Exploring Officer Arrest Discretion Following State Policy Changes in Intimate Partner Violence

Abstract

In 2015, South Carolina passed the Domestic Violence Reform Act giving officers increased arrest discretion and rescinding mandatory arrest laws for intimate partner violence (IPV). Analyses using incident reports (2016/2017) from South Carolina show that officers primarily rely on the presence of five variables to determine arrest: children, injury, property destruction, offender at scene, and weapons. Cases with Black female victims are more likely to result in arrest, suggesting that Black males are being criminalized. We conclude that officers are reluctant to use individual discretion and rely on a form of structured discretion provided by administrators that shapes local culture and decision-making.

Keywords

intimate partner violence officer discretion mandatory arrest policies

Introduction

Criminologists readily recognize that a certain amount of police discretion is involved with most arrests and, to varying degrees, discretion centers on extralegal variables. As intimate partner violence (IPV) fails to challenge public order and is traditionally private, many officers are reluctant to arrest offenders. Numerous studies have investigated the subject with diverse conclusions explaining police behavior (Bonner, 2015; Goldstein, 1963; Skogan & Frydl, 2004), but the factors affecting officers’ IPV arrest decisions are not well understood. Most research on police responses to IPV in the United States examine official reports from law enforcement agencies or victimization surveys that disproportionately sample people in large urban areas. Few studies analyze data extracted from responding police officers’ narratives in IPV incident reports as we do (Karlsson et al., 2018; Sorenson, 2017), but these more nuanced data sources may provide greater insight into what factors are important to officers when making arrest decisions.

This study examines officers’ incident reports for IPV in Aiken, South Carolina, to explore officers’ use of arrest discretion in a small city. These data are particularly well suited for continued research on arrest decisions due to South Carolina recently changing mandatory arrest laws to allow arrest decisions to be at the sole discretion of the responding officers. The timing of this change is ideal for research on local arrest decisions, as the city recently received national attention regarding alleged racial discrimination by law enforcement. Increased discretion resulting from the removal of mandatory arrest laws could allow for implicit biases of local officers to affect arrest decisions. Using smaller city data also allows researchers to investigate how local police culture affects decisions. The process of guided discretion, which provides officers some parameters to consider when deciding whether to make an arrest, is explored in this study. Finally, this study follows the lead of others that call for continued research on arrest in IPV cases to inform police training and policy (Tatum & Pence, 2015).

The desirableness of policies mandating arrest is mixed. Since the mid-1980s, many states in America changed to pro-arrest policies based on prior research showing a deterrent effect of an arrest, but research is ambiguous in determining whether mandatory arrest policies deter future violence and, in certain cases, whether it may increase violence (see Barlow & Walklate, 2018; Johnson & Goodlin-Fahncke, 2015; Robinson et al., 2018). Critics of mandatory arrest statutes emphasize that removing them may actually benefit low-income and minority women by providing a greater sense of agency in deciding whether to press charges as arrest may negatively affect victims financially and/or with regard to child custody (Barlow & Walklate, 2018).

There are also many legal and extralegal factors that influence officer discretion (Bonner, 2015; Durfee & Fetzer, 2016; Tatum & Pence, 2015). One key factor increasing the likelihood of arrest is the presence of injury (Bachman & Coker, 1995; Bonner, 2015; Robinson & Chandek, 2000), although the “arrest if there is an injury” guideline may become an ignored preference if officers are unsympathetic to the victim (Bonner, 2015). Other fairly well-established predictors of arrest are the presence of drugs and/or alcohol (Avakame & Fyfe, 2001; Durfee & Fetzer, 2016; Novak et al., 2002) and weapons (Durfee & Fetzer, 2016; Sorenson, 2017). The influence of minors’ presence at the scene is mixed, but is often included as a control variable (Buzawa & Austin, 1993; Hirschel & Buzawa, 2013; Hirschel et al., 2010; Poon et al., 2014; Tatum & Pence, 2015). Previous research also finds that the offender’s presence at the incident scene is a significant positive predictor of arrest (Buzawa et al., 1999; Feder, 1996). While offenders who remain on scene are more likely to reoffend and are more violent (Buzawa et al., 1999), it is possible that officers mistakenly believe that if the offender is off scene that the episode has deescalated, resulting in a decreased need to make an arrest (Tatum & Pence, 2015).

Less obvious but implicit mediating variables that affect officers’ arrest decisions are bias toward the victims’ and offenders’ race and sex. Several studies document that arrest is more likely in IPV incidents when there is a racial match other than a Black suspect and Black offender (Durfee & Fetzer, 2016; Hirschel et al., 2017; Lee et al., 2013). It appears counterintuitive that the demographic group typically at the largest risk of arrest has a lower likelihood of IPV arrest, but officers responding to incidents may regard the case less seriously if it is with a Black victim and Black offender due to negative stereotypes that the case is “ordinary” (Klinger, 1997; Lee et al., 2013). Victimization is historically rooted in stereotyped characteristics to defend slavery institutions that result in a paradigmatic victim being a White woman (Larance et al., 2019; Romain & Freiburger, 2016). Generally, harsher punishments are found in cases with Black offenders and White victims, as well as cases with White female victims (Paternoster, 1984; Williams et al., 2007). This results in a greater likelihood that racial and ethnic minority women are less likely to be seen as victims and their offenders are less likely to be arrested compared to White women (e.g., O’Neal et al., 2016).

The extent to which race and other extralegal variables influence officers’ decisions may be affected by mandatory arrest policies. Mandatory arrest policies may reduce officer discretion and the consideration of extralegal variables by limiting the role of implicit biases in officer decision-making (Eitle, 2005). Research is ambiguous regarding how mandatory arrest policies link to race. In states without mandatory arrest laws, some research finds Whites are more likely to be arrested than minorities (Eitle, 2005; Hirschel & Buzawa, 2013). Other studies report higher rates of arrest among minorities (Bachman & Coker, 1995; Novak et al., 2002). Black women are additionally more likely to report all types of violent crime in part due to Whites having greater economic resources that allow them access to private services (e.g., physicians, safe shelter) and consequently be less reliant on law enforcement (Bachman & Coker, 1995). Thus, higher rates of arrest may be influenced by greater contact with public safety and not because of implicit racial biases.

Gender and racial considerations may be powerful subconscious factors in determining arrest, but the effects of racial bias may vary by social geography and history. South Carolina is of particular interest in studying IPV because it has the sixth highest female homicide rate (Violence Policy Center, 2018) and continues to hold the ignominious distinction of one of the top 10 most dangerous states for women in the United States for the last two decades (Smith, 2018). In 2014, the Post and Courier highlighted this dangerous and often lethal trend of IPV and female homicide in a Pulitzer Prize series. This exposé of the state’s relative tolerance of domestic violence and IPV became the impetus in the passing of the 2015 gender-based policy, the Domestic Violence Reform Act (DVRA; Robins, 2017).

The DVRA significantly altered state statutes by allowing prosecutors to seek harsher punishments and giving officers more arrest discretion by rescinding mandatory arrest laws and the ability to charge first-time offenders with felonies.¹ While the DVRA contains attitudes encouraging stronger responses to IPV by the state, the removal of mandatory arrest laws also means that officers may make fewer arrests. This is particularly true in cases where an officer may not be certain one has enough evidence to establish probable cause. Under mandatory arrest, officers were required to make an arrest at the sign of any visible injury, and this removed some of the subjectivity involved in establishing probable cause (see Barlow & Walklate, 2018; Myhill & Johnson, 2016). As other research notes, officers with less experience, training, and knowledge may be less confident in discerning reasonable suspicion (Kinports, 2010) and may believe that a case needs to be successful at trial to justify arrest (Brown & Frank, 2005). In the absence of mandatory arrest laws, officer culture and the use of departmental guidelines to consider for arrest can help officers in determining whether to make an arrest.

Setting

This study focuses on officers’ use of arrest discretion for IPV in Aiken, South Carolina post-DVRA. Aiken is a midland city located about 20 miles from Georgia and has an approximate population of 31,000 people with almost 69% who identify as White and 27% as Black (U.S. Census Bureau, n.d.). It is a conservative city that recently experienced national attention due to discordant race relations when the Washington Post reported on police abuse in Aiken and included a video (Balko, 2016). In this video, officers stopped a Black couple for having a temporary paper license tag on a recently bought vehicle and then performed a controversial drug search allegedly exposing the woman’s breasts and executing a roadside cavity search on the man. While this is a singular incident, it highlights race and gender as potentially prejudicial factors of arrest in this city.

Given that the officers in this study rely on their own discretion, it is possible that implicit racial biases affect arrest decisions. As other research notes, discretion can liberate one to act on biases in cases where evidence is ambiguous (Devine et al., 2009; Quillian, 2006). The IPV cases examined in this study are intra-racial. If Black females are socially devalued, one might expect that cases with White victims would be more likely to result in a more severe punishment of arrest. Given the recent attention drawn to local policing regarding the treatment of Black persons, however, Black offenders suspected of a crime may be treated more severely resulting in more arrests with Black female victims. Conversely, the removal of mandatory arrest laws may result in officers arresting in cases only when they perceive they can establish strong evidence against the alleged offender, limiting the role of racial bias in arrest decisions.

Data and Method

The Aiken Department of Public Safety provided the researchers with 571 IPV incident reports that occurred through January 2016 and December 2017. The authors coded and hand-entered all values into SPSS. In addition to the standard demographic and incident type fields, each report includes a narrative option. While some fields in the incident report are forced entry fields (e.g., victim’s race), most are not. Often officers fail to report details in the standardized sections, but describe events in the narrative, which were coded. Accurate thematic coding occurs most often when data sources meet pre-defined criteria, are relatively structured, and describe similar experiences (Young & Casey, 2018), like officer incident report narratives. Occasional unclear incident reports were marked and later coded as a team to ensure consistency.

Dependent Variable Operationalization and Analytic Approach

The dependent variable is whether an arrest occurred as a direct result of IPV (1 = yes). We do not include arrests resulting from an outstanding warrant or other infraction. Our data include misdemeanor and felony cases of IPV and do not include sexual assaults or other types of family offenses or household disturbances. Because our dependent variable is dichotomous, logistic regression is used to predict odds of arrest.

Independent Variables Operationalization

Due to the literature’s description of the salience of race and sex surrounding perceptions of victimhood, we create interaction variables for race and sex. Our categories include Black female, White female (= reference), Black male, and White male. The racial composition of the data lack diversity but are similar to that of Aiken. Ninety-five percent of Aiken’s residents identify as Black (26.9%) or White (68.5%; U.S. Census Bureau, n.d.). The majority of cases were intra-racial, with 87% of cases with Black female victims having a Black offender and 90% of cases with White female victims having a White offender. Due to the small sample size and overall low arrest rate, 54 cases involving interracial couples are omitted as these interracial cases impact the findings (discussed in the following). Ten cases are omitted where the victim or offender identified as Hispanic, Asian, or “other” race/ethnicity, due to a lack of variation in the measure. Because of covariation, offenders’ characteristics are not included in the multivariate analysis.

Additional independent variables are based on the information that the Aiken police require dispatchers to indicate to the responding officers and the structure of the incident reports which officers complete. This information appears in the Aiken Department of Public Safety Policies and Procedures and is designed to assist responding officers in determining the best course of action. The structure of incident reports guides officers’ decisions in assessing the “risk” of current and future threat (see Robinson et al., 2018) and provides justification for an officer’s arrest decision (Barlow & Walklate, 2018). The format of the incident forms varied over the 2-year period, but some included a check box to indicate the presence of variables. It is noteworthy that officers did not always check the boxes on the incident forms and instead relied on the narrative section to provide that information.

Using incident reports’ fixed fields and narrative sections, the analysis focuses on variables that reflect the written guidelines provided to officers responding to IPV calls. These variables are dummy coded and include whether the suspect is present on scene; weapons are involved (e.g., blunt objects, gun, knife, screwdriver); the responding officer observes victim injuries (not simply the victim reported injuries); the offender is under the influence of drugs and/or alcohol; children are present. Each of these variables is coded as “present” (=1) if they are noted in the report (0 = “not present”). We similarly analyze and code the narratives to determine the cause of the incident. These codes include custody or childrearing issues; suspected infidelity; money issues; property destruction; and threats. Relationship type (1 = “married”; 0 = “unmarried”) is included as a control variable.

Consistent with other research (Karlsson et al., 2018; Muftić et al., 2015), incidents involving dual aggressors (n = 18) and same-sex couples (n = 7) are removed from the data due to the small number of cases which present methodological concerns. Cases where the victim was the city or a business entity (n = 15), cases where the primary victim was a juvenile or had an atypical/unclear relationship (e.g., power of attorney) (n = 8), and cases with missing demographic characteristics (n = 11) are also omitted. Listwise deletion resulted in 450 incidents.

Results

Table 1 presents descriptive statistics of the 450 police incident reports. It shows frequencies of race and gender interaction terms for offenders and victims and situational variables of the cases. Table 2 presents the results of the logistc regression of arrest on race and gender interaction variables and covariates that reflect officer discretion.

Table 1.

Descriptive Statistics for Arrest and Covariates (N = 450).

Variable	n	%	M	SD
Demographics
Black female offender	51	11.3	.113	.317
Black male offender	232	51.6	.516	.500
White male offender	130	28.9	.289	.454
White female offender	37	8.2	.082	.275
Black female victim	232	51.6	.516	.500
Black male victim	51	11.3	.113	.317
White female victim	130	28.9	.289	.454
White male victim	37	8.2	.082	.275
Married
Yes	103	22.9	.229	.421
Case attributes
Child present
Yes	122	27.1	.271	.445
Victim injured
Yes	138	30.7	.307	.462
Drugs/alcohol
Yes	71	15.8	.158	.365
Argument over child custody
Yes	35	7.8	.078	.268
Argument over infidelity
Yes	61	13.6	.136	.343
Argument over money
Yes	19	4.2	.042	.201
Argument included destruction of property
Yes	71	15.8	.158	.365
Threats
Yes	23	5.1	.051	.220
Weapon present
Yes	35	7.8	.078	.268
Offender on scene
Yes	256	56.9	.569	.496
Arrest made
Yes	105	23.3	.233	.423

Note. Sample based on listwise deletion of missing data.

Table 2.

Unstandardized Coefficients (Standard Errors) and Odds Ratios From Logistic Regression of Arrest on Victim and Case Attributes (N = 450).

	Model 1	Model 2	Model 3
Variable	B Exp(b)	B Exp(b)	B Exp(b)
Black female victim	0.843 (.29) 2.32**	0.770 (.30) 2.16**	0.705 (.36) 2.02*
Black male victim	0.586 (.41) 1.80	0.726 (.42) 2.07#	0.078 (.54) 1.08
White male victim (White female = reference)	0.477 (.47) 1.61	0.381 (.49) 1.46	0.165 (.58) 1.18
Married (1 = yes)		0.256 (.27) 1.29	0.424 (.34) 1.53
Child present (1 = yes)		1.07 (.24) 2.92***	1.28 (.34) 3.58***
Victim injured (1 = yes)			2.64 (.31) 14.06***
Drugs/alcohol (1 = yes)			0.617 (.38) 1.85#
Argument over child custody (1 = yes)			0.046 (.53) 1.05
Argument over infidelity (1 = yes)			−0.191 (.40) 0.826
Argument over money (1 = yes)			−0.336 (.76) 0.714
Argument included destruction of property (1 = yes)			0.789 (.35) 2.20*
Threats (1 = yes)			0.049 (.79) 1.05
Weapon present (1 = yes)			2.05 (.46) 7.75***
Offender on scene (1 = yes)			0.661 (.31) 1.94*
−2 log likelihood	479.511	458.535	328.146
Model χ²	9.435*	30.410***	160.800***

^#p < .10. *p < .05. **p < .01. ***p < .001. Two-tailed tests of significance.

As displayed in Table 2, nested models introduce sets of covariates into the regression equation to predict odds of arrest. The more conservative two-tailed test of significance is employed given that proposed directional hypotheses are not provided for all of the model variables. Model 1 and Model 2 include demographic variables. Model 3 includes situational factors that could affect the cases’ severity and consequently police officers’ decisions to arrest. Model 1 indicates that in cases with Black female victims, the odds of arrest are 2.3 times higher than cases with White female victims (Exp[.84] = 2.32; p ≤ .01).

Model 2 includes variables representing family demographics by adding dummy variables for marital status and whether a child was present at the time of the incident. This model reveals that incidents where at least one child was present almost tripled the likelihood of arrest relative to cases where no child was present at the time of the incident (Exp[b] = 2.92; p ≤ .001). When including the family demographics, cases with Black female victims are still 2 times as likely to result in an arrest than cases with White female victims (Exp[b] = 2.160; p ≤ .01). The significant increase in the model chi-square indicates an improvement in the model fit.

Model 3 includes the situational covariates and reveals that just four of these variables significantly increase the odds of an arrest. Cases that include a visible victim injury, a weapon, and destruction of property significantly increase the likelihood of arrest. Cases where the offender was on the scene at the time of the arrival of law enforcement also had a higher likelihood of resulting in an arrest. Cases where the victim had a visible injury are 14 times as likely to result in an arrest compared to cases without a visible injury (Exp[b] = 14.06; p ≤ .001). The destruction of property had an odds of arrest 2 times larger than cases without the destruction of property (Exp[b] = 2.20; p ≤ .05). Incidents where a weapon was present are almost 8 times more likely to result in an arrest (Exp[b] = 7.75; p ≤ .001). The odds of arrest are 94% higher when the offender is at the scene when police arrived (Exp[b] = 1.94; p ≤ 05). The drastic increase in the model chi-square indicates a large and significant improvement in model fit. In this full model, the Black female victim variable retains its significant effect net of these additional variables. Results indicate that overall, although police officers have discretion in the decision to arrest, they are primarily relying on five variables to determine whether to arrest and that cases with Black female victims relative to White female victims are more likely to result in an arrest, even when accounting for other variables that affect arrest decisions.

Discussion

South Carolina remains one of the deadliest states for women. The passage of the DVRA in 2015 allows prosecutors to seek harsher punishments (including the ability to charge first-time offenders with felonies), makes receiving bond after arrest more difficult, provides officers with more arrest discretion by removing mandatory arrest laws, and importantly reflects the public sentiment that IPV needs to be addressed.

Despite South Carolina having passed ostensibly more progressive domestic violence reform, the reliance on evidence and probable cause to make an arrest indicates that officers may make fewer arrests than under mandatory arrest laws. Studies on arrest discretion and police policies illustrate tensions between the motivation of the laws, the local senior officers/administration and their ability to train guided discretion, and officers’ ability to exercise discretion with the physical evidence at hand (Barlow & Walklate, 2018). As Barlow and Walklate (2018) note, discretion needs to be understood holistically and discretion is “the requirement for accountability that sets limits on what is considered reasonable, actionable, and doable rather than discretion per se” (p. 3). The variables reflecting officer discretion in this small police jurisdiction reinforce the principles of guided discretion with little variation demonstrated by police officers.

Cases that do not neatly fit into an officer’s idea of a victim may further amplify arrest decision uncertainty. Best (1997) acknowledges that victims are readily recognized when the relationships are clear, unambiguous, and have strong claims of suffering and exploitation. Officers often find the relationships between the aggressors and victims ambiguous due to uncertainty and an inability to identify the catalyst of the incident and whether actions were defensive or aggressive outside of witness statements. Physical injuries are clear evidence that help officers determine the primary aggressor. Overreliance on the observed injury variable may understate the reality that many women who engage in violence do so defensively, and to equalize their size disadvantage, may turn to defensive fighting, which often obfuscates the relationship of who is being exploited to responding officers (Miller, 2001). The likelihood of arrest being influenced by a visible injury may also undermine the severity of other types of IPV and IPV threats (Robinson et al., 2016). The presence of verbal threats, which attempts to capture the victims’ perceptions of future harm, is not a significant predictor of arrest. Officers may recognize threats as a commonly exaggerated theme in IPV incidents with victims overstating the severity of those threats. It may also suggest that departmental expectations/perceptions exist that a case will be successful at trial to justify an arrest (Brown & Frank, 2005) and perceptions of future harm are difficult to justify without clear evidence.

We find that cases with Black female victims are more likely to result in arrest. This suggests the possibility that Black male offenders are being criminalized. In separate analyses (available upon request), we include interracial couples and find that cases with Black female victims were less likely to result in an arrest in the full model. Thus, when including cases with Black female victims and White male offenders, cases with Black female victims were not more likely to result in an arrest relative to cases with White female victims. This lends support to the supposition that Black male offenders are potentially being criminalized as opposed to treating these crimes as “ordinary.” While this dataset is small and does not allow for examination of responding officers’ race, future studies would do well to explore this. Given the public accusations of racism toward Black citizens, it is possible that a racial bias exists toward Black male offenders. Conversely, the increase in arrests may serve as a protective factor toward Black female IPV victims. These findings underscore the value in examining local police data to capture nuances of culture that may be lost in large datasets.

Second, these findings contribute to the literature by illustrating a clear pattern that although officers have arrest discretion, they are choosing to arrest in cases with the clearest and strongest evidence due to officers’ accountability for their arrest decisions (Barlow & Walklate, 2018). As the policy change increased officer discretion, it is possible that responding officers are giving more weight to the severity of the incident to help them determine an arrest decision as a visible injury suggests a potentially greater danger to the victim. In addition, visible injury was included in the initial policy and thus some officers may continue to rely on it as an objective measure in establishing probable cause to make an arrest. Local police guidelines and concerns over establishing probable cause are likely explanations. For example, the Aiken Department of Public Safety encourages officers’ use of discretion, but if a primary aggressor is unable to be determined and there are not immediate threats to safety, officers are cautioned that an arrest may exacerbate the situation.

Although arrests are not required if children are present, children’s presence is a significant predictor of arrests in this study, but is not in others (Hirschel & Buzawa, 2013; Hirschel et al., 2010). Children’s presence may be an extralegal variable officers use to make arrest decisions and evaluate the gravity of the situation by viewing children as indirect IPV victims, or perhaps elevating officers’ perceptions of victimhood if the female victim is a mother. The passage of the DVRA allows for charge enhancements if IPV occurs in the presence of children and our findings support the idea that officers are taking this into consideration. Alternatively, the variable measuring whether the conflict involved custody or childrearing does not approach significance. The absence of statistical significance on this variable may indicate that children’s indirect involvement may matter very little in officers’ decision-making. Future work should more closely investigate the role of the presence of children in officers’ decisions to arrest.

The finding of increasing odds of arrest when children are present may be linked to enhanced charges and mandatory reporting laws associated with children’s presence at the IPV scene, which reflect departmental expectations. If officers choose not to arrest when children are present, this could reflect poorly on officers’ decisions when the mandatory reporter reviews the incident case report. Incident report narratives are used as the justification for their decisions. Depending on how favorably the local culture views arrest, officers may limit discretion if clear objective measures are not demonstrated or reportable on the form. If local cultures are critical of arrest, they may emphasize maintaining safety without an arrest. For example, not all cases indicating alcohol abuse resulted in an arrest; some of these incidents were resolved by separation for the night. This suggests arrest occurs in cases that are the most threatening to victims’ safety and/or a child’s presence triggers mandatory reporting by senior administrators.

Written guidelines affect officers’ decisions to arrest; however, officers still demonstrated limited individual discretion outside of these guidelines. For example, the written guidelines instruct officers to take into consideration the offender’s complaint history at the location, but we see little evidence of this practice in the incident reports. Anecdotally, we did recognize a handful of cases of repeat offenders just in the 2-year span we investigated, but very rarely was there mention of this, even if it was a call earlier that day. Similarly, the offender being located on scene significantly increased the odds of arrest, although there is no mention of offender’s presence in the written guidelines.

In sum, these data show few IPV arrests despite increased discretion and public outcry in larger circulation South Carolina newspapers for IPV to be taken more seriously. The DVRA resulted in more severe punishment of convicted offenders, but also allowed for fewer arrests by removing mandatory arrest and allowing more discretion. We conclude that Aiken officers remain loyal to key legal factors that used to require arrest and choose not to evaluate other factors more heavily. There is evidence that officers are exercising discernment in regard to the presence of children. Guided discretion provides parameters for police officers to make arrests, but senior administrators in our data shape the parameters that encourage or discourage arrests. This study augments literature addressing the extent to which larger organizational features, such as mandatory arrest policies, and guidelines for arrest, shape how police officers make arrest decisions (Eitle, 2005).

Future research should further explore how local culture can shape officer use of discretion. In this research, the factors related to arrest are those that are the most easily defendable or justifiable, objective, and pose a higher risk for harm. In small police departments without mandatory arrest policies, the objectives of senior administrators may shape culture. Furthermore, this research suggests the need for qualitative research to supplement the existing quantitative studies that examine the role of organizational elements in the decision to arrest in IPV cases (Eitle, 2005; Robinson et al., 2018).

Limitations

There are several limitations with this study. The first is that we only examine 2 years of data post-DVRA and do not compare to pre-DVRA data. Consequently, it is unknown how the arrests have changed in Aiken, only that officers are not readily using officer discretion at this time. Due to changes in the Aiken Department of Public Safety’s database management system, we can only compare the IPV arrest rate to 2015. We find a decrease in IPV offender arrests since the passage of the DVRA. The IPV arrest rate in 2017 was 12.6% compared to 16.6% in 2015. These arrest numbers reflect a larger aggregate of cases that include family offenses and household disturbance cases. The ADPS changed their coding and reporting system in 2016. The 2017 arrest rate of 12.6% provides a meaningful comparison including all of these cases. Our analysis, however, only includes IPV incidents. The larger percentage of arrests that we report is reflecting arrests only related to the smaller sample of IPV-only incidents. The changing of local police department incident report coding further emphasizes the ways in which smaller departments are unique from those often represented in larger secondary datasets.

Another limitation is that the incident reports capture only one moment of specific abuse by one or both partners. Like many agency sample studies, this study fails to distinguish between different types of violence (e.g., intimate terrorism versus situational violence) and those who are most at risk (Myhill, 2015). The significance of the victim sustaining a visible injury suggests that cases of nonphysical abuse and cases of long-term emotional abuse are not being addressed by arrest. In addition, only IPV incidents in Aiken are examined, preventing any sort of comparison to other South Carolina cities. We recognize these limitations, but also argue that focusing on one jurisdiction avoids potential variability between policy and documentation practices that would confound small sample sizes and is a strength of this study.

The final limitation of this study, however, surrounds the subjective nature of the police incident reports. Two independent officers faced with the same IPV event may not only interpret severity of the incident differently and use arrest discernment differently, but they likewise may record the event uniquely. Some officers may not have been as motivated to complete the narrative section as thoroughly as others, regardless of whether omitted information influenced their decisions. Therefore, we cannot say with certainty that things did or did not occur, but we can confidently state that some officers did find some factors significant enough to record and these factors affected their judgment. It is possible that more experienced officers are more comfortable with IPV and how to complete the incident report’s narrative section.

Footnotes

Acknowledgements

The authors gratefully acknowledge Timothy McClure for his feedback on an earlier version of this study and the Aiken Department of Public Safety for providing access to the data analyzed.

Authors’ Note

An earlier version of this article was presented at the 2018 meeting of the Southern Sociological Society.

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: The first author acknowledges support in the form of a summer research grant RISE Grant no. 17110-19-50518 from the University of South Carolina.

ORCID iD

Philip B. Mason

Notes

Author Biographies

Philip B. Mason is an associate professor of sociology at the University of South Carolina Aiken. His research interests include health, religion, and domestic violence. Within these domains, he has a special interest in rural–urban comparisons and often uses local data to inform policy. He also serves on the board and volunteers as a victims’ advocate for his local domestic violence organization.

Michelle A. Petrie is an associate professor of sociology at the University of South Carolina Aiken. Her primary research interests revolve around social inequality with a particular focus on race and gender. Her prior research applies these interests to a wide range of institutions including the workplace, the environment, and the criminal justice system.

References

Avakame

E. F.

Fyfe

J. J.

(2001). Differential police treatment of male-on-female spousal violence: Additional evidence on the leniency thesis. Violence Against Women, 7, 22–45. https://doi.org/10.1177/10778010122182280

Bachman

Coker

A. L.

(1995). Police involvement in domestic violence: The interactive effects of victim injury, offender’s history of violence, and race. Violence and Victims, 10, 91–106. https://doi.org/10.1891/0886-6708.10.2.91

Balko

(2016, April 1). Video shows white cops performing roadside cavity search of black man. Washington Post. https://www.washingtonpost.com/news/the-watch/wp/2016/04/01/video-shows-white-cops-performing-roadside-cavity-search-of-black-man/?utm_term=.51de9fd22533

Barlow

Walklate

(2018). Policing intimate partner violence: The “golden thread” of discretion. Policing A Journal of Policy and Practice, 14, 404–413. https://doi.org//10.1093/police/pay001

Best

(1997). Victimization and the victim industry. Society, 34, 9–18.

Bonner

H. S.

(2015). Police officer decision-making in dispute encounters: Digging deeper into the “black box.” American Journal of Criminal Justice, 40, 493–522. https://doi.org/10.1007/s12103-014-9274-2

Brown

R. A.

Frank

(2005). Police-citizen encounters and field citations: Do encounter characteristics influence ticketing? Policing: An International Journal of Police Strategies & Management, 28, 435–454. https://doi.org/10.1108/13639510510614546

Buzawa

Austin

(1993). Determining police response to domestic violence victims. American Behavioral Scientist, 36, 610–623. https://doi.org/10.1177/0002764293036005006

Buzawa

Hotaling

Klein

Byrne

(1999). Response to domestic violence in a pro-active court setting: Final report. National Institute of Justice. https://www.ncjrs.gov/pdffiles1/nij/grants/181427.pdf

10.

Devine

Buddenbaum

Houp

Studebaker

Stolle

(2009). Strength of evidence, extraevidentiary influence, and the liberation hypothesis: Data from the field. Law and Human Behavior, 33, 136–148. https://doi.org10.1007/s10979-008-9144-x

11.

Durfee

Fetzer

M. D.

(2016). Offense type and the arrest decision in cases of intimate partner violence. Crime & Delinquency, 62, 954–977. https://doi.org/10.1177/0011128714540277

12.

Eitle

(2005). The influence of mandatory arrest policies, police organizational characteristics, and situational variables on the probability of arrest in domestic violence cases. Crime & Delinquency, 51, 573–597. https://doi.org/10.1177/0011128705277784

13.

Feder

(1996). Police handling of domestic calls: The importance of offender’s presence in the arrest decision. Journal of Criminal Justice, 24, 481–490. https://doi.org/10.1016/S0047-2352(96)00034-7

14.

Goldstein

(1963). Police discretion: The ideal versus the real. Public Administration Review, 23, 140–148. https://doi.org/10.2307/973838

15.

Hirschel

Buzawa

(2013). The impact of offenders leaving the scene on the police decision to arrest in cases of intimate partner violence. Violence Against Women, 19, 1079–1103. https://doi.org/10.1177/1077801213501843

16.

Hirschel

Hutchison

I. W.

Shaw

(2010). The interrelationship between substance abuse and the likelihood of arrest, conviction, and re-offending in cases of intimate partner violence. Journal of Family Violence, 25, 81–90. https://doi.org/10.1007/s10896-009-9272-1

17.

Hirschel

McCormack

P. D.

Buzawa

(2017). A 10-year study of the impact of intimate partner violence primary aggressor laws on single and dual arrests. Journal of Interpersonal Violence. https://doi.org/10.1177/0886260517739290

18.

Johnson

Goodlin-Fahncke

(2015). Exploring the effect of arrest across a domestic batterer typology. Juvenile & Family Court Journal, 66, 15–30. https://doi.org/10.1111/jfcj.12024

19.

Karlsson

M. E.

Quińones

K. R.

López

C. M.

Andrews

A. R.

III Wallace

M. M.

Rheingold

(2018). Sociodemographic and incident variables as predictors of victim injury from intimate partner violence: Findings from police reports. Violence Against Women, 24, 1413–1432. https://doi.org/10.1177/1077801217733558

20.

Kinports

(2010). Veteran police officers and three-dollar steaks: The subjective/objective dimensions of probable cause and reasonable suspicion. Journal of Constitutional Law, 12, 751–784.

21.

Klinger

D. A.

(1997). Negotiating order in patrol work: An ecological theory of police response to deviance. Criminology, 35, 277–306.

22.

Larance

L. Y.

Goodmark

Miller

S. L.

Dasgupta

S. D.

(2019). Understanding and addressing women’s use of force in intimate relationships: A retrospective. Violence Against Women, 25, 56–80. https://doi.org/10.1177/1077801218815776

23.

Lee

Zhang

Hoover

L. T.

(2013). Police response to domestic violence: Multilevel factors of arrest decision. Policing: An International Journal of Police Strategies & Management, 36, 157–174. https://doi.org/10.1108/13639511311302524

24.

Miller

(2001). The paradox of women arrested for domestic violence. Violence Against Women, 7, 1339–1376. https://doi.org/10.1177/10778010122183900

25.

Muftić

L. R.

Finn

M. A.

Marsh

E. A.

(2015). The victim-offender overlap, intimate partner violence, and sex: Assessing differences among victims, offenders, and victim-offenders. Crime & Delinquency, 61, 899–926. https://doi.org/10.1177/0011128712453677

26.

Myhill

(2015). Measuring coercive control: What can we learn from national population surveys? Violence Against Women, 2, 355–375. https://doi.org/10.1177/1077801214568032

27.

Myhill

Johnson

(2016). Police use of discretion in response to domestic violence. Criminology & Criminal Justice, 16, 3–20. https://doi.org/10.1177/1748895815590202

28.

Novak

K. J.

Frank

Smith

B. W.

Engel

R. S.

(2002). Revisiting the decision to arrest: Comparing beat and community officers. Crime & Delinquency, 48, 70–98. https://doi.org/10.1177/0011128702048001003

29.

O’Neal

E. N.

Beckman

L. O.

Spohn

(2016). The sexual stratification hypothesis: Is the decision to arrest influenced by the victim/suspect racial/ethnic dyad? Journal of Interpersonal Violence, 34, 1–24. https://doi.org/10.1177/0886260516651093

30.

Paternoster

(1984). Prosecutorial discretion in requesting the death penalty: A case of victim-based racial discrimination. Law & Society Review, 18, 437–478.

31.

Poon

Dawson

Morton

(2014). Factors increasing the likelihood of sole and dual charging of women for intimate partner violence. Violence Against Women, 20, 1447–1472. https://doi.org/10.1177/1077801214557954

32.

Quillian

(2006). New approaches to understanding racial prejudice and discrimination. Annual Review of Sociology, 32, 299–328. https://doi.org/10.1146/annurev.soc.32.061604.123132

33.

Robins

(2017). Domestic violence law: State of fear: Domestic violence in South Carolina. South Carolina Law Review, 68, 629–664.

34.

Robinson

A. L.

Chandek

M. S.

(2000). The domestic violence arrest decision: Examining demographic, attitudinal, and situational variables. Crime & Delinquency, 46, 18–37. https://doi.org/10.1177/0011128700046001002

35.

Robinson

A. L.

Pinchevsky

G. M.

Guthrie

J. A.

(2016). Under the radar: Policing non-violent abuse in the US and UK. International Journal of Comparative and Applied Criminal Justice, 40, 195–208. https://doi.org/10.1080/01924036.2015.1114001

36.

Robinson

A. L.

Pinchevsky

G. M.

Guthrie

J. A.

(2018). A small constellation: Risk factors informing police perceptions of domestic abuse. Policing and Society, 28, 189–204. https://doi.org/10.1080/10439463.2016.1151881

37.

Romain

Freiburger

T. L.

(2016). Chivalry revisited: Gender, race/ethnicity, and offense type on domestic violence charge reduction. Feminist Criminology, 11, 191–222. https://doi.org/10.1177/1557085115580389

38.

Skogan

Frydl

(2004). Fairness and effectiveness in policing: The evidence. National Academics Press.

39.

Smith

(2018, September 30). Despite reforms, SC still ranks among nation’s deadliest states for women. Post & Courier. https://www.postandcourier.com/news/despite-reforms-sc-still-ranks-among-nation-s-deadliest-states/article_8ba80818-c31c-11e8-86ae-7f181e23e903.html/

40.

Sorenson

S. B.

(2017). Guns in intimate partner violence: Comparing incidents by type of weapon. Journal of Women’s Health, 26, 249–258. https://doi.org/10.1089/jwh.2016.5832

41.

Tatum

K. M.

Pence

(2015). Factors that affect the arrest decision in domestic violence cases. Policing: An International Journal of Police Strategies & Management, 38, 56–70.

42.

U.S. Census Bureau. (n.d.). Quick facts: Aiken City, South Carolina. https://www.census.gov/quickfacts/aikencitysouthcarolina/

43.

Violence Policy Center. (2018). When men murder women: An analysis of 2016 homicide data. Violence Policy Center. http://vpc.org/studies/wmmw2018.pdf/

44.

Williams

M. R.

Demuth

Holcomb

J. E.

(2007). Understanding the influence of victim gender in death penalty cases: The importance of victim race, sex-related victimization, and jury decision making. Criminology, 45, 865–891. https://doi.org/10.1111/j.1745-9125.2007.00095.x

45.

Young

D. S.

Casey

E. A.

(2018). An examination of the sufficiency of small qualitative samples. Social Work Research, 43, 53–58. https://doi.org/10.1093/swr/svy026