1. Introduction
Intimate partner violence (IPV) is a social problem with serious, even potentially lethal consequences. Increased awareness of intimate partner violence in the United States has been accompanied by a decrease in overall rates of IPV reported. The national decrease in rates of intimate partner violence from 1993 to 2010 was estimated at 64% [
1]. Despite this trend, concerns remain. The first year findings from the National Intimate Partner & Sexual Violence Survey of the Centers for Disease Control (CDC) focus on intimate partner violence, sexual violence, and stalking in the United States, and indicate that these forms of violence are widespread. Indeed, one in four women and one in seven men indicated they have been subject to “severe physical violence” by an intimate partner [
2]. Economic costs of intimate partner violence in the United States have been estimated at around $12.6 billion annually [
3].
While ideas of what constitutes intimate partner violence vary, the common conceptualization of IPV is included in the recently re-authorized Violence against Women Act [
4] as: domestic violence, dating violence, sexual violence, and stalking. Data from criminal statistics and crime-related surveys typically support the notion that, in intimate partner violence, women are significantly more likely to be victims and less likely to be perpetrators. For example, the National Crime Victimization Survey, an ongoing survey since 1973 conducted by the United States Census Bureau indicates that from 1994 until 2010, about four in five victims of intimate partner violence were female [
1]. Similarly, the well-known Violence against Women Survey [
5] found that over 22% of women indicated they had been victims of intimate partner violence at some point in their lifetime, while this was true for only 7.4% of men.
The view that intimate partner violence is primarily a crime against women has been strongly contested by some who argue that research indicates “gender symmetry” in intimate partner violence perpetration [
6]. It has been suggested that male victims of IPV are less likely to report their victimization in crime surveys [
7]. Supporter of gender symmetry point primarily to community surveys that use a behavioral measure such as the Conflict Tactics Scales (CTS) [
8] to assess the occurrence of IPV more accurately. The CTS and the revised version, the CTS-2, have been considered important in introducing a standardized way of assessing violence in intimate relationships [
9], but have also been strongly criticized for simply quantifying when a violent act occurs (e.g., a slap) without the additional relevant information such as the motivation for the violent act, the context in which the violence occurs, and the meaning and consequences of the violent act for the parties involved [
10]. As a result, critics suggest that studies that rely on measures such as the CTS do not accurately reflect the occurrence and contextual elements of IPV.
2. Intervention in Intimate Partner Violence
Awareness regarding IPV’s prevalence in the United States has increased in recent years, yet controversy remains about the nature of intimate partner violence itself, how it presents, and most importantly, how to provide effective prevention and intervention [
11]. As Hanna ([
12], p. 1458) writes, “One of the most perplexing problems for the law has been how to curb men’s violence against their intimate partners”.
A primary focus of efforts to effectively intervene in intimate partner violence has been law enforcement response, even though law enforcement contact is not prompted in most IPV incidents. Research has suggested that 73% of women and 86.5% of men who were identified as being victims of intimate partner violence did not report their victimization to law enforcement [
13]. Regarding the social problem of IPV, the legal system cannot remedy all contributing factors, but it is an essential component to providing for victim safety and offender accountability [
12]. While intimate partner violence is considered to be one of the most under-reported crimes in the United States [
14], these calls are still the most common to be received by law enforcement [
15].
Decades ago, the general law enforcement response to intimate partner violence was to treat it as a private family matter and avoid intervention and arrest. While this response may have been reflective of the limited understanding many had of the occurrence of IPV at the time, it was also related to officers’ desire to avoid working with these types of cases, as well as limitations in arrest abilities in most states that did not allow for officers to make warrantless arrests in misdemeanor cases unless they had witnessed the crime. IPV acts have been, and continue to be primarily, considered misdemeanors [
16].
The work of activists and advocates since the 1970s has prompted a shift in the criminal justice response to IPV, with an increased focus on holding IPV offenders accountable for their use of violence. The role of the criminal justice system is strongly emphasized in efforts to ultimately reduce IPV, as Hart ([
17], p. 406) writes, “The use of criminal sanctions has been a critical, and indeed even a necessary, response to violence against women, of course”. This emphasis has led to pro-arrest and even mandatory arrest policies for IPV-related crimes [
18]. Maxwell, Garner, and Fagan [
19] examined over 4000 cases of domestic violence to explore the impact of arrest upon defendants’ likelihood to recidivate (reoffend), and found “consistent but modest” reductions in domestic violence recidivism, therefore providing limited support for policies that emphasized arrest. A recent meta-analysis using the National Crime Victimization Survey found that the odds of a victim being re-victimized were reduced by over 43% if her perpetrator was arrested [
20].
Similarly, past research has also yielded modest support for the deterrent effects of post-arrest interventions. Generally, research examining the impact of court disposition upon reducing recidivism in IPV offenders has been “meager” ([
21], p. 76). In one community study in Hamilton County, Ohio researchers reviewing the dispositions of 3662 IPV arrestees examined factors significantly associated with recidivism. Jail, with or without an accompanying fine, was not significantly related to a reduction in recidivism. Recidivism was significantly higher for those who did not have charges filed against them, while probation was associated with reduced recidivism [
21]. Modest significant effects in reduced recidivism were associated with participation in a batterer intervention program [
21]. Babcock, Green, and Robie [
22] also found small but significant effects regarding the impact of batterer intervention programs (BIPs), psycho-educational groups charged with teaching offenders about abuse and equality in relationships, upon recidivism. However, a recent review of over thirty studies of criminal sanctions beyond arrest, examined the research on the effects of prosecution, of conviction, and of sentencing. This review found that, most frequently, these sanctions were not statistically significant regarding their impact on the prevalence of reoffending [
19]. Regarding the question of the effects of sentencing, Klein, Centerbar, Keller, and Klein [
23] found in their study of IPV offenders who had both IPV and non-IPV cases that approximately 57% of the offenders were sentenced more severely for their IPV crimes than the non IPV crimes, and that these offenders, if prosecuted during the early years of their adult criminal offending, were significantly less likely to be arrested for subsequent IPV offending.
4. Gender of IPV Perpetrator and Law Enforcement Response
Coordinated community response efforts prompted lobbying for new laws to address how the criminal justice system treated IPV victims and offenders. This included the promotion of mandatory arrest laws in which law enforcement officers were required to arrest an IPV perpetration suspect if there was probable cause that an IPV offense occurred [
30]. While the enactment of pro-arrest and mandatory arrest policies appear to have led to an increase in arrest rates for IPV perpetration [
31], concerns have been raisedthat the most notable change has been the increase in the arrests of women in IPV situations [
32,
33], with the primary concern being that police officers are incorrectly identifying IPV victims as IPV perpetrators [
34]. The arrests of women include cases in which they are identified solely as the perpetrator of IPV in the incident, as well as “dual arrest” cases in which both parties are arrested either because the evidence supports that they are both at fault, or the responding officer cannot make a determination of fault [
35]. Support for these concerns can be found in research, which indicates this increase in arrests of women for IPV perpetration, does not coincide with evidence of an increase in the use of violence by women [
36]. Thus far, the relationship between gender and arrest for IPV since the enactment of pro-arrest and mandatory arrest policies in the research literature has remained unsettled. Feder [
37] found that in a law enforcement jurisdiction with strong pro-arrest policies, arrests were more likely to be made in IPV-related calls for service than for non-IPV related calls, victims were most often female, but the gender of the identified IPV perpetrator was not significantly related to the decision to arrest. Ho [
38] found similar arrest rates for both female and male suspects of IPV perpetration (36.4% and 36% respectively). Based on national criminal justice statistics from 1998–2002, females were the identified IPV victims in 84% of the cases of spousal assaults and were the victims in 86% of cases of IPV where the assailant was their boyfriend or girlfriend [
39].
5. Age, Intimate Partner Violence, and Criminal Justice Response
Risk for IPV offending is often correlated with younger ages [
40], with the average range of ages for perpetrators identified as 18–35 years [
15]. According to the United States Bureau of Justice Statistics from 1998–2002, the average age of victims of reported intimate partner violence was 27 years for those assaulted by their boyfriend or girlfriend and 35 years for those assaulted by spouses. For reported IPV perpetrators during this same time frame, 73% of those who perpetrated against their spouse were age 30 years or older while only 36.6% of those who assaulted their girlfriend or boyfriend fell into this age range [
39]. There is general consensus in the research literature regarding the average age of IPV perpetrators, with the national mean identified as 31 years of age [
41].
6. Ethnicity/Race, Intimate Partner Violence, and Criminal Justice Response
There have been inconsistent findings in the literature regarding the relationship between race and the involvement of IPV perpetrators within the criminal justice system (for further discussion of this, see McClosky, Sitaker, Grisby, & Malloy, [
42]). Hirschel [
35] found that arrests in IPV cases are more likely when perpetrators are white. In his study regarding IPV perpetrators in Sacramento County, California, Kingsnorth [
43] found whites, as well as those of Asian or Pacific Islander descent to be underrepresented in the sample of perpetrators in relation to overall population of the county, while Hispanic perpetrators were proportional to the population. In the same study, African Americans were 35.5% of IPV perpetrators, while African Americans constituted only 10% of the overall population. The over-representation of men of color identified as perpetrators of IPV is not typically interpreted as indication that they are more likely to be involved in IPV, but rather as a result of other aspects of the criminal justice system such as police response and prosecutorial issues [
42]. Additionally, the effects of race and ethnicity may be mediated by other demographic factors such as age and income [
44].
7. Current Study
This study examines IPV-related incidents in one CCR community over a 14-year time period, focusing specifically on the outcomes of citations and findings or convictions. This study is unique due to the extent of the data collection and large sample size. It examined the relationship between gender and the outcomes of citation and conviction for IPV-related offenses. Citation includes both the arrest and lodging of the IPV perpetrator at the time of the initial response to the incident, as well as a citation (an arrest) after an investigation of the IPV-related crime. Conviction refers to cases in which the prosecution has filed formal charges against the identified IPV perpetrator, and the case has resulted in the finding of guilt by a judge or jury. Conviction also refers to cases in which the IPV perpetrator has accepted a plea deal. The researchers selected age and race in addition to gender to explore their relationship to citation and convictions of guilt including those pleading guilty.
7.3. Law Enforcement Protocol for Intimate Partner Violence
This jurisdiction is located in a state where state statutes and municipal laws allow for law enforcement officers’ discretion to arrest based on the specifics of the incident itself (such as imminent threat or an established protection order in place). The local police department however, has adopted policy that does specify mandatory arrest for IPV-related crimes when probable cause has been established. There may be cases, however, where officers may not believe they have sufficient evidence to “lodge” the perpetration suspect at the time of initial contact, and then later believe an arrest is warranted. In these cases, the IPV perpetrator may be cited for the IPV-related crime and released. Due to this department-specific policy and the manner in which the community data was collected, only incidents responded to by the local police department were included for analysis.
This sample includes investigations resulting in both misdemeanor and felony citations and incidents in which there was a single charge and multiple charges. In this jurisdiction the following charges are characterized as IPV-related: third degree assault (within the context of an intimate or formerly intimate relationship), domestic assault, violation of protection order (or harassment order), terroristic threats, strangulation, vandalism, child neglect or child abuse (when a child is exposed to domestic violence), criminal mischief and disturbing the peace.
8. Methods
Source of the Data
The data was collected by the local domestic violence coordinating council and originated from a variety of sources including: law enforcement reports, county corrections, the local county attorney’s office, the local courts, probation, and local batterer intervention programs. The data was entered into a computerized format by staff at the agency, and analyzed by the authors of this study.
The sample consisted of 23,614 investigations that were recorded. The perpetrator and the victim were identified for each incident as well as the ticket number, residential status, date of incident, time of incident, address at which the incident occurred, officer number of the responding police officer, age of victim and of perpetrator, race of victim and of perpetrator, indication of whether a citation was given, offenses related to each citation, disposition of each case, and fines and/or length of jail time received. The sample included all domestic violence investigations during the time frame in this jurisdiction, including investigations in which women were identified as perpetrators, men were identified as victims, same sex couples were involved, and incidents in which both parties were cited. Not all investigations lead to citations, so citation of an IPV crime was identified as our first outcome variable. We then examined, using the cited cases, what predictors were linked to a finding (or plea) of guilty.
The authors of this study then recoded some of the variables before running their analyses. Statistical analyses were conducted in Microsoft Excel and SPSS. Descriptive, bivariate statistics, and logistic regression were used.
9. Findings
Over three quarters of the incidents, victims in the sample were women (n = 18,448, 78%) and about three quarters of the perpetrators were men (n = 16,607, 70%). The mean age of victims at the time of the incident was 31 and the mean age of the perpetrators was 32. In the majority of the incidents, victims were White (n = 17,505, 74%) and the majority of perpetrators were also White (n = 12,747, 55%). However, as with many other communities regarding criminal justice data, there is an overrepresentation of racial/ethnic minority group members for both victims and perpetrators, with the exception of Asian victims and perpetrators.
Repeat perpetrators were identified and coded by name after their first incident in the 14-year time span. Of the 23,614 incidents, 47% (n = 9858) involved repeat perpetrators. Over 86% of incidents had males as repeat perpetrators compared to 14% of incidents with females as repeat perpetrators χ2 (1, N = 21,164) = 789.503, p < 0.000 indicating that repeat perpetrators overwhelmingly, tend to be males.
9.1. Predicting Citation for IPV Offenses
The summary of logistic regression analysis for variables predicting citation is presented in
Table 1. Although the results show a significant model, the variance explained is very small (3%) because of the large sample size and a small variance in the dependent variable itself—nearly 80% being cited. The odds ratio—Exp(
B) indicate a
small but significant effect of gender, race and age. Incidents in which perpetrators were males have higher odds of being cited over incidents in which females are perpetrators
(OR = 1.4). With regard to race, the effects vary by each race category: Incidents in which the perpetrator was identified as Black (
OR = 1.2), Hispanic (
OR = 1.4), and Native American (
OR = 1.3) have higher odds, although incidents with Asian perpetrators (
OR = 1.2) are not significantly higher. Age has a very small effect (
OR < 1.0)—the lower the age, the less likelihood of being cited.
Table 1.
Summary of logistic regression analysis for variables predicting citation (No = 0, Yes = 1, N = 23,614) and pleading (No = 0, Yes = 1, N = 16,619) controlling for gender, race and age of suspects.
Table 1.
Summary of logistic regression analysis for variables predicting citation (No = 0, Yes = 1, N = 23,614) and pleading (No = 0, Yes = 1, N = 16,619) controlling for gender, race and age of suspects.
| Citation | Pleading |
---|
Predictor | B | S.E. | eB (Odds Ratio) | B | S.E. | eB (Odds Ratio) |
---|
Gender (male = 1) | 0.319 *** | 0.04 | 1.376 | 0.693 *** | 0.04 | 1.999 |
Suspect Race | | | | | | |
Asian | 0.173 | 0.116 | 1.189 | −0.465 *** | 0.103 | 0.628 |
Black | 0.196 * | 0.042 | 1.217 | −0.032 | 0.038 | 0.969 |
Hispanic | 0.305 *** | 0.068 | 1.357 | −0.035 | 0.058 | 0.966 |
Native American | 0.278 ** | 0.094 | 1.321 | 0.297 *** | 0.085 | 1.345 |
Age | −0.01 *** | 0.002 | 0.99 | 0.001 | 0.002 | 1.001 |
Constant | 1.292 *** | 0.064 | 3.641 | −0.345 *** | 0.062 | 0.709 |
χ2 | 146.145 *** | | | 344.319 *** | | |
Df | 6 | | | 6 | | |
9.2. Predicting Plea/Conviction for IPV Offenses
By dividing the number of incidents that led to citation by those that were coded as having a “guilty” plea or finding, it was determined that there was a 55.6% conviction rate from the citations issued. However, this did not account for cases in which charges might be combined into one court case, or for other factors that might affect a final outcome of conviction. Therefore, the researchers removed these cases, which left them with a conviction rate of 73.3%.
The summary of logistic regression analysis for variables predicting a guilty plea or finding are presented in
Table 1. This analysis included all incidents in which the charges were filed and resulted in convictions. Findings of guilt, either through pleas or court findings in this jurisdiction, however differ from citations regarding relationship to the selected independent variables. Incidents in which the IPV perpetrators are males are twice as likely to result in a finding of guilt (
OR = 2.0) than those in which the identified perpetrator was female. While incidents in which perpetrators were identified as Black (
OR ≈ 1) and Hispanic (
OR ≈ 1) are no more likely to plead or be found guilty than Whites, and incidents in which perpetrators were identified as Asian were less likely than those with White perpetrators (
OR < 1). Only incidents in which Native Americans were identified as perpetrators were more likely (
OR = 1.35) to plead guilty. Age does not have an effect on plea.
10. Discussion & Conclusions
Professionals who study intimate partner violence are still attempting to determine the components of effective intervention. However, there is support for the coordinated community response model in responding to cases of IPV, in which various agencies and organizations work with law enforcement to provide for offender accountability and victim safety. This study examined the dataset of one community coordinated response which has maintained active efforts at collaboration, training and rapport-building since its inception.
Regarding its efforts to hold IPV perpetrators accountable, this community has a higher arrest rate (over 70% of incidents lead to citations) than the national average of 37% [
35]. The conviction rate for this CCR was found to be 73.3% of those cited, which is similar to what has been reported in other studies, but it is also important to note that conviction rates differ across jurisdictions [
45].
Findings of this study indicate that men are far more likely to use violence challenging the claim that women are as likely to use violence as much or more than men [
46]. This study supports the fact that men are far more likely to be cited. This is not surprising because there are studies which indicate that women are more likely to sustain criminally prosecutable injuries as a result of intimate partner violence [
47]. Indeed, a recent meta-analysis examining intimate partner homicide internationally indicates that women are six times more likely than men to be killed by their partner [
48]. IPV incidents in which there was a repeat perpetrator were more likely to have males as the perpetrator.
Race, with the exception of Asians, of IPV perpetrator, has a small but significant predictor of citation. While differences have been found incongruously in other samples related to IPV-related citations, this community coordinated response team has worked closely with law enforcement in effort to reduce the influence of race in IPV-citations. While age has been found to be a fairly consistent predictor in other studies of citation for IPV-related crimes, this study only found that those cited were only one year older than those who were not but age did not have an effect on pleading guilty. The average age of IPV perpetrators in our sample was consistent with national research, approximately 32 years of age.
Regarding guilty pleas, women were less likely to plead or be found guilty. One explanation for that is, this is a community that uses Pre-Trial Diversion as an option for offenders who do not have a criminal history, and this sample included incidents in which 86% of the IPV offenders were repeat offenders, and this was the case in only 14% of incidents with female offenders. It should be noted that even when sent to Pre-Trial Diversion, IPV perpetrators often incur many of the same consequences as those who are convicted, such as being required to attend a batterer intervention program.
Our findings indicate that incidents in which IPV perpetrators were identified as Black, Hispanic and White all had similar rates of pleading (or being found) guilty. This may be the result of the consistency in which the prosecutors and judges in this community address IPV related crimes. Most IPV offenders in this community are provided the opportunity to attend an approved batterer intervention program to avoid a jail sentence. The incidents in which the IPV perpetrators who were identified as Asian were actually less likely to plead guilty should be explored further. However, we cannot draw definite conclusions about the effectiveness of CCR because we do not have pre CCR data to make comparisons. The literature review did not provide us with an equivalent non-CCR community study to compare our results and draw definite conclusions about the effectiveness of the CCR model. Given the large data set, the statistics yielded by this study may provide other and future researchers with similar research questions parameters to compare and draw more definitive conclusions about the effectiveness of the CCR model.