According to the Bureau of Justice Statistics’ annual victimization reports, approximately 691,000 nonfatal intimate partner violence (IPV) victimizations occurred annually in the United States between 2013 and 2017.[1] In a small number of incidents, the violence became lethal. Incident data reported by law enforcement agencies to the FBI over the past 15 years indicate that, each year, about 1,400 people were killed by their current or former intimate partners. In the majority of these homicides, the victim was a woman killed by her male (ex)partner.[2] As a result, there were approximately 6,400 more female intimate partner homicide victims than male victims over the past decade. While homicides in the United States predominately involve people who are young adult males, intimate partner homicides largely involve men killing their female partners.[3]
Beginning in the 1980s, advocates and legal scholars sought to criminalize IPV by implementing pro-arrest and mandatory arrest policies, supporting preferred prosecution policies, enforcing protective orders, or requiring intervention programs for those convicted of abuse.[4] They also sought consistent and coordinated responses across criminal justice and community-based organizations that were effective and focused on the safety of victims.[5] Over the past several decades, these efforts have ranged from establishing second-responder programs within law enforcement agencies, where crisis response teams make home visits following an initial police response, to instituting innovative pretrial strategies, such as increased involvement by the judge in managing those committing IPV during the pretrial period, restructured court procedures, and expanded victim services.[6]
A more recent development in how law enforcement responds to IPV incidents involves assessing the victim, person who committed the offense, or both for risk and needs and then connecting one or both parties to appropriate resources regardless of whether an arrest occurs. In 2015, the Police Executive Research Forum (PERF) found that 42% of surveyed law enforcement agencies reported using a structured risk assessment to determine if a victim is in danger of future violence. The PERF report also indicated that 39% of U.S. agencies use a risk-assessment approach to identify persons who repeatedly commit crimes.[7]
Currently, two victim-focused models of IPV risk assessment are used in the United States: the Lethality Assessment Program (LAP) and the Domestic Violence High-Risk Team (DVHRT) model. Although current evidence about these interventions is promising, particularly in terms of an LAP leading to women’s use of more protective strategies, a more in-depth study is needed to establish these collaborative risk assessment models as evidence-based.[8] As a result, in 2012 NIJ and the Office on Violence Against Women (OVW) launched the Domestic Violence Homicide Prevention Demonstration Initiative to further expand the evidence base about how these models work in practice and how they impact survivors and the criminal justice and service systems.
Victim-Focused Models
LAP is a police-led model that largely follows the Lethality Assessment Program–Maryland Model, which was created in the early 2000s through a collaboration of advocates, researchers, and law enforcement practitioners.[9] The LAP assessment is based on Dr. Jacquelyn Campbell’s Danger Assessment instrument, which she developed with the support of NIJ and other federal agencies to determine the likelihood that a man would kill his female intimate partner. The assessment instrument uses a calendar to guide the victim’s recall and to document the frequency and severity of assaults over the past year, along with 20 questions about the individual’s behaviors and threats or about other circumstances in the survivor’s life, to calculate a weighted risk score for each victim. A validation study illustrated that danger levels or scores that resulted from the Danger Assessment instrument captured more than 90% of potentially lethal cases.[10] This finding led others to assert that if IPV homicides are predictable, they are preventable — this has been the impetus behind OVW funding the implementation and testing of such tools.[11]
The LAP protocol — an 11-item screener conducted by law enforcement as well as other allied professionals — focuses on identifying those who are most at risk of becoming a homicide victim or experiencing a serious, lethal-like assault. Officers at the scene of an incident connect high-risk victims with an advocate by phone; the advocate has a brief discussion with the victim about safety planning. Following the incident, the law enforcement agency and service program personnel are encouraged to follow up with victims, particularly those who are most at risk, to connect them with services.
The immediate connection with an advocate — initiated by law enforcement — is likely a critical component, as victims who call the police do not often follow up with victim services. Although the rate of women who call the police to seek help increased from 40% in 1992 to 60% in 2014, the rate of victims across the United States who seek victim services remained at about 30% during this same period.[12] Thus, LAP aims to better connect victims who come in contact with law enforcement to victim advocacy services for assistance with safety planning, obtaining orders of protection, and shelter.
The second victim-focused model of IPV risk assessment, the DVHRT or high-risk team approach, has also gained traction across the country over the past decade.[13] Under this approach, law enforcement personnel seek to identify victims at high risk for lethal-like violence using the Danger Assessment–Law Enforcement tool.[14] They then forward information about the highest risk victims to their community’s multidisciplinary DVHRT. These teams often include representatives from victim service organizations, law enforcement agencies, pretrial services, prosecutors’ offices, corrections, and batterer intervention programs. They meet regularly to review the referred victims and decide, as a team, which victims warrant intensive assistance and monitoring.
Victims identified as needing intensive assistance must agree to share their information with the DVHRT. The staff then gathers more information about the victim and individual who abused them from the team’s member agencies — as well as from the victim — to develop a protection plan that allows each agency to contribute in a manner that fits its mission and resources. The team also collects case-level outcome information and reviews the status of existing cases each month to plan follow-up actions as needed. Prosecutors who are members of the team can use this information to inform bail and pretrial release recommendations. Eventually, the team decides if a victim’s risk has decreased to a point where this level of monitoring is no longer indicated.
By design, the DVHRT serves many fewer victims than LAP, which serves all high-risk victims. At least one U.S. municipality is trying to implement both protocols simultaneously.
Model Focused on Persons Committing the Offense
Several U.S. law enforcement agencies are adopting an IPV risk assessment approach built on the focused deterrence model that served as the framework for the U.S. Department of Justice’s Project Safe Neighborhoods initiative.[15] This model — a problem-oriented policing initiative — seeks to prevent violence by directly contacting persons convicted of a violent offense and sending them a message that their “violence is no longer tolerated.” When violence does occur, an enforcement strategy guided by state and federal statutes is used.[16]
With support from the Office of Community Oriented Policing Services, North Carolina’s High Point Police Department and researchers evaluated whether an approach that focuses on the person committing the offense reduces IPV homicides.[17] Their approach uses criminal history information to classify them into four groups organized by severity of risk. The lowest risk group includes individuals who have no criminal charge for IPV-related cases but have had a complaint filed against them. The second group is made up of those with one IPV-related charge. The third group includes persons with a second IPV charge or who were in the second group but have now committed a court-prohibited behavior (e.g., violation of a protection order). The fourth group comprises individuals with three or more IPV-related charges, a violent criminal record that could include at least one IPV-related incident, a violation of a civil protection order, a weapon used during an IPV-related incident, or a prior felony conviction.
The department deploys a series of graduated tactics for each group, beginning with a notification procedure — that is, those in the lowest risk group receive a letter stating that law enforcement sees them as a person that may commit an offense. The department then uses adaptable monitoring (e.g., an individual is flagged in the records management system) and service provisions such as referring them to mental health and substance abuse programs, and at the highest level, a series of increasingly swift and severe criminal justice sanctions (e.g., enhanced prosecution resources). This sanction regimen includes periodic call-in meetings, during which law enforcement officials and community members meet with selected persons who have committed a violent offense to review both the sanctions used to punish those who commit the most serious offenses and the community services available that may help them avoid these sanctions.[18]
The State of the Evidence
Over the last two decades, various federal agencies (including NIJ) have supported research that has led to the development and testing of programs to assess for serious and lethal violence, particularly among females.
For example, Campbell and colleagues — with support from NIJ, the Centers for Disease Control and Prevention, and the National Institutes of Health — completed a multicity study to assess which factors in relationships, beyond the occurrence of IPV, predict intimate partner femicide.[19] This research identified 11 factors that distinguished victims who were abused and killed by their partners from those who experienced abuse only. These factors include the person's employment status, the type relationship between the victim and the person committing the offense, the presence of a child from the victim’s previous relationship, the individual's use of control tactics after separation, the their threatened or actual use of a weapon, and the lack of a prior arrest among active individuals. The findings suggest that criminal justice practitioners could combine these risk factors into a score that would identify more than 90% of potentially lethal IPV cases.[20]
In 2008, Jill Theresa Messing and colleagues — with support from NIJ — examined whether the use of the LAP screen decreased the rate of repeat, severe, lethal, and near-lethal violence and increased the rate of emergency safety planning and help-seeking.[21] Their research found that a majority of victims who were willing to participate in a study interview also spoke to the hotline advocate during their encounter with law enforcement. Their outcome analyses further indicated that women who spoke to the advocates reported using significantly more protective strategies over the next six months and experiencing significantly less victimization than women in a comparison group.[22] The women who participated in the LAP intervention were also significantly more satisfied with law enforcement’s response and more likely to report that their advocate was at least somewhat helpful.[23]
These findings are promising, but more rigorous research is needed to determine the effectiveness of these IPV interventions. To fill in gaps and expand the knowledge base, in 2012 NIJ and OVW collaborated to launch the Domestic Violence Homicide Prevention Demonstration Initiative.[24] Through this initiative, NIJ supports a multidisciplinary team of researchers and practitioners who are examining the implementation of the LAP and DVHRT models across six sites. The project, which began in 2014, is seeking to validate previous findings related to LAP and to establish baseline evidence on the functioning and outcomes of the DVHRT model. The evaluation design includes collecting data about intimate partner incidents from each site’s criminal justice system and linking these data with information from victim service providers and confidential victim interviews conducted in two sites. This data collection will allow policymakers to see how overlapping systems can manage risk and how this process influences further victimization.
In addition to studying how communities are implementing the interventions, the NIJ-supported evaluation team[25] is interviewing victims, victim service professionals, police officers, and key executive members of the local law enforcement communities at multiple points during the implementation process. These interviews will provide information about program implementation in each community and also reveal how implementation has affected the community’s response to IPV.
“As a direct result of being part of the Domestic Violence Homicide Prevention Demonstration Initiative, I now incorporate much thought, along with specific conversations surrounding data collection and the potential for future research, into any decision that will impact our agency from the perspective of policy and/or practice,” explains John Guard, Chief Deputy of the Pitt County Sheriff’s Office in Greenville, North Carolina.
“The days of making agency decisions based on one’s gut feeling are fleeing,” Guard adds. “We need programs supported by solid research that can be replicated. This is why NIJ’s support of research within the criminal justice system is so valuable to law enforcement agencies around the country.”
Results from the multisite evaluation effort are expected to be available in 2021.
For More Information
Learn more about NIJ’s IPV work.
Read more about the Domestic Violence Homicide Prevention Demonstration Initiative.
About this Article
This article discusses the following grants:
- “Risk Assessment Validation Study,” grant number 2000-WT-VX-0011
- “Police Department’s Use of Lethality Assessments: An Experimental Evaluation,” grant number 2008-WG-BX-0002
- “Evaluation of the Domestic Violence Homicide Prevention Initiative,” grant number 2013-ZD-CX-0001
This article was published as part of NIJ Journal Issue Number 282.