reasons for reporting and not reporting domestic violence to the police. Criminology 40:617–647.[CrossRef][ISI][find it @ Stanford]

    Ferraro K. (1989) Policing women battering. Social Problems 36:61–74.[Medline][find it @ Stanford]

    Finkelhor D, Hotaling GT, Lewis IA, Smith C. (1990) Sexual abuse in a national survey of adult men and women: Prevalence, characteristics, and risk factors. Child Abuse and Neglect 14:19–28.[CrossRef][ISI][Medline][find it @ Stanford]

    Finkelhor D and Dziuba-Leatherman J. (1994) Children as victims of violence: A national survey. Pediatrics 94:4413–420.[Abstract/Free Full Text]

    Finkelhor D, Wolak J, Berliner L. (2001) Police reporting and professional help seeking for child crime victims: A review. Child Maltreatment 1:617–30.

    Fleury RE. (2002) Missing voices: Patterns of battered women's satisfaction with the criminal legal system. Violence Against Women 8:181–205.[Abstract/Free Full Text]

    Garner J, Fagan J, Maxwell C. (1995) Published findings from the spouse assault replication program: A critical review. Journal of Quantitative Criminology 11:13–28.

    Garner J and Maxwell C. (1999) What are the lessons of the police arrest studies? In Ward S and Finkelhor D (Eds.). Program evaluation and family violence research(Haworth, New York) pp. 83–114.

    Hart TC and Rennison C. (2003) Reporting crime to the police, 1992–2000(Bureau of Justice Statistics. Special report. U.S. Department of Justice, Washington, DC).

    Hirschel JD and Hutchison IW. (1992) Female spouse abuse and the police response: The Charlotte, North Carolina, experiment. Journal of Criminal Law and Criminology 83:73–119.[CrossRef][ISI][find it @ Stanford]

    Kluger J. (2006) New York's Integrated Domestic Violence court system Retrieved November 11, 2006, from http://www.nycourts.gov/ip/domesticviolence/index.shtml.

    Labriola M, Rempel M, Davis R. (2005, November) Testing the effectiveness of batterer programs and judicial monitoring. Final report(National Institute of Justice, Washington, DC).

    Maxwell C, Garner J, Fagan J. (2001) The effects of arrest on intimate partner violence: New evidence from the spouse assault replication program. Research in brief(National Institute of Justice, Washington, DC).

    Mills LG. (1998) Mandatory arrest and prosecution policies for domestic violence: A critical literature review and the case for more research to test victim empowerment approaches. Criminal Justice and Behavior 25:306–318.[Abstract]

    Mills LG. (1999) Killing her softly: Intimate abuse and the violence of state interventions. Harvard Law Review 113:551–613.[find it @ Stanford]

    Mills LG. (2003) Insult to injury: Rethinking our responses to intimate abuse(Princeton University Press, Princeton, NJ).

    National Research Council. (2004) Advancing the federal research agenda on violence against women(The National Academies Press, Washington, DC).

    Pate A and Hamilton E. (1992) Formal and informal deterrents to domestic violence: The Dade County spouse assault experiment. American Sociological Review 57:691–697.[CrossRef][ISI][find it @ Stanford]

    Radford J. (1989) Women and policing: Contradictions old and new. In Hanmer J, Radford J, Stanko B (Eds.). Women, policing and male violence(Routledge & Kegan Paul, London) pp. 13–45.

    Felson Richard B, Messner Steven F, Hoskin Anthony, Deane Glenn. (2002) "Reasons for Reporting and not Reporting Violence to the Police". Criminology 40:617–648.

    Roberts A. Classification typology and assessment of five levels of woman battering. Journal of Family Violence in press.

    Roberts A and Everly G. (2006) A meta-analysis of 36 crisis intervention studies. Brief Treatment and Crisis Intervention 1–12.

    Roberts AR and Roberts B. (2005) Ending intimate abuse: Practical guidance and survival strategies(Oxford University Press, New York).

    Sherman LW, Schmidt JD, Rogan DP, Smith DA, Gartin PR, Cohn EG, et al. (1992) The variable effects of arrest on crime control: The Milwaukee domestic violence experiment. Journal of Criminal Law and Criminology 83:137–169.[CrossRef][ISI][find it @ Stanford]

    Taxman F, Young D, Byrne J. (2002) From prison safety to public safety: Innovations in offender reentry. One of a series of articles on reentry issues to be published by the Office of Justice Programs(National Institute of Justice, Washington, DC).


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
7/1/55    most recent
mhl020v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Buzawa, E.
Right arrow Articles by Byrne, J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Buzawa, E.
Right arrow Articles by Byrne, J.
Social Bookmarking
 Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?