Generalized and racialized consequences of the police response to intimate partner violence in the U.S.: A systematic scoping review

Abstract

Intimate partner violence (IPV) impacts more than 40% of people in the U.S. Since the 1980s, the U.S. has maintained a police-centric response to IPV, which relies on arrest—via policies like mandatory arrest laws—as its primary intervention. There is mixed evidence on whether IPV policing decreases subsequent IPV at the individual level, but less is known about IPV policing’s broader collateral consequences. This systematic scoping review is the first to synthesize existing evidence for the generalized consequences of IPV policing in the U.S. We searched Web of Science, ProQuest, and EBSCO Host, and identified 36 relevant articles. Survivor criminalization was the most studied generalized consequence of IPV policing and existing studies have documented positive associations between mandatory arrest laws and risk of survivor arrest. We also found numerous methodologically rigorous studies on the effects of mandatory arrest laws on population-level IPV victimization. The review also identifies gaps in the evidence base: there is a need for research on additional potential consequences of IPV policing such as police violence against survivors, involvement of child protective services, and psychosocial and physical health outcomes of survivors.

1.1. Introduction

Intimate partner violence (IPV) impacts more than 40% of people in the U.S. in their lifetime and constitutes a growing proportion of violent crime (Catalano, 2012; Leemis et al., 2022). IPV includes physical violence, sexual violence, psychological abuse, stalking, and other forms of coercion between current or former spouses or dating partners. Since the 1980s, the primary response to IPV in the U.S. has been policing, prosecution, and incarceration. This “carceral” response to IPV has persisted despite mixed evidence of its effectiveness, increased recognition of the harms of mass criminalization and incarceration, and growing calls for criminal legal reform (Kim, 2018; National Institute of Justice, 2016). However, the carceral response to IPV—especially policing—may have generalized consequences for survivor¹ health and safety that have not been systematically documented. In this scoping review, we conduct the first known summary of the empirical evidence concerning generalized consequences of IPV policing in the U.S., including how these consequences may vary by survivor race.

1.1.1. The problem of IPV and the rise of IPV policing

IPV is a highly prevalent problem with severe consequences for health and safety (Breiding et al., 2015). Most recent national estimates demonstrate that more than 40% of people experience contact sexual violence, physical violence, and/or stalking by an intimate partner in their lifetime (Leemis et al., 2022). There are also pronounced racialized disparities in the prevalence and severity of IPV (Caetano et al., 2005). For example, in New York City, Black women represent 13% of the city’s population but 30% of intimate partner homicide victims in the city (NYC Mayor’s Office to End Domestic and Gender-Based Violence, 2021). The health and safety consequences of IPV are manifold, including physical injury and death, as well as long-term chronic health consequences such as heart disease, reproductive disorders, gastrointestinal disorders, and psychiatric disorders (Campbell, 2002; Catalano et al., 2009; Sugg, 2015). Additionally, from the early 1990s to 2010, the U.S. experienced dramatic reductions in reported rates of overall violent victimization, while reported rates of IPV declined at far slower rates (Catalano, 2012). Moreover, rates of IPV homicide have been increasing in recent years (Fridel & Fox, 2019). The need for effective strategies to reduce and prevent IPV is at a critical stage for public health and public safety.

Prior to the 1980s, IPV had historically been viewed as a private matter that did not require external intervention (Erez, 2002; Oppenlander, 1982). However, through the confluence of the Battered Women’s Movement and the rise of the “tough-on-crime” era, IPV was redefined in the public eye as a criminal legal issue (Erez, 2002; Young & Stein, 2004). Applying a criminal legal perspective to the problem of IPV has in turn shaped strategies to reduce and prevent IPV. Namely, the U.S. employs a “carceral feminist” response to IPV, meaning that it relies on policing, prosecution, and incarceration as its primary tools to advance the ostensible goal of protecting women from gender-based violence (Kim, 2018; National Institute of Justice, 2016). This carceral feminist response to IPV arose as part of a larger carceral evolution in the U.S., as racialized criminalization was applied to a growing number of social problems leading to unprecedented levels of policing and criminal legal punishment, particularly for Black Americans and other communities of color—what is now widely known as mass incarceration (Thompson, 2010). This national shift toward a carceral response to IPV was consistent with a larger global movement to criminalize IPV in the 1990s, though the scope and scale of the U.S. criminal legal system is unmatched globally (Fair & Walmsley, 2021; Htun & Jensenius, 2020).

A primary example of the carceral feminist response to IPV is mandatory arrest laws. Mandatory arrest laws require police to make an arrest when responding to IPV calls, regardless of police discretion and survivor wishes (Hirschel et al., 2007). These laws grew in national popularity due in part to the findings from a 1984 randomized trial called the Minneapolis Domestic Violence Experiment (MDVE). The MDVE was a six-month trial with 314 subjects in which police officers were randomly assigned to make an arrest, send the abusive partner away for 8 hours, or offer counseling in response to misdemeanor IPV incidents (Sherman & Berk, 1984). Results from the MDVE showed that men who were arrested had reduced likelihood of reoffending after six months (Sherman & Berk, 1984). Although the findings were not successfully reproduced in subsequent trials (as discussed below), by 1989, over 75% of jurisdictions across the country changed their domestic violence laws to allow for warrantless arrests for domestic violence misdemeanors (Erez, 2002) and by 1992, 15 states and the District of Columbia had adopted mandatory arrest laws for domestic violence (Schmidt & Sherman, 1993). Mandatory arrest laws are one type of warrantless arrest law for IPV. Other types include discretionary arrest laws (officers may arrest without a warrant) and preferred arrest laws (arrest is the preferred response). Today, mandatory arrest and other pro-arrest policies remain a major component of the nation’s police response to IPV. The policing response to IPV also includes policing practices, such as enforcement of no-contact orders, and staffing decisions, such as the funding of specialized police units for domestic violence.

1.1.2. Potential generalized consequences of IPV policing

While the ostensible goal of IPV policing is to prevent and reduce IPV victimization, since the growth of the carceral response to IPV, many stakeholders (e.g., survivor groups, activists, criminologists, and lawyers) have raised concerns that this response may not effectively meet survivors’ needs and may have adverse consequences for survivors, with experts highlighting that poor and Black survivors were likely to face the brunt of such adverse consequences (Goodmark, 2018, 2023; Hinerfeld, 2020; Richie, 2000; Ruttenberg, 1993; Sherman, Schmidt, & Rogan, 1992). First, critics have argued that IPV policing policies and practices, like mandatory arrest laws, restrict survivor agency and choice (Leisenring, 2008). There are numerous reasons why survivors of IPV may not want to engage with the criminal legal system or may not want their abusive partner to be arrested. For example, they may have strong emotional and/or financial ties with the abusive partner, they may fear that the abusive partner will become more violent in retaliation for having the police called on them, or they may not trust the system to deliver safety (Dugan et al., 2003; Leisenring, 2008). As a result, the presence of mandatory arrest laws may have a chilling effect that discourages survivors from reporting IPV and seeking help (Dugan, 2003).

Second, a common critique of the policing response to IPV is that it extends the reach of carceral control, contributing to mass criminalization and incarceration, which has known harmful economic, psychosocial, and health consequences, particularly for Black Americans (Alexander, 2012; Anonymous, 2003; Davis et al., 2022). Further, investments in this carceral response come at the expense of more significant investments to address the structural factors that give rise to IPV, such as housing insecurity and financial dependence (Goodmark, 2018; Snyder, 2019).

Lastly, critics have argued IPV policing may result in adverse downstream consequences for survivors. The arrest and/or incarceration of a romantic partner, spouse, or co-parent can impact the psychosocial and economic well-being of survivors in the short- and long-term and, as a result, may have generalized consequences for survivor health. For example, partner incarceration is longitudinally associated with negative health outcomes for women including increased risk of substance use, major depressive disorder, and AIDS (Bruns & Lee, 2020; Johnson & Raphael, 2009; Wildeman et al., 2012, 2019). Partner incarceration is also associated with financial consequences (e.g., loss of a source of income) and civil disqualifications (e.g., exclusion from public housing), which can in turn impact the health of a survivor and their family (Wheelock & Uggen, 2006).

Survivors can also face direct adverse consequences including the criminalization of survival, or the arrest, prosecution, and/or incarceration of survivors of violence for acts tied to their own survival, such as self-defense (Goodmark, 2023); police violence against survivors (Coker et al., 2015; Ritchie, 2017); and child protective services involvement (Coker et al., 2015; Roberts, 2020). IPV policing can trigger investigations from child protective service agencies because there is typically strong coordination between government systems of policing and family surveillance, and some jurisdictions treat IPV as evidence of child maltreatment (Edwards, 2019; Rebbe et al., 2021; Roberts, 2001, 2022). A child maltreatment investigation can be conceptualized as an adverse consequence for survivors because it can lead to long-term family regulation, child removal, and termination of parental rights (Roberts, 2001, 2022).

These potential generalized consequences of IPV policing have been largely ignored in the research literature and by policymakers. The U.S. has maintained a police-centric response to IPV despite increased recognition of the harms of mass criminalization and incarceration and growing calls for criminal legal reform. Criminal legal reform efforts have largely excluded violent criminalized behavior generally, and IPV specifically (Beckett et al., 2016, 2018; Goodmark, 2018, 2019; Siegel, 2016). In fact, criminal legal reform efforts often include specific exceptions for IPV or even include harsher criminal legal punishments for IPV (Beckett et al., 2016, 2018; Goodmark, 2018, 2019; Siegel, 2016).

In maintaining and reinforcing a police-centric response to IPV, policymakers are ostensibly weighing the societal costs of increased criminalization against the purported public safety benefits of IPV policing and determining that the benefits outweigh the costs. However, the public safety benefits of IPV policing have not been consistently empirically demonstrated and evidence concerning the generalized consequences of IPV policing has not been comprehensively evaluated.

1.1.3. Evidence concerning the effects of IPV policing

After nearly three decades of research on the topic, the evidence concerning whether IPV policing decreases subsequent IPV victimization at the individual level remains mixed. Following the MDVE, the National Institute of Justice funded a series of replication studies in five U.S. cities to measure the effect of arrest vs. non-arrest alternatives such as mediation or separation for misdemeanor domestic violence assault causes on subsequent revictimization (Berk et al., 1992; Dunford et al., 1990; J. D. Hirschel & Hutchison, 1992; Pate & Hamilton, 1992; Sherman, Schmidt, Rogan, et al., 1992). Those trials, known as the Spouse Assault Replication Program (SARP), resulted in conflicting findings, with some showing evidence of a deterrent effect of arrest, some showing no difference between arrest and non-arrest, and one showing an increased risk of subsequent violence after arrest. Variation in study locations/populations, sample size, and alternate police responses tested (e.g., mediation vs. separation) may have driven some of these differences in findings.

Since then, multiple pooled analyses and meta-analyses have attempted to synthesize the findings from these trials and more recent trials. Two such synthesis efforts focused on the SARP trials found that there was no statistically significant effect of arrest on subsequent victimization based on police reports (Maxwell et al., 2002; Sugarman & Boney-McCoy, 2000). However, based on survivor interviews, arrest was associated with reductions in subsequent self-reported victimization of 25% (Maxwell et al., 2002) and 5% (Sugarman & Boney-McCoy, 2000). The most recent meta-analysis, published in 2020, included a total of 11 trials evaluating the individual effects of arrest vs. non-arrest on subsequent IPV victimization using official records (arrest reports, police reports, and hotline reports) (Hoppe et al., 2020). The summary estimate across the 11 studies was null (OR=0.946; 95% CI: 0.787, 1.138), demonstrating no effect of arrest on subsequent IPV victimization based on official records (Hoppe et al., 2020). Together, existing evidence from individual randomized trials suggests that arrest reduces subsequent victimization based on self-report data but not official records data.

Not only is the evidence base on the individual-level effect of partner arrest on subsequent IPV victimization mixed, but it also provides an incomplete picture of the potential consequences of IPV policing. This research has been focused on the individual-level effect of arrest (and not population-level effects of mandatory arrest laws) and has primarily focused on short-term (e.g., 6-month) outcomes. When evaluating how IPV policing impacts survivor health and safety, it is important to consider broader measures of survivor health and well-being beyond individual re-victimization alone. Additionally, longer-term analysis is important because, for example, in the case of arrest and incarceration for an IPV incident, the violence may stop in the immediate term while the abusive partner is incarcerated, but the violence may resume or worsen in the longer term if the root cause of the violence goes unaddressed upon release from incarceration. Indeed, a meta-analysis of 57 studies showed that incarceration of abusive partners was associated with a significant increase in their subsequent rate of offending (Garner et al., 2021). Systematic study and verification of these generalized consequences at the population level may suggest that the police-centric response to IPV in fact does more harm than good for survivor safety, health, and well-being. Namely, if IPV policing is not consistently effective at decreasing subsequent IPV victimization at the individual level, and also leads to harmful consequences for survivor safety and well-being, then it is critical to reimagine our response to IPV.

1.1.4. Purpose of the present review

The empirical evidence concerning the generalized effects of IPV policing beyond individual-level IPV revictimization has not been comprehensively reviewed and synthesized. Further, extant studies are siloed within different fields, including criminology, law, economics, and women’s studies, making evidence difficult to assess as a coherent body of research. The objective of the present scoping review is to summarize, for the first time, the existing U.S. research concerning the potential generalized consequences of policing and arrest as a response to IPV, including effects on population-level IPV rates, reporting of IPV to police, survivor criminalization, and other aspects of survivor health and well-being. We also assess the extent to which prior research has considered racialized differences in those consequences and where gaps in the evidence base exist.

1.2. Methods

1.2.1. Literature search

We conducted a search of Web of Science, ProQuest (which includes MedLine), and EBSCO Host (which includes the National Criminal Justice Reference Service and PsycInfo). We used the following search terms: (“mandatory arrest” OR “arrest policy” OR “arrest rate” OR “rate of arrest”) AND (“domestic violence” OR “intimate partner violence” OR “domestic abuse” OR “spous* assault” OR “wife abuse” OR “wife beating”). We additionally restricted the search to articles published in English in scholarly journals in the year 1980 or later because the police-centric response to IPV arose in the 1980s. The last search was completed on October 16, 2022.

1.2.2. Inclusion/exclusion criteria

Eligible articles included all empirical studies (quantitative and qualitative) focused on a U.S. population that assessed consequences of the police-centric response to IPV. The study population could be local or national so long as the outcome was a survivor-focused outcome. The specific exposures of interest were policing policies and practices in response to IPV, such as mandatory arrest or other warrantless arrest legislation/policies, individual arrests for IPV, interactions with police, or the implementation of specialized IPV policing units or specialized training. The outcomes of interest were any survivor-based outcome, excluding individual-level IPV revictimization following arrest because of the many existing reviews and meta-analyses focused on individual-level IPV revictimization (Garner et al., 2021; Garner & Maxwell, 2000; Gelles, 1993; Homant, 1985; Hoppe et al., 2020; Leisenring, 2008; Maxwell et al., 2002; Maxwell & Garner, 2012). Outcomes of interest included population-level measures of IPV victimization rates or IPV homicide rates and survivor-based outcomes other than individual risk of revictimization, such as risk of dual arrest (arrest of both the perpetrator and survivor of IPV), rates of reporting to police, and measures of survivor health and well-being.

1.2.3. Screening and abstraction

We used the systematic review management software Covidence to conduct the screening and abstraction process. After removing duplicates, one author (SK) individually conducted the title and abstract screen. The full-text screen and data abstraction process were conducted in collaboration with two of the co-authors (MSA and EB). Two reviewers independently screened each full-text article for inclusion. Any disagreements between the two reviewers were resolved with input from the third reviewer. We also conducted a review of the reference lists of all included articles to identify additional articles that met the inclusion criteria. Using a standardized form, data were abstracted independently by two reviewers on study characteristics such as study design, study population, unit of analysis, operationalization of exposure and outcome, main findings, and whether the study included a racial analysis. We pilot tested the form on three studies and edited as necessary before using it for all included studies. Data abstraction forms were compared, and any disagreements were resolved with input from the third reviewer. Methods were consistent with the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) (checklist in Appendix A). The review was registered with Prospero, an international database of systematic review protocols, on March 12, 2021 (CRD42021288849).

1.3. Results

1.3.1. Search results

Figure 1 depicts a flowchart summarizing the search and screening process. The literature search yielded 1,089 articles after duplicates were removed. Of those, 968 were excluded during the title and abstract screen, resulting in 121 articles for the full-text screen. An additional 87 articles were excluded during the full-text screen. The most common reasons for exclusion were: (1) exposure is not IPV policing (n=22), (2) non-empirical (e.g., systematic reviews or legal scholarship) (n=19), (3) outcome is probability of offender arrest (n=10), and (4) descriptive (n=10) (e.g., describing characteristics of a population with no analytic comparisons). Following the full-text screen, 34 eligible articles were identified. Two additional articles were identified through searches of reference lists of included articles, resulting in a total of 36 articles that met the inclusion criteria.

Figure 1: Flowchart summarizing article search and selection process.

1.3.2. Description of included articles

This section provides an overview of the publication year, journal discipline, study design, study population, exposure(s), and outcome(s) of each study as well as a discussion of whether the study included an analysis of differential impacts of IPV policing by race. Characteristics of included studies are also summarized in Table 1.

Table 1:

Characteristics of included studies

First author	Year	Journal discipline	Study design	Study setting	Exposure(s)	Outcome(s)	Racial analysis	Findings summary
Yegidis	1994	Social work	Cross-sectional	4 IPV shelters in FL county	Police interactions under preferred arrest policy	Experience with police	Yes	Discrepancy between survivor wishes and police action; Black women less satisfied with police response than white women
Erez	1998	Violence (multidisciplinary)	Qualitative	2 OH counties	Police contact	Experience with police	No	Survivors wish to retain autonomy in attempts to stop abuse
Stephens	2000	Criminology	Qualitative	Semi-rural NY county	Abusive partner arrest under new mandatory arrest law	Experience with police	No	Survivors wish to retain autonomy in attempts to stop abuse
Dugan	2002	Criminology	Longitudinal, population-level	48 large US cities	Warrantless arrest laws	IPV homicide rates	Yes	Mandatory arrest associated with lower white femicide rate, though not robust to city-dependency test
Bui	2003	Women’s studies	Qualitative	4 US cities	Police contact	Help seeking decisions	Yes	Experiences of survivors and help-seeking behaviors are complex
Dugan	2003a	Criminology; Public policy	Longitudinal, population-level	National	Warrantless arrest laws	Household violence; reporting to police	No	Mandatory arrest associated with lower odds of spousal violence and informing police; no effect on family or dating violence
Dugan	2003b	Law	Longitudinal, population-level	48 large US cities	Warrantless arrest laws; local pro-arrest policies; local police commitment index	IPV homicide rates	Yes	Pro-arrest law and policies associated with lower IPV homicide for unmarried subgroups; Local specialized unit and training associated with higher IPV homicide for Black unmarried women
Miller	2003	Criminology	Randomized trial	Miami-Dade County, FL	Abusive partner arrest	Willingness to call police; satisfaction with police; feelings of safety; feelings of agency	No	Arrest positively related to perception of safety, negatively related to perception of legal power, no effect on perception of personal power
Hickman	2003	Law	Cohort	Miami-Dade County, FL	Victim preferred arrest outcome	Reporting to police	No	Victim preference associated with increased reporting to police
Wilson	2004	Criminology	Cross-sectional	National	Abusive partner arrest; whether victim reported incident to police	Satisfaction with police	No	Arrest associated with satisfaction with police; victim reporting incident associated with lower satisfaction vs. someone else reporting
Shim	2005	Violence (multidisciplinary)	Qualitative	New York metropolitan area	Arrest policies; victim exposure to arrest	Any consequence of being arrested	Yes	Experiences of survivors and help-seeking behaviors are complex
Rajah	2006	Women’s studies	Qualitative	New York City	Warrantless arrest laws	Victim identity challenges	Yes	Documented negative consequences of mandatory arrest policies for survivors
Salazar	2007	Violence (multidisciplinary)	Longitudinal, population-level	2 GA counties	Coordinated community response aiming to increase IPV legal sanctions	Female exposure to criminal legal system	No	Intervention associated with increased in female arrests
Hirschel	2007a	Criminology	Cross-sectional	19 states	Warrantless arrest laws	Dual arrests	No	Odds of dual arrest higher in mandatory vs. discretionary agencies
Frye	2007	Violence (multidisciplinary)	Cross-sectional	New York City	Warrantless arrest laws	Arrest outcome	Yes	Of cases reviewed, 5% were unwanted arrests, 9% were dual arrests, 24% were retaliatory arrests; Female arrest was more likely when partner was white
Hirschel	2007b	Criminology	Cross-sectional	19 states	Warrantless arrest laws	Dual arrests	No	Mandatory arrest laws increased likelihood of dual arrest for all relationship categories; Same sex couples more likely to experience dual arrest
Schwartz	2009	Sociology	Longitudinal, population-level	National	Triangulation of crime and violence data sources	Arrest gender gap	No	No evidence that underlying violence by women is increasing; Increases in women’s arrests more likely driven by increasing pro-arrest policies
Iyengar2	2009	Economics	Longitudinal, population-level	National	Warrantless arrest laws	IPV homicide rate; family homicide rate	Yes	Mandatory arrest laws associated with 60% increase in IPV homicide rate; Effect stronger among Black people
Hovmand	2009	Methods	Simulation	N/A	Warrantless arrest laws	Victim arrests	No	Mandatory arrest laws associated in increase in victim arrests
Zeoli	2010	Health	Longitudinal, population-level	46 large US cities	Warrantless arrest laws; laws reducing firearm access in IPV cases; police staffing levels	IPV homicide rate; IPV firearm homicide rate	No	Mandatory arrest laws not associated with IPV homicide rate or firearm homicide rate; Warrantless arrest laws associated with decrease in IPV homicide rate and firearm homicide rate; Police officers per capita negatively associated with both outcomes
Zeoli	2011	Public policy	Longitudinal, population-level	46 large US cities	Warrantless arrest laws	IPV homicide rate	No	Law classification schemas influence estimated effects; For one classification, mandatory arrest laws associated with decrease in total IPV homicide rate and male IPV homicide rate; Using 3 other classifications, no effects found
Leisenring	2011	Sociology	Qualitative	2 jurisdictions in Western state	Warrantless arrest laws	Victim identity challenges	No	Documented negative consequences of mandatory arrest policies for survivors
Xie	2012	Criminology	Longitudinal, population-level	40 large MSAs	Warrantless arrest laws; police force size; social service staff size	IPV victimization	Yes	No effect of mandatory arrest laws on IPV victimization; Police officers and social service workers per capita associated with lower IPV victimization
Durfee	2012	Women’s studies	Cross-sectional	28 states and DC	Warrantless arrest laws	Arrest outcomes	No	Mandatory arrest positively associated with female arrest and dual arrest
Dichter	2013	Criminology	Qualitative	Large East Coast city	Victim arrest as result of IPV police call	Any consequence of being arrested	No	Documented negative consequences of mandatory arrest policies for survivors; Need for new IPV response involving alternatives to arrest
Brame	2015	Criminology	Randomized trial	Lexington County, SC	Proactive enforcement of no-contact orders	Victim experiences of aggression, knowledge, safety, and contact	No	Intervention associated with higher levels of stalking/threatening behaviors and experiences of physical aggression, not associated with survivor knowledge or safety
Novisky	2015	Women’s studies	Cross-sectional	5 IPV shelters in Midwestern state	Victim support for mandatory arrest policy	Reporting to police	No	Victim support for mandatory arrest positively associated with likelihood of reporting to police
Li	2015	Criminology	Qualitative	IPV shelter in Northeastern state	Victim arrest for IPV charge	Any consequence of being arrested	No	Experiences of survivors and help-seeking behaviors are complex; Need for new IPV response involving alternatives to arrest
Sherman	2015	Criminology	Randomized trial	Milwaukee, WI	Abusive partner arrest	Victim mortality	Yes	Victim mortality rate higher among people whose partner was arrested; Effect concentrated among Black participants
Hirschel	2017	Women’s studies	Cross-sectional	4 states	Primary aggressor laws	Single and dual arrests	No	Primary aggressor laws negatively associated with likelihood of dual arrest
Dasgupta	2018	Economics	Longitudinal, population-level	National	Warrantless arrest laws	Domestic violence homicide rates; youth mental health and behavioral outcomes	No	No effect of warrantless arrest laws on domestic violence homicide rates; No effect on 4 of 5 youth behavioral outcomes; Warrantless arrest laws negatively associated with youth suicidal ideation
Chin	2019	Economics	Longitudinal, population-level	National	Warrantless arrest laws	IPV homicide rates	No	No effect of mandatory and preferred arrest laws on homicide rates; Discretionary arrest laws associated with decrease in current spouse homicide and increase in common-law spouse homicide
Cook	2019	Economics	Longitudinal, population-level	National	Warrantless arrest laws	IPV intimidation and assault rate	No	Mandatory arrest laws negatively associated with intimidation rate only when unemployment rates between 4.75% and 8.5%; No effect on assault rate
Gezinski	2020	Violence (multidisciplinary)	Qualitative	Utah	Warrantless arrest laws	Experience with police	No	Documented negative consequences of mandatory arrest policies for survivors; Need for new IPV response involving alternatives to arrest
Hirschel	2021	Criminology	Cross-sectional	37 states	Primary aggressor laws	Single and dual arrests	No	Primary aggressor laws not associated with likelihood of dual arrest
Amuedo-Dorantes	2022	Law; Economics	Longitudinal, population-level	1,148 counties	Sanctuary policies	Domestic violence homicide rates	Yes	Sanctuary policies reduce domestic violence homicide rate for Hispanic female victims; This effect dissipates in states with mandatory arrest laws

²Corrigendum to Iyengar 2009 issued in 2019 noting potential but unverified coding error

Publication year.

All of the articles were published in 1994 or later, with the majority published after 2002.

Journal discipline.

Criminology was the most common discipline of the journals in which the articles were published: 36% (n=13) of the articles were published in a criminology journal. Other common journal disciplines were women’s studies (n=5), violence (multidisciplinary) (n=5), and economics (n=5).

Study design.

Of the 27 quantitative studies, nearly half (n=13) were longitudinal, population-level analyses using data from multiple jurisdictions over time. Of the remaining 14 quantitative studies, nine were cross-sectional, three were randomized trials, one was a cohort study, and one was a simulation study. Nine studies were qualitative.

Study setting.

About 22% (n=8) of the studies used a national sample and another 25% (n=9) used a study sample that covered many states, metropolitan statistical areas, or cities. The remaining studies, excluding the simulation study, were focused on one to five jurisdictions, spanning various regions across the country.

Exposures.

Half of the articles (50%, n=18) assessed the impacts of warrantless arrest laws for IPV as the IPV policing exposure, typically operationalized as mandatory arrest laws alone or as a combination of mandatory, preferred, and discretionary arrest laws. In four additional articles, the exposure of interest was the individual arrest of the abusive partner. Other common exposures included measures of general police contact or interactions during an IPV incident (n=3) and police force size/staffing (n=3).

Outcomes.

The most common survivor-related outcomes of IPV policing analyzed in the included studies were criminalization of survivors (n=12), various measures of survivors’ experience or satisfaction with police (n=10), population-level measures of IPV or other domestic violence homicide rates (n=8), willingness to or likelihood of reporting IPV to police (n=4), and population-level measures of IPV victimization other than homicide rates (n=3).

Racial analysis.

The majority of studies (69%, n=25) did not include an analysis of differences in the consequences of IPV policing by racialized groups, despite evidence of pronounced racialized disparities in impacts of IPV victimization and general exposure to policing (Caetano et al., 2005; Crutchfield et al., 2009, 2012; NYC Mayor’s Office to End Domestic and Gender-Based Violence, 2021). Only 11 studies included a racial analysis, which typically involved a stratified analysis or test for interaction in quantitative studies, or a thematic focus on impacts of IPV policing for specific racialized groups in qualitative studies.

1.3.3. Graphical summary and evidence gaps

Figure 2 provides a graphical summary of the studies by study design, exposure, and outcome, elucidating important patterns and gaps in the evidence base. First, we found that the studies evaluating population-level IPV victimization outcomes were focused on population-level exposures, such as warrantless arrest laws, police force size/staffing, and, in one case, sanctuary policies (i.e., policies that restrict cooperation between local law enforcement and federal immigration authorities). These studies all used longitudinal data.

Figure 2: Graphical summary of included articles by study design, exposure, and outcome.

³ Survivor experience with police includes challenges with identity work

As noted above, the most studied outcome of IPV policing was survivor criminalization, often dual arrest or survivor arrest specifically. However, many of these studies used cross-sectional data, limiting the ability for valid causal inference. There have been only two longitudinal studies focused on survivor criminalization as an adverse consequence of IPV policing, and in those studies the IPV policing exposure was (1) a specific coordinated community response involving increased IPV legal sanctions in two Georgia counties (Salazar et al., 2007) and (2) the triangulation of various population-level crime and violence data sources (Schwartz et al., 2009). There have been no longitudinal studies to date on the impact of warrantless arrest laws on survivor criminalization.

An additional notable gap in the evidence base is the lack of studies on police violence and child welfare involvement as outcomes of IPV policing. In a 2015 national survey of anti-violence service providers, advocates, and attorneys, one-third of respondents reported that police sometimes or often used inappropriate force against survivors of sexual assault and domestic violence and 89% of respondents reported that contact with police sometimes or often resulted in involvement with child protective services (Coker et al., 2015). Additionally, in two of the qualitative studies included in this review, study participants reported that police threatened to remove children from their home or reported actual investigation or involvement with child protective services (Dichter, 2013; Stephens & Sinden, 2000). These findings will be discussed further in the following section. Despite existing evidence that police violence and child welfare involvement are common consequences of IPV policing, no research to date on IPV policing has explicitly focused on these outcomes.

Finally, Figure 2 shows that there have only been two studies published on the generalized health consequences of IPV policing beyond IPV revictimization. Those two studies focused on youth mental and behavioral health outcomes (Dasgupta & Pacheco, 2018) and victim all-cause mortality (Sherman & Harris, 2015). This lack of research concerning health-related consequences of IPV policing may be reflective of the disciplinary focus of past research on criminology with little to no research in the public health field. There is a large body of evidence demonstrating the public health consequences of systems of criminalization and incarceration (Frank et al., 2013; Freudenberg, 2001; Kajeepeta et al., 2020, 2021; Weidner & Schultz, 2019). Therefore, it is likely that IPV policing practices and policies that lead to increased criminalization and incarceration would have generalized public health consequences, and such consequences should be explored further in the IPV literature.

1.3.4. Summary of findings

Quantitative studies grouped by outcome type

Survivor criminalization.

As mentioned above, the majority of research on survivor criminalization as an adverse consequence of IPV policing consists of cross-sectional studies of mandatory arrest laws. There is also one simulation study of mandatory arrest laws. These studies all provided evidence to suggest that mandatory arrest laws are associated with an increased risk of survivor criminalization, measured as higher rates of female arrests, victim arrests, and dual arrests (Durfee, 2012; Frye et al., 2007; Hirschel, et al., 2007a; Hirschel, et al., 2007b; Hovmand & Ford, 2009). A longitudinal study evaluating the effects of a coordinated community response to IPV with the goal of increasing legal sanctions against male abusive partners found that the intervention also increased the rate of female arrests (Salazar et al., 2007). A study of the impact of primary aggressor laws, which require police to determine the primary aggressor in an IPV dispute before making an arrest (Coffino, 2020), in mandatory arrest jurisdictions found that dual arrest was less likely than single arrest in states with primary aggressor statutes (OR=.452, p<.0001) (Hirschel & Deveau, 2017). However, a follow-up study using data covering many more years and jurisdictions, found that the enactment of primary aggressor laws had no significant effect on the risk of dual arrest (OR=.973, p=.550) (Hirschel et al., 2021). Finally, Schwartz et al. (2009) triangulated multiple data sources (e.g., arrests, convictions, victim-based reports of crime) to explore whether a narrowing of the gap between men’s and women’s violent crime arrest rates was driven either by women committing more violent offenses or by changes in policing practice/mobilization of law, such as through mandatory arrest laws. Findings strongly supported the latter theory of police widening the net of criminalization.

IPV and domestic violence homicide rates.

There have been seven studies using large, longitudinal datasets to assess the effect of warrantless arrest laws on rates of IPV homicide. With respect to mandatory arrest laws, four of these seven longitudinal studies demonstrated no significant effect of mandatory arrest laws on IPV homicide rates in most analytic models (Chin & Cunningham, 2019; Dasgupta & Pacheco, 2018; Zeoli et al., 2011; Zeoli & Webster, 2010). Two of those studies used an observational time-series approach and focused on large U.S. cities only (Zeoli et al., 2011; Zeoli & Webster, 2010) while the other two conducted national state-level difference-in-differences analyses (Chin & Cunningham, 2019; Dasgupta & Pacheco, 2018). In a longitudinal time-series study, Dugan (2002) found that mandatory arrest laws specifically for violations of orders of protection were associated with lower rates of White femicide but had no effect on Black femicide over the study period. The author additionally notes, however, that the effect on White femicide was not robust as it was dependent on the inclusion of specific cities in the dataset. In a difference-in-differences study, Iyengar (2009) found that mandatory arrest laws were in fact associated with an increase in the IPV homicide rate, however Chin and Cunningham (2019) identified potential coding errors in Iyengar’s analysis, which could not be verified. Discrepancies in the classification and recorded timing of mandatory arrest statutes may explain disparate findings (Zeoli et al., 2011).

With respect to exposures other than mandatory arrest laws, Chin and Cunningham (2019) found evidence that discretionary arrest laws are associated with a decrease in spousal IPV homicide rates relative to no warrantless arrest laws (effect for current spouse homicides: reduction of 0.1553 per 100,000; effect for former spouse homicides: reduction of 0.0321 per 100,000). Dugan et al. (2003) assessed associations between four relevant exposures (warrantless arrest laws, mandatory arrest laws, local pro-arrest policies, and specialized police training or staffing) and intimate partner homicide rates among 12 victim groups (defined by combinations of race, gender, and marital status). The authors found that state and local pro-arrest policies were associated with lower intimate partner homicide rates among multiple unmarried subgroups (Dugan et al., 2003). They also found that local police agencies with a specialized domestic violence unit or specialized training were associated with higher homicide rates for unmarried Black women (RR=1.359) (Dugan et al., 2003). Finally, Amuedo-Dorantes and Deza (2022) applied a difference-in-differences analysis to the study of sanctuary policies that restrict cooperation between local police and federal immigration authorities and found that sanctuary policies were associated with a decreased rate of domestic violence homicides with a Hispanic female victim (reduction of 0.007 homicides per 100,000). They also found that the protective effect of sanctuary policies dissipated when the state had a mandatory arrest law (Amuedo-Dorantes & Deza, 2022). The authors found no effect of sanctuary policies on the domestic violence homicide rates of non-Hispanic White victims.

Other IPV and domestic violence victimization rates.

Three longitudinal studies using national data focused on how mandatory arrest laws and other measures of IPV policing impact rates of violent victimization other than homicide (Cook & Taylor, 2019; Dugan, 2003; Xie et al., 2012) and the findings were somewhat inconsistent. Xie et al. (2012) found that mandatory arrest laws had no effect on IPV victimization rates among women. Meanwhile, Dugan (2003) found that mandatory arrest laws that specifically require arrest for violations of orders of protection were associated with lower odds of spousal violence (OR=.885) but not family or dating violence. When assessing interaction by county unemployment rates, Cook and Taylor (2019) found that mandatory arrest was associated with lower rates of intimidation at unemployment rates roughly between 4.75% and 8.5%, but not at lower or higher rates of unemployment. The authors found no effect of mandatory arrest laws on assault rates across any level of unemployment.

Xie et al.’s (2012) multilevel analysis used data from 40 metropolitan statistical areas over a 16-year period (1989–2004). Cook and Taylor’s (2019) analysis was conducted at the county level over a more recent 16-year period (2000–2015); however, the authors did not specify how many counties were included in the analysis. Dugan’s (2003) analysis was a national household-level analysis over a 6.5-year period (January 1992-June 1998) using data from over 500,000 households, however, the author did not account for clustering by household given the repeated measures over time. Importantly, Cook and Taylor (2019) measured the outcome of IPV victimization using crime data collected and reported by the Federal Bureau of Investigation (FBI), thereby only representing victimization reported to police. This could bias the findings given that mandatory arrest laws may result in underreporting of IPV to police (as discussed below). Xie et al. (2012) and Dugan (2003) used data from the National Crime Victimization Survey, which is still subject to reporting bias but is expected to be a more valid measure of victimization than crime data given that it includes information on crimes both reported and not reported to police (Dugan, 2003). Xie et al. (2012) additionally found that metropolitan statistical areas with more police officers per capita and more social service workers per capita had lower rates of IPV victimization (β=−.322 and −.197, respectively).

Survivors’ experience with police.

The evidence base on IPV survivors’ experience with police is heterogenous and lacks consistency across constructs and measures (Brame et al., 2015; Miller, 2003; Wilson & Jasinski, 2004; Yegidis & Renzy, 1994). The literature includes two randomized trials, one testing the effects of individual arrest of the abusive partner (Miller, 2003) and the other testing the effects of proactive enforcement of no-contact orders (Brame et al., 2015). The first randomized trial showed that abusive partner arrest was negatively associated with survivor perception of legal power (i.e., perceived agency in criminal legal decisions) (β=−.102), positively associated with perception of safety (β=.327), and not associated with perception of personal power (i.e., perceived control over economic and social resources) (β=.006) (Miller, 2003). These outcomes were all assessed shortly after the police intervention (Miller, 2003). The second trial showed that proactive enforcement of no-contact orders—involving proactive contact between police and survivors to inform survivors of the requirements of the orders, advise them on how to collect evidence in case of a violation, and monitor compliance—was not associated with likelihood of partner recidivism according to crime records. Based on victim interviews, proactive enforcement of no-contact orders was associated with higher levels of reported stalking/threatening behaviors and experiences of physical aggression (Brame et al., 2015). The authors argue that this “stalker effect” may be due to the survivor’s increased awareness of what constitutes stalking/threatening behaviors after receiving guidance from law enforcement. Consistent and validated measures of survivor experiences with and attitudes toward IPV policing are needed.

Reporting of IPV to police.

Four studies evaluated the rate of reporting of IPV to police as a consequence of IPV policing. One used nationally representative data to assess the association between mandatory arrest laws for violation of protection orders and informing police of IPV and demonstrated that mandatory arrest laws for protection order violations were associated with lower odds of informing police (OR=.875) (Dugan, 2003). Two of the four studies used victim interview data from the Spouse Assault Replication Program randomized trial in Miami-Dade County (Hickman & Simpson, 2003; Miller, 2003). One study demonstrated no association between the abusive partner being randomly assigned to arrest and the survivor’s willingness to call police in the future (Miller, 2003) and the other found that survivors receiving their preferred arrest outcome (arresting the abusive partner vs. not arresting) were more likely to report future victimization to police (OR=1.794) (Hickman & Simpson, 2003). Finally, a cross-sectional study of women in domestic violence shelters in a Midwestern state found that survivors who supported mandatory arrest policies had higher odds of reporting IPV to police (OR=1.692), though this association may be the result of reverse causation given the use of cross-sectional data (Novisky & Peralta, 2015).

Other outcomes.

Two studies evaluated other public health outcomes, namely victim all-cause mortality and youth mental and behavioral health outcomes (Dasgupta & Pacheco, 2018; Sherman & Harris, 2015). The first was a 23-year follow-up of the Spouse Assault Replication Program randomized trial in Milwaukee (Sherman & Harris, 2015). The authors found the rate of all-cause mortality among IPV victims whose partners had been randomly assigned to arrest was 64% higher than the rate among victims whose partners were not assigned to arrest. The second study included a multi-level analysis of warrantless arrest laws and five youth mental and behavioral health outcomes (Dasgupta & Pacheco, 2018). The authors found no effect of warrantless arrest laws on four of the five outcomes; warrantless arrest laws were associated with a 5.1% reduction in the probability of youth suicidal ideation. In gender-stratified analysis, this effect was present among girls but not boys.

Qualitative studies

Four of the nine qualitative studies included in this review focused on the experience of survivors who were wrongfully arrested or the challenges survivors face in successfully presenting themselves to police as the person experiencing violence, referred to as “identity work,” under mandatory arrest policies (Dichter, 2013; Leisenring, 2011; Li et al., 2015; Rajah et al., 2006). The studies detailed the psychological and emotional trauma of being wrongfully arrested as well as the material consequences of arrest such as incurred expenses, loss of employment, involvement from child protective services, and loss of child custody (Dichter, 2013; Leisenring, 2011; Li et al., 2015; Rajah et al., 2006). In two of the studies, some participants reported that one positive consequence of the arrest was that it served as a catalyst for them to leave the abusive relationship (Dichter, 2013; Li et al., 2015). The other qualitative studies primarily focused on survivors’ experiences and interactions with police (Bui, 2003; Erez & Belknap, 1998; Gezinski, 2020; Shim & Hwang, 2005; Stephens & Sinden, 2000). Common themes included: 1) inadequate police response involving victim-blaming and verbally abusive language; 2) hesitancy toward calling police because survivors relied on the abusive partner financially or feared retaliation from family or the abusive partner; 3) lack of autonomy and choice; and 4) survivors having difficulty navigating the criminal legal system and feeling that the abusive partner was better able to manipulate the system in their favor.

Together, the authors of these nine qualitative studies drew a number of shared conclusions based on their findings. First, the experiences of IPV survivors and their help-seeking behaviors are complex, shaped by both individual factors as well as structural and cultural forces (Bui, 2003; Li et al., 2015; Shim & Hwang, 2005). Second, authors identified adverse consequences of mandatory arrest policies for survivors, including failed identity work, criminalization, and arrest (Dichter, 2013; Gezinski, 2020; Leisenring, 2011; Rajah et al., 2006). Third, there is an urgent need for new emergency intervention services for IPV reduction and prevention that include alternatives to arrest (Dichter, 2013; Erez & Belknap, 1998; Gezinski, 2020; Li et al., 2015). Finally, authors concluded that survivors often wish to retain some autonomy in their attempts to stop the experience of abuse and that IPV interventions must be informed by survivor voices (Erez & Belknap, 1998; Stephens & Sinden, 2000).

Studies with a racial analysis

As mentioned above, 11 studies included a racial analysis of the consequences of IPV policing. Three of these 11 studies were qualitative, assessing how the experience of police contact or survivor arrest varied for specific racial/ethnic groups and how survivors of color faced unique challenges (Bui, 2003; Rajah et al., 2006; Shim & Hwang, 2005). One qualitative study was focused on IPV in Vietnamese American communities (Bui, 2003) and another was focused on IPV in Korean American communities (Shim & Hwang, 2005). The authors identified some shared themes concerning unique barriers to help-seeking that immigrant survivors experience including language barriers, economic dependency, social and cultural isolation, and a reliance on abusive partners for immigration status. In the third qualitative study with a racial analysis, Rajah et al. (2006) discussed that racism influenced how police treated Black women survivors in New York including survivor accounts of verbal abuse and physical roughness by police.

The majority of the quantitative studies that included a racial analysis assessed if the effect of IPV policing on survivor outcomes differed among Black survivors compared with White survivors (Dugan, 2002; Dugan et al., 2003; Iyengar, 2009; Sherman & Harris, 2015; Xie et al., 2012; Yegidis & Renzy, 1994). Three studies found that the adverse effects of IPV policing or mandatory arrest laws were more pronounced among Black survivors than White survivors (Iyengar, 2009; Sherman & Harris, 2015; Yegidis & Renzy, 1994). For example, in Iyengar’s (2009) study, the effect of mandatory arrest laws on intimate partner homicides was an increase of 0.621 per 100,000 among Black couples compared to 0.508 per 100,000 among White couples and 0.014 per 100,000 among Asian couples. In Sherman and Harris’ (2015) study, the adverse effect of arrest on victim all-cause mortality was largely explained by the effect among Black participants: arrest increased mortality by 98% among Black victims compared to 9% among White victims. Additionally, as discussed above, one study found that mandatory arrest laws were associated with lower rates of White femicide but had no such protective effect on Black femicide (Dugan, 2002). Two studies found no significant differences in the effect of state mandatory arrest laws on homicide rates or victimization rates by race, respectively (Dugan et al., 2003; Xie et al., 2012).

1.4. Discussion

In this scoping review, we summarized the evidence base concerning the consequences of IPV policing in the U.S. that extend beyond individual revictimization. We found that survivor criminalization is the most studied consequence of IPV policing and existing studies have documented a positive association between mandatory arrest laws and risk of survivor arrest. However, the evidence is primarily limited to cross-sectional studies; additional longitudinal studies would strengthen understanding of the causal nature of this relationship. We also found that there have been numerous rigorous studies of the effects of mandatory or other warrantless arrest laws on population-level measures of IPV victimization, including IPV homicide rates. Conclusions cannot be drawn about whether mandatory arrest laws reduce population-level rates of IPV victimization given mostly null findings and study limitations. An additional key finding is that there is a dearth of research on police violence, child protective services involvement, and health outcomes as potential consequences of IPV policing. Finally, only 31% of studies (n=11) included a racial analysis of the consequences of IPV policing. Of the studies that assessed differences among Black and white survivors, some found that Black survivors experienced more pronounced adverse outcomes or no protective outcomes compared to White survivors and others found no significant differences by race.

This scoping review provides the first-ever overview of existing evidence on the potential generalized consequences of IPV policing beyond individual IPV revictimization in the U.S., which allows for a more informed and nuanced calculus concerning the costs and benefits of the police-centric response to IPV, and specifically mandatory arrest laws. The present review demonstrates that (1) there are studies that have identified generalized consequences of IPV policing beyond individual revictimization, such as studies identifying a relationship between mandatory arrest laws and the risk of survivor arrest, and (2) there are studies that have documented more pronounced generalized consequences of IPV policing among Black survivors. Policymakers should consider potential generalized consequences beyond individual revictimization when evaluating the effectiveness of the current police-centric response to IPV and weighing it against potential alternative responses such as non-policing specialized emergency responders and investments in emergency and long-term housing and survivor grant programs.

The present review also highlights gaps in the evidence base. Specifically, additional research is needed on potential consequences of IPV policing such as police violence against survivors, child protective services involvement, and measures of the psychosocial and physical health of survivors. To date, the study of the police-centric response to IPV has primarily been the domain of criminology. Applying a public health perspective to the study of the police-centric response to IPV demands consideration of generalized consequences, expanding focus from narrow, short-term measures of recidivism and revictimization to broader measures of overall survivor health and well-being.

1.4.1. Limitations of the review and individual studies

There are important limitations of the present scoping review. First, the existing research on the consequences of IPV policing is quite heterogenous making exhaustive synthesis of findings difficult. For example, numerous themes were interrogated in the qualitative research, all of which could not be summarized in depth here. Additionally, many of the quantitative studies used different constructs, measures, and classifications for common exposure and outcome variables, making quantitative synthesis of findings impractical. Most notably, efforts should be made to reach consensus on the classification of warrantless arrest statutes.

Second, this review only included peer-reviewed articles published in scholarly journals. There is an extensive grey literature, including non-peer reviewed publications from the U.S. Department of Justice and the American Civil Liberties Union, as well as case studies collected by IPV service provider organizations that offer valuable insight concerning the consequences of IPV policing (Coker et al., 2015; Goodmark, 2022). Those contributions to the evidence base are not covered in this scoping review.

Finally, it is important to note some general limitations of the evidence base. Many studies used robust methods to test the effects of warrantless arrest laws, such as using a difference-in-differences approach to take advantage of the differential timing at which states passed laws and including a number of potential confounders that may have influenced both passage of the laws and violent victimization rates. However, many studies did not apply a causal inference framework to their analysis and either did not include important confounders that may result in a spurious relationship, or included a large number of predictors in their models without theoretical justification for their inclusion. Additionally, many studies used administrative data sources, such as the National Incident Based Reporting System (NIBRS) or the Supplementary Homicide Report (SHR), and did not evaluate the potential sources of selection bias and information bias that may result from the use of these data sources or discuss strategies to mitigate such sources of bias. These sources of administrative crime data generally only reflect IPV incidents that are reported to police and rely on good faith, consistent reporting from law enforcement agencies. The field of IPV research would benefit from increased scope and accessibility of administrative data sources beyond FBI data sources, such as the National Crime Victimization Survey, as well as investment in local primary data collection efforts from survivor cohorts. Lastly, it is important to note that a corrigendum was issued in reference to Iyengar’s 2009 publication due to a potential but unverified coding error (Iyengar, 2019).

1.4.2. Conclusions

Given the growing prevalence of severe IPV in the U.S. and increased reexamination of our reliance on policing and prisons as primary responses to many social and public health problems, there is an urgent need to critically evaluate the current police-centric response to IPV, including the widespread use of mandatory arrest laws. Such carceral policies have established societal costs, given the collateral consequences of mass criminalization and incarceration, and their disparate impacts on Black communities. To date, debate concerning the effectiveness of the policing response to IPV and mandatory arrest laws has primarily focused on its specific deterrent effect in the prevention of short-term individual-level revictimization. More research is needed on the impacts of IPV policing beyond individual revictimization and how these impacts may vary by survivor race. This emerging evidence base should be considered when evaluating our nation’s response to IPV.

Highlights.

• This is the first review of evidence on generalized consequences of IPV policing in the U.S.

• Survivor criminalization was the most studied consequence of IPV policing.

• There are also numerous studies on the effect of mandatory arrest laws on population-level IPV victimization.

• There is a need for more research on other potential consequences of IPV policing.