Abstract
New York State law mandates specific intimate partner violence (IPV) documentation under all circumstances meeting the enumerated relationship and crime criteria at the scene of a domestic dispute. Law enforcement compliance with this mandate is unknown. We reviewed law enforcement completion rates of Domestic Violence Incident Reports (DVIRs) and assessed correlations with individual or legal factors. Law enforcement officers filed DVIRs in 54% of the cases (n = 191), more often when injury occurred (p < .01) and the defendant had prior court contact (p < .05). The discussion explores policy implications and potential means to rectify the gap between mandated processes and implementation.
Recognition of the high prevalence of intimate partner violence (IPV) has led to widespread public health initiatives, updated medical practices (Eisenstat & Bancroft, 1999), and innovations in the civil and criminal legal arenas to address the problem. Coordinated responses to IPV include, but are not limited to, the justice system, the health care system, social services, and the education system.
The economic toll of IPV on communities and victims has been well documented. The Centers for Disease Control (CDC) estimates the yearly national costs at US$5.8 billion. Figures released in 2005 estimate the cost of IPV for medical care, mental health services, lost time from work due to injury and death at US$8.3 billion (Max, Rice, Finklestein, Bardwell, & Leadbetter, 2004). However, both of these estimates exclude the costs associated with policing, prosecution, defense, and incarceration of cases that come into contact with civil and criminal courts.
Since the early 1980s, the police have been modifying their IPV response policies after research and successful civil suits unveiled early concerns that police were reluctant to get involved in domestic disputes (Epstein, 1999). After increased attention to this problem, from both the legal and research communities, mandatory arrest policies were initiated, tested, and replicated in a number of communities with mixed reviews (Maxwell, Garner, & Fagan, 2001; Sherman, 1992; Sherman & Berk, 1984). Many states require that officers complete a mandatory reporting form specific for domestic violence calls for service. Attorneys may use these forms to assess whether the police collected evidence, who was present at the scene, and whether there were any injuries. Whether the victim had injuries may predict health care service use, which in turn may lead to additional evidence and possible medically related witnesses.
This study examines law enforcement compliance with completion of a New York State mandated Domestic Violence Incident Report (DVIR) form. In addition, the study examined whether victim injury predicts compliance and if compliance affects the perpetrator’s conviction.
Current Study
We know little about law enforcement compliance with mandated IPV reporting requirements. Under many state laws, police complete required paperwork at the scene of an IPV investigation. This document may be the standardized police reporting form or a special IPV supplemental form. In New York State, legislation was enacted in 1994 requiring the police to complete the DVIR at the scene of all IPV investigations. The DVIR provides space to record the victim’s narrative, any perpetrator admissions, injuries, and whether medical care was called to the scene of the incident, N.Y.S.C.P.L. §140.10 (4)(c). Although we confirmed whether law enforcement filed a DVIR by reviewing documents at the police department, many prosecutors’ offices do not have resources to conduct such reviews. The DVIR has the potential to provide essential information to the prosecutor about what evidence exists, where to obtain it, and what, if any, statements may need to be entered at trial. Whether these mandated forms are in fact being completed and provided to prosecutors has not been studied in depth. An earlier study demonstrated that police complete the DVIR in nearly 50% of IPV police calls for service (Cerulli & Tomkins, 2002). If prosecutors need to rely on medical documentation, they must first know whether there has been an injury, and whether and where medical care was provided. DVIRs are critical in holding perpetrators accountable.
This study reviews a subset of IPV cases adjudicated by a Western New York prosecutor’s office. Our first hypothesis is that law enforcement compliance with completing the DVIR is significantly correlated with whether there is a recognized injury at the scene and subsequent medical care. The presence or absence of an injury plays a key role in the immediate victim services needed at the scene, as well as the level of the crime charged and subsequent punishment should a conviction be secured. A second exploratory hypothesis is whether the filing of a DVIR affects the level of case disposition. Independent variables include victim and perpetrator sociodemographic information (gender, relationship, age), injury, use of a weapon, and medical documentation. The dependent variable is whether law enforcement completed the DVIR.
Method
Researchers reviewed IPV cases adjudicated between October 1, 2002, and October 31, 2003. During the 13-month period, the prosecutor handled a total of 3,337 cases in a specialized domestic violence and sexual assault unit. We reviewed all case files (1,141; 34.2%) resulting in a conviction or guilty plea. Only closed cases were examined. Under state law, dismissed and sealed cases were unable to be reviewed without court order. The study was conducted 6 years after the statute was implemented. Police had all participated in mandatory trainings offered by the district attorney. There have been no statutory changes at a state level or policy changes at a local level that alter reporting requirements. The definition of “when” forms must be completed has broadened and is now consistent with the department’s internal policy at the time of data collection. The state office responsible for assisting with statewide efforts created an interdisciplinary team that included law enforcement, prosecutors, and advocates to make the form more user-friendly. Ideally, the completion of such forms at the scene of calls for service might assist individual citizens, via the adjudication of these cases, and at a community level, to evaluate the success of efforts to ameliorate violence as measured by case volumes.
Based on the file reviews, we included all cases in which the victim was an older adult (older than the age of 60), a teenager (12-18 years of age), Hispanic, or disabled, as these victims are largely underrepresented in the literature (n = 97). We then drew a random sample of 94 of the remaining cases in which violence occurred between intimate partners as defined by those married, formerly married, dating, or having a child in common. Same-sex couples were included. Our total sample was composed of 191 IPV cases. There is no definition of “domestic violence” per se in the statute, nor an enumerated crime titled “domestic violence.” Rather, in New York State, these cases are charged with the same statutory provisions utilized for stranger or acquaintance crimes: assault, stalking, harassment, disorderly conduct, aggravated harassment. It is the party relationship that defines the crime as an IPV offense.
In a retrospective chart review, we recorded the variables of interest on a coding sheet on reviewing the DVIR, which law enforcement completed at the scene of the offense. The DVIR contained much of the demographic information about the victim and the defendant, as well as case-level variables. For files missing this form, researchers reviewed supplemental crime reports on file with the police department to complete the coding sheet. The university institutional review board (IRB) approved all aspects of this study.
Analysis
In New York State, a violation-level conviction is essentially equivalent to a traffic infraction, resulting in a fine and, at the most, 15 days in the local correctional facility. Felony and misdemeanor convictions are more serious, as they can result in more sanctions, including longer incarceration periods, fines, and probation supervision. Felony convictions were merged with misdemeanors as there were only six of the former. Race for “other” was merged with “Black” as well given the small sample size.
For the first hypothesis, we conducted a descriptive analysis to assess sample sociodemographics and chi-square to assess for statistically significant correlations between the independent variables, perpetrator’s age, gender, race, prior contact with the district attorney’s office, use of a weapon, level of charge at arrest, and injury requiring medical treatment, respectively, with the presence of a DVIR.
For the secondary exploratory hypothesis, we utilized logistic regression analysis (Hosmer & Lemeshow, 2000) to assess conviction level, felony/misdemeanor versus violation. Goodness of fit was assessed with the Hosmer–Lemeshow statistic (Hosmer & Lemeshow, 2000). We converted two of the variables into one variable, injury (yes/no) and medical treatment (yes/no), to create and analyze a three-category variable to assess the seriousness of the injury: no injury, injury without treatment, and injury with treatment. As victim and offender gender and race were highly correlated (both at p < .001), only perpetrator data were used in the final analyses. The final model is provided below.
Findings
Characteristics of subjects are described in Table 1. Victims were predominantly female and perpetrators were predominantly male. They were of similar age and racial background. Approximately half of victims and perpetrators had prior contact with the court, thus also with this prosecuting attorney’s office. All but one of the victims’ prior contacts had been as a victim. All the perpetrators’ prior contacts had been as offenders and 40% of those contacts involved the same victim. Despite this, there was one stalking charge to address multiple offenses against the same partner. New York was the last state to pass a broad stalking law, which was in effect at the time of this study. Based on findings from this review, the district attorney’s office offered training to improve implementation of this law.
Sample Description.
Note. Ethnicity is not reported due to the large number of missing cases.
Because criminal court systems do not routinely record race, 76 cases are missing that information.
Three cases did not have this information in the file.
New York State Penal Law mandates DVIRs for all police calls for service when a family offense has occurred. While the definition of “family” under New York law is strictly defined (married, formerly married, related by blood or marriage), the local police department has employed an expanded definition of family that includes all those subjects included in the present study; according to the mandate, 100% of the cases should have the DVIR. However, our review showed that out of 191 cases, police only completed the DVIR in 103 (54%) cases, consistent with earlier research noted above (Cerulli & Tomkins, 2002).
In light of evidence that some police do not file DVIRs in accordance with the statutory requirement, it became essential to understand when and why police do complete and file them. We next explored whether police are more or less likely to file a DVIR when an injury has occurred, when medical treatment was provided, when the perpetrator had a weapon, or when the level of offense was high (see Table 2 for details). Chi-square analyses revealed there were no statistically significant correlations between whether the police filed a DVIR and when the perpetrator had used a weapon, whether emergency medical professionals provided medical treatment at the scene or when the level of the crime was high. However, there was a statistically significant relationship between completion of the DVIR and both an injury occurring (p < .01) and the defendant having prior contact with the court (p < .05).
Case Circumstances and Prevalence of Domestic Violence Incident Report.
Note. DVIR = Domestic Violence Incident Report.
Fisher’s Exact Test with 2df.
In New York State, the Penal Law classifies a violation as a charge with the potential sentence of 15 or less days in jail (e.g., harassments, disorderly conduct), a misdemeanor as a charge receiving no more than 1 year in jail (e.g., assault 3rd, menacing, etc.), and a felony as carrying the potential for more than 1 year incarceration at a state correctional facility as opposed to a county-run facility.
Along with an increase in filing of DVIRs when an injury was present, police were more likely to file a DVIR when the crime was an assault. Police were less likely to file DVIRs for criminal contempt, although earlier violations may affect the degree of the offense. DVIRs were less likely to be filed for property crimes, and equally as likely for harassment charges (see Table 3 for more). After understanding the nature of the charge, we utilized logistic regression to assess what variables discussed above increased the odds ratio of a conviction for a felony/misdemeanor versus a violation, including the filing of a DVIR.
Type of Charge and Prevalence of Domestic Violence Incident Report Filed.
Note. Charges that did not involve a crime to a person or crime to property were deleted including possession of drugs or illegal possession of weapons. DVIR = Domestic Violence Incident Report.
In both analyses shown in Table 4, two variables were statistically associated with a conviction: the defendant’s age and whether the defendant had prior contact with the district attorney’s office as a defendant. In the unadjusted analysis, prior contact with the court was associated with (odds ratio, 95% confidence interval) a 3.61 [1.97, 6.61] times greater likelihood of having a felony or misdemeanor conviction for the index event as opposed to a violation conviction. The odds associated with a felony conviction or misdemeanor conviction with prior contact in the multivariate model was similar (3.37 [1.72, 6.61]). In the unadjusted analysis, younger age was associated with receiving a felony or misdemeanor conviction. However, after adjusting for covariates, older perpetrators are more likely to be convicted of a felony or misdemeanor, suggesting that a 1-year increase in age is associated with a 5.4% [1.9%, 8.9%] greater likelihood of a felony or misdemeanor conviction. The other variables were not statistically associated with the greater likelihood of a felony or misdemeanor conviction.
Results of Univariate and Multivariate Logistic Regression Analyses.
Note. OR = odds ratio; CI = confidence interval.
Discussion
According to the statutory mandate, 100% of the cases reviewed should have had a DVIR filed. Unfortunately, findings demonstrate that only 54% of the cases had these forms on file with the District Attorney. There is a gap between mandated policy and implementation. We find support for the hypothesis that police more often file DVIRs when an injury occurs. In addition, police more often file DVIRs when the defendant had prior contact with the court, which may become known at the scene of the arrest via the victim, police familiarity with the perpetrator, or the presence of an earlier issued protection order, which is electronically kept on a statewide level.
Although there is a state mandate for police to file DVIRs, findings reveal that whether the police file these forms is not associated with the level of conviction. There was no support for the second hypothesis that DVIRs were related to conviction level. Rather, we learned that the defendant’s prior contact with the system and age are associated with the level of charge at conviction. Findings also revealed that use of a weapon and charge at arrest are not associated with the final conviction. As is common practice, many district attorneys consider prior record when determining whether to offer a reduced charge in a plea negotiation. Because there is often a reduction in the charge, weapon use no longer appears linked to a particular outcome.
This study occurred in a district where police are trained to observe and document injuries at the scene. However, they are not necessarily trained to follow a case through medical treatment to further document injuries and treatment postarrest. If police file DVIRs only when a victim’s injuries are apparent at the scene, they risk failing to document a victim’s potential hidden injuries, such as internal injuries or the evolution of bruising evidence. Victims may seek treatment later, and/or discuss IPV medical and safety concerns with their physician. Better coordination between police and medical providers to document medical treatment given following the violent incident, either by a hospital or primary care provider, may be beneficial. In the absence of a police-conducted investigation to compile all potential evidence, prosecutors should inquire whether an injury was identified after the police call for service.
Rather than view victims in separate silos (police call for service, medical care provider’s office, court) we must begin to view a domestic violence call for service and subsequent response as a continuum. These interrelated encounters for the victim with helping professionals must be continually documented to facilitate perpetrator accountability. Medical professionals are encouraged and often mandated to educate their patients about their legal rights in the setting of IPV. Similarly, a coordinated community response to address IPV should attempt to connect IPV survivors with health care resources when law enforcement officers respond to calls for domestic violence incidents; even if there are no immediately observable injuries, the victim could suffer injuries at a later time, or long-term emotional and psychological harm.
Efforts to identify barriers or misperceptions limiting DVIR completion and filing might improve future compliance. It is possible that despite statutory language to the contrary, police do not file such forms when the victim requests that no arrest is made. It is also possible that the victims may state a wish to pursue the matter in Family Court, obviating, from the police perspective, the need for the DVIR. However, even in Family Court, DVIRs can provide important details to assist the court officials in decision making around safety plans, custody, and visitation issues, and recommending concurrent criminal prosecution now permissible in many jurisdictions.
Despite evidence that the DVIR completion did not impact the level of conviction in this study, the instrument remains a valuable tool for prosecutors in criminal court and civil attorneys in Family Court. The DVIR contains important information that may inform the strategy of the case, how to reach a victim at a safe or alternate contact, and captures the event in the victim’s voice through the narrative portion of the document. Prosecutors rely heavily on the information contained in them to know what evidence to secure, including medical records and possibly medical providers’ testimony at trial. Although it may not affect the level of case, it may affect what is charged due to the injuries noted and/or the presence of a weapon. In addition, DVIRs may be used to track prevalence and severity of IPV at a community or state level. The accuracy of the reports is important beyond the immediate use in a criminal trial.
Limitations
These findings are based on a sample of cases prosecuted in New York with dispositions favorable to the state. Therefore, caution is warranted in applying the findings to cases outside the jurisdiction or those dismissed for failure to prosecute, lack of evidence, lack of victim signature, or procedural defects. Also, as this study was for a city-level court, more serious offenses may be transferred to a higher level court. However, these exploratory data provide insight into police utilization of state-mandated reporting forms, as well as the role of injury, and subsequent documented treatment, on convictions. These two issues appear clearly interwoven. The ideal study is a prospective cohort study in which researchers respond to the call for service with the police, independently document injury and treatment, and then follow the cases through final adjudication. These data do not indicate whether there was injury detected subsequent to the police call for service. It is possible that those victims who did not receive DVIRs suffered injuries, had medical treatment, and/or may have been involved in previous cases where other evidence, such as statements or admissions, may have been helpful to proving their cases.
Conclusion
Further research is needed to explore creating better linkages between police, medical providers, and the legal community to keep victims safe and hold perpetrators accountable. In our study, injury clearly affected police documentation, which in turn affected the prosecutor’s ability to identify and secure needed evidence. Forty-six percent of our sample cases did not have DVIRs provided to the prosecution. If the prosecutors are operating in the dark, they do not know what evidence to secure, and therefore they may alter the case strategy from one grounded in evidence-based prosecution to one relying on the victim. This undue burden on the victim, to face her abuser in court, may be avoidable if the original police call for service provides the prosecutor with all the needed and required paperwork. How to enforce compliance with this state mandate remains an unanswered question and merits further research. Such research, conducted in the confines of balancing victim safety, may require a mixed-method approach that includes observations, police interviews, and document reviews post-calls for service. By triangulating the data, we may be able to better understand low compliance with state mandates and assist efforts for improved implementation. It is also important to interview prosecutors to assess their perceptions on the usefulness of the DVIR and its utility in the courtroom. Ultimately, improved coordination between all advocates for IPV victims may improve the welfare of victims, the accountability of perpetrators, and resource utilization for law enforcement with reduction of community prevalence of IPV.
Footnotes
Acknowledgements
We wish to thank Christina Raimondi and Jennifer Thompson Stone for manuscript preparation, and Corey Nichols-Hadeed, JD, and Neil McLaughlin for their assistance with editing.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported in part by Grants T32MH18911 (E. D. Caine, PI), K01MH075965 (C. Cerulli, PI), and P20 MH071897 (E. D. Caine, PI).