Mandatory arrest: A flawed policy based on a false premise

In sum, the labeling of all acts of physical aggression as violent [italics added] can have unintended social implications.

(K. Daniel O’Leary)

This was presented recently at a domestic violence conference in Sacramento, Calif. Some of the information below appears in my previous columns; however, much of the information and studies have been updated. Of interest to those who work in or with the criminal justice system is the fact that a number of new hyperlinked studies are in the reference section. In the opinion of this author, in the 21st century far too many public policies continue to reflect ideologically held beliefs from 20th century. It is imperative for public policy makers to be aware of 21st century empirical findings.

The flawed policy

The National Institute of Justice report, Controlling Violence Against Women: A Research Perspective on the 1994 Violence Against Women Act’s Criminal Justice Impacts, cautions that advocates and public policy makers need to know that their policies and practices will not endanger women (Ford, Bachman, Friend & Meloy, 2002). What should be troubling to all criminal justice professionals is that most advocates and almost all public policy makers do not know or are unwilling to recognize that mandatory “one-size-fits-all” policies can endanger some women, men and children (Dugan, Nagin & Rosenfeld, 2003; Eng, 2003; Hanna, 1998).

Most reasonable and prudent people — particularly those familiar with the criminal justice system — acknowledge that it is unrealistic to expect restraining/protection orders or arrest alone to be effective in deterring domestic violence. The research documents that without a coordinated community response in place, mandatory policies alone are not enough to protect victims from harm. National Institute of Justice studies clearly document that most communities do not have a coordinated community response (Hirschel & Dawson, 2003).

Despite the lack of empirical evidence for the efficacy of mandatory policies, they are in place in 24 states. And all 50 states have some form of de facto mandatory (preferred or pro-arrest) domestic violence polices (Miller, 2005). The obvious problems created by placing the dangerous cart (mandatory policies) before the protective horse (a coordinated community response) continues to be ignored by advocates and public policy makers (Davis, 2008).

An ever-growing number of studies by feminist researchers document that mandatory “one-size-fits-all” policies (intervening as if victims and offenders constitute a homogeneous group), create the false illusion of hope that the criminal justice system is actually capable of protecting (other than by the incarceration of offenders) families (Crisler, 2005; Davis, 1998; Eng, 2003; Hanna, 1998; Stake, 2005; Wells & DeLeon-Granados, 2005). And, by their very nature, mandatory one-size-fits-all policies ignore the diverse needs of families (Eng, 2003).

The contemporary criminal justice dichotomy created by mandatory policies is that far too many minor offenders, including some victims, are being arrested while far too many chronically violent offenders are not being incarcerated (Toon, Hart, Welch, Corando, & Hunting, 2005). Mandatory one-size-fits-all policies eviscerate the diverse support and varied resources that some families require (Eng, 2003; Wells & DeLeon-Granados, 2005).

Advocates and public policy makers ignore the fact that mandatory policies are placing some families, because of their distinctive characteristics (including race, ethnicity, class, education, employment and religion) in greater danger (Crisler, 2005; Dugan, Nagin, & Rosenfeld, 2003; Eng, 2003; Hanna, 1998; Maxwell, Garner, & Fagan, 2001; Sherman, 1992; Stake, 2005; Wells & DeLeon-Grandos, 2005).

Most advocates and public policy makers claim that the Minneapolis Domestic Violence Experiment documents that “arrest works best.” They ignore that the results of the experiment were specific to a particular demography and the results do not imply that all suspected offenders be arrested (Sherman & Berk, 1984). Advocates and public policy makers also do not consider the fact that the effort was just one of six NIJ sponsored experiments.

What the advocates, public policy makers and the literature almost universally omit is that the research from all six studies document that the majority of offenders discontinued their abusive behavior without being arrested (Maxwell, Garner & Fagan, 2001). Not one of the six studies document that mandatory arrest should be the preferred policy. In fact, the lead author of the Minneapolis Domestic Violence Experiment continues to publicly argue against mandatory arrest (Sherman, 1992).

The warning that mandatory policies may unnecessarily limit a community’s resources from identifying and prosecuting the most chronic violent offenders and that they restrict resources needed by families at greatest risk appears to have fallen through the cracks of most research studies (Davis, 1998; Maxwell, Garner & Fagan, 2001; Sinden & Stephens, 1999).

The false premise

The Minneapolis Domestic Violence Experiment was conducted by academic researchers, but advocates and researchers do not acknowledge that without the concerns of three law enforcement officers the experiment would not have taken place (Sherman & Berk, 1984). Despite what advocates claim, these three officers represent the beliefs of the majority of law enforcement and in fact the entire criminal justice system. (Davis, 1998; Sinden & Stehphens, 1999; Toon, Hart, Welch, Coronado & Hunting, 2005).

Public policy makers appear to have passed mandatory policies based almost solely on anecdotal stories that do not present the context and circumstances of individual incidents, and that lack information about the seriousness of the incident or the diverse needs of individual victims. Mandatory arrest is a spurious hypotheses based on an unsubstantiated and false premise that most law enforcement officers do not care about domestic violence victims and the anecdotal claims by some advocates that officers care less about female than male victims.

The reports of law enforcement failing to respond appropriately or refusing to take proper action are primarily found in retrospective studies of law enforcement domestic violence responses. The context and circumstances of the individual incident is usually ignored. In addition, researchers often do not separate the serious and chronic violent assaults from minor or infrequent physically aggressive behavior (Davis, 1998; Sinden & Stephens, 1999).

Context and circumstances

In their book, “Crisis Intervention,” the authors write that, “To understand intimate partner violence, it is important to make a distinction between common couple [minor] violence and chronic [serious] battering” (Hendricks, McKean, Hendricks, 2003, p. 228). Regardless of state legislation, de facto mandatory arrest polices along with the twin fiscal concerns of law enforcement administrators concerning federal funding and federal lawsuits have coerced administrators into implementing mandatory polices (Davis, 1998).

The National Violence Against Women Survey (Tjaden & Thoennes, 2000 p. 11) documents that “most physical assaults committed against women and men by intimates are relatively minor and consist of pushing, grabbing, shoving, slapping and hitting.” Mandatory policies demand that criminal justice professionals ignore the difference between serious (battering) and minor (family conflict) (Davis 2008; Johnson & Ferraro, 2000).

It is now suggested by advocates, perhaps because of the dramatic rise in the numbers of women being arrested, that law enforcement needs to consider the context and circumstances of incidents before they make an arrest for minor assaultive behavior. However, advocates and law enforcement training suggest that officers should consider the context and circumstances of the behavior of females and not males (Davis, 2008). One former prosecutor notes that the evidentiary standards for proving “abuse” are now so low that any man who is charged or appears in court is considered guilty (Hanna, 1998).

During my 21 years of experience in law enforcement, most officers — when not constrained by mandatory policies — did consider the context and circumstances of incidents. Incidents that were considered by the officer and society in general to be minor most often did not result in an arrest. Most serious incidents that were considered to be serious crimes or felonies did result in an arrest (Davis, 1998; Sherman, 1992).

Domestic violence homicide

Regardless of claims by advocates and public policy makers, the NIJ study, The Decline of Intimate Partner Homicide, reports that, “There was no statistically significant relationship between any criminal justice system response and victimization for either gender or for any racial or ethnic group” (Wells & DeLeon-Grandos, 2005, p. 33).

Contemporarily, the definition of “domestic violence or abuse” varies from state to state, from law enforcement agency to law enforcement agency, advocate to advocate and victim to victim. Intimate partner homicide is far more objective than subjective domestic violence assault claims. Most domestic violence advocates and public policy makers agree that the primary goal of the Violence Against Women Act is to reduce intimate partner homicides.

The Bureau of Justice Statistics online report, Homicide Trends in the United States, under, “The proportion of all homicides involving intimates by gender of victim, 1976-2005,” the data documents that prior to the Violence Against Women Act from 1976 to 1993 the number of female intimate partner homicides decreased from 34.5 percent to 28.2 percent. However, after the passage of the law from 1994 to 2005 the same report documents that while the percent of nonintimate or unknown homicides of women decreased from 72.0 percent to 66.7 percent the number of intimate partner homicides of women increased from 28.0 percent to 33.3 percent.


Those most responsible for interventions, policies, prevention and solutions still have much to learn. An interesting way to learn what law enforcement officers think about domestic violence might be an extensive survey rather than relying on anecdotal stories (Davis, 1998; Toon, Hart, Welch, Coronado & Hunting, 2005; Sinden & Stephens, 1999). Professionals in one area of specialization need to understand and appreciate the difficulties that other professionals face. There needs to be less proffering by researchers and advocates that they have already discovered “the answer.” Until researchers are willing to agree just what “domestic violence” is or is not, it will be improbable to impossible for them to discover the cause or provide a cure (Davis, 1998; Davis, 2008; Johnson & Ferraro, 2000; Pence & Dasgupta, 2006; Soler, 2007; Wallace, 2002).

For references to articles, studies and reports mentioned in this story, click here.

About the author

Richard L. Davis completed studies in criminal justice management at LaSalle University. He has a graduate degree in criminal justice from Anna Maria College, and another in liberal arts with a concentration in history from Harvard University.

Contact Richard Davis.

Request product info from top Police Training companies

Thank You!

= required Error occured while sending data

By submitting your information, you agree to be contacted by the selected vendor(s).

Join the discussion

Thanks! You've been successfully signed up for the P1 career newsletter!

Brand focus

Sponsored content
5 biggest LMS implementation 'fails' (and how to avoid them)

5 biggest LMS implementation 'fails' (and how to avoid them)

Make sure your learning management system rollout is successful by avoiding these mistakes

logo for print