©1999 - 2012
Edward D. Reuss
All rights reserved. Including the right of reproduction in whole or part in any form


Domestic Violence and Default Arrest

In sum, the labeling of acts of physical aggression as violence can have unintended social implications.

K. Daniel O’Leary

Approximately one year before I retired in 1996 a police officer on my department shot his wife with his service revolver and then took his own life. I was involved with some of the criminal justice and social efforts to prevent this horrific tragedy. Obviously those efforts failed. Almost every day since that horrific homicide/suicide I have read, written or thought about domestic violence and its intersection with the criminal justice system.

I remain deeply troubled that in this 21st century we continue with failed 20th century criminal justice efforts that are based on hypotheses to prevent domestic violence incidents. We continue to proffer the belief that “arrest works best in all domestic violence incidents” and that “arrest deters repeat abuse.” These two hypotheses remain reality to many people despite the fact that over the past decades no single individual or groups of researchers have produced a single empirical evidence-based study that supports those hypotheses.

Domestic violence remains a complex and multifaceted issue and we continue to have great difficulty in understanding and engaging its complexities because far too many people claim we have discovered the cause. Reams of empirical evidence-based data documents that it should crystal clear that there is no single cause, no single intervention and no single prevention. 

Approximately 17 years ago, in my home state of Massachusetts, domestic violence was so epidemic that the governor declared a state of emergency. And similar the United States Attorney General in 1984 [perhaps the original default position] he declared that the criminal justice system was going to lead the way.

There can be no argument that since then the criminal justice system has engaged the issue of domestic violence far more than any other segment of society. However, in June of 2008 the current governor declared another domestic violence emergency; at that second declaration the Lieutenant Governor stated that the trend was moving in the wrong direction [domestic violence homicides and suicides were rising not falling] and the administration’s goal is to better understand the pattern of domestic violence.

It should be clear by now that many if not most contemporary interventions are not working and that criminal justice intervention, while often necessary, is reactive and not preventative. While many public policy makers and interveners point towards the drop in domestic violence, the facts are that domestic violence has traditionally risen or fallen at a similar rate to violence in general. 
While it is true that some rates of domestic violence have fallen, the same is true of some rates of violence in general.

The belief that the criminal justice system will prevent domestic violence and that domestic violence is primarily caused by the patriarchic behavior of males are red herrings that continue to mislead us in our attempt to prevent or minimize all forms of domestic violence.

Over the last four decades the central theme of public policy makers and interveners continues to be if only the police would arrest the abuser the victim would be protected. Many people cling desperately to that belief despite the ever rising number of empirical evidence-based studies that provide data that arrest is reactive and not preventative. Anyone remotely familiar with the criminal justice system understands that in cases of serious, injurious or repeat abuse arrest is necessary. However, simply because arrest is necessary in some instances of domestic violence does not demonstrate arrest is preventative and necessary in all instances.

I recognized in 1995 as I do now in 2009 that families, particularly those at the lower end of the socioeconomic/educational strata of society who suffer chronic or serious abuse are often not afforded the domestic violence intervention they so often need or desire.

There are reams of data that clearly document that the criminal justice system does not provide an equal opportunity for all our citizens. One only needs to visit the Bureau of Justice Statistics website to discover that there is a direct correlation between socioeconomic/educational status and offending, victimization, arrest, conviction and incarceration. While many citizens continue to argue that those dramatic criminal justice differences are primarily because of race, when socioeconomic/educational status is controlled the issue of race all but disappears.

What should be obvious to all of us by now, given the data over the last 25 years, is that we are not going prevent domestic violence by arrest or incarceration. Arrest, by its very nature is reactive and, despite a widely held belief by many researchers [again is far from being alone in holding this belief], public policy makers and domestic violence interveners there is still not a single empirical evidence-based study that can document arrest deters abuse for all domestic violence incidents.

All Domestic Violence Incidents

On page 17 of a recent report sponsored by the U.S. Department of Justice, Practical Implications of Current Domestic Violence Research the author, Andrew R. Klein, concludes that, “Arrest should be the default position for law enforcement in all [emphasis added] domestic violence incidents.” Thus mandated, encouraged, and preferred law enforcement arrest policies are now joined by default position policies. I have to assume [Klein does not define what he means by a default position] that arrest is where we should start. 

I agree with Klein that arrest should be the default position for law enforcement for all serious, injurious or repeat domestic violence incidents. However, I believe that an every growing number of studies warn us of the dangers of arresting for all domestic violence incidents in general.

Klein writes on page 16 that; “A major re-examination of a series of fairly rigorous experiments in multiple jurisdictions finds that arrest deters repeat reabuse [emphasis added]…” Paradoxically, I have read the studies Klein cites to support his claim and have discovered, after reading these studies again, that these studies provide little to no support for the deterrent effect of arrest.

On page 16 and 17 of Klein’s report he provides a number of citations that, I assume are intended to support his conclusion. The fact is that the studies he cites provide no empirical evidence-based data to support his conclusion that “arrest can deter reabuse in all domestic violence incidents.” The studies he provides produce, at most, moderate yet statistically insignificant data to support his claim. In fact many clearly refute his beliefs.

I know Andy Klein and I respect his beliefs and intentions. This is not intended to rebuke Klein. My intent is only to question his conclusions because his report is sponsored by the U.S. Department of Justice and his report will be accepted as fact by most law enforcement officers, prosecutors, judges and certainly most certainly domestic violence interveners who cling tenaciously to the belief that arrest is best and arrest deters repeat abuse.

Klein and I agree on a great many issues concerning domestic violence intervention by the criminal justice system What then causes this apparent schism? I believe that ideological held beliefs may be causing Klein and in fact a great many other intelligent and highly respected researchers to replace objective empirical evidence-based data they observe with subjective interpretations of that data. Klien helps explain this premise in the preface of his Practical Implications of Current Domestic Violence Research: For Law Enforcement, Prosecutors and Judges. In the preface Klein notes the following:

For example, although Jacobson and Gottman’s findings regarding the typology of batterers have been questioned, their reported observations [emphasis added], if not their conclusions, have been confirmed [emphasis added] (p. vi).

I have read the Jacobson and Gottman’s finding without preconceived or ideological held beliefs. Jacobson and Gottman observed that both males and females exhibited aggressive and assaulted behavior and females often initiated that behavior. Jacobson and Gottman did not observe that women were most always the passive and docile victims at the hands of aggressive and violent males. In fact, almost half, 28/57 of the women agreed that they were the primary aggressor during these observations and they agreed that they might have qualified for placement in a domestic violence program. 

I know Andy Klein and similar to the vast majority of researchers sponsored by the U.S. Department of Justice, I know that Klein , despite the observations by Jacobson and Gottman, does not believe that men and women can be equally aggressive or assaultive. Perhaps he does not believe that because the majority of perpetrators he observed in the Quincy, MA domestic violence court were primarily male, not female.

In fact, concerning injurious, serious, and lethal domestic violence I had the same observations. As a police officer I never once responded to a domestic violence incident where I observed the female standing over a beaten and battered male who had received a black eye or broken nose as a result of their altercation. And there is no denying that far more women were the victims of homicide than men. However, if were are to prevent domestic violence it must be understood that injuries and lethality are results not causes. To prevent the result we must understand the cause.

However, what I also did respond to and observed, as often as not, were a males and a females who appeared to be equally responsible for the lower levels of assaultive, aggressive and coercive behavior that is far more prevalent in relationships than injurious and lethal violence.

Klein and the majority of people who work in our courts do not make these same observations because they most often only observe the serious or lethal cases. And, I recognized then as I do now that many families, particularly those at the lower end of the socioeconomic/educational strata of society who suffer chronic abuse and serious or lethal violence are often not afforded the proper intervention or provided the resources they so often need or desire. 

The data in many of the studies provided by Klein, and others I have researched, clearly document that the many if not most of domestic violence incidents law enforcement respond to are often not injurious or lethal. Perhaps too many researchers are more inclined to report what they observe and believe in their own area of specialization. However, I have discovered that far too often many are unwilling and some are unable to understand or appreciate the difficulties being experienced by others.

The vast majority of the data cited in a summary of his findings clearly documents he was not observing data that provides an overview of the general population. Thus most of his conclusions are based on, “… domestic violence perpetrators that come to the attention of the criminal justice or court authorities have a prior criminal history for a variety of nonviolence and violence offenses against males as well as females…” Klein notes that, in a Massachusetts study 84.5% of the perpetrators had a prior criminal record that averaged a little more than 13 charges (p. 16).

Klein is far from alone in holding the belief that arrest works best in all domestic violence incidents and that there is data that actually confirms that arrest deters repeat abuse. Klein is just one of a majority of authors and domestic violence interveners who hold this belief. In fact, the belief that there are studies that document that arrest deters repeat abuse may be the most ubiquitously held belief among domestic violence researchers, public policy makers and interveners.

Domestic violence program workers cling firmly to the “arrest deters repeat abuse” and “arrest works best” beliefs even as the studies cited by Klein actually refute that belief and clearly document there is no empirical evidence-based data to support either of those two beliefs. Perhaps they do because they know something must be done.

However, are we doing the right thing by accepting arrest as the default [beginning?] position in the belief that will deter repeat abuse? Does it make sense to begin at the end of this enigma? I have yet to meet a single domestic violence program worker that actually believes the issue of domestic violence begins the day heterosexuals reach adulthood. Arrest is a very flawed and simple reaction to a very complex problem. 

As a medical analogy, we remain fixated in believing that our response should be to increase our client base by building more and better hospitals, to define almost all coercive or assaultive behavior as an act of domestic violence and to declare that we have discovers both the cause and the cure.

The Studies Cited

The first study Klein cites, we assume to support the belief that arrest is best in all domestic violence incidents is the The Effects of Arrest on Intimate Partner Violence: New Evidence From the Spouse Assault Replication Program . In fact on page 13 of the study the authors of this study suggest quite opposite:

This [study] suggests that policies requiring arrest for all [emphasis added] suspects may unnecessarily take a community’s resources away from identifying and responding to the worst offenders and victims most at risk.

The study that the vast majority of researchers claim supports the belief that arrest deters domestic violence repeat abuse is the Minneapolis Domestic Violence Experiment. However, if you actually take the time to read the study it provided no such data. All serious [felony] cases were screened out and only minor [misdemeanor] assaults were reported. Hence, it is improbable to impossible to use this study to make the claim that arrest works best for all domestic violence incidents.

In fact the study Klein uses to support his implication for law enforcement, prosecutors and judges that arrest be their default position actually warns against using arrest in all domestic violence incidents. On page 13 the report notes that, “a majority of suspects discontinued their aggressive behaviors even without an arrest.” Given that to be true it is impossible to claim that arrest alone was responsible for the lack of reabuse even by those who were arrested.

The next study Klein cites is Police Intervention and the Repeat of Domestic Assault. This study does not support using arrest in all domestic violence incidents. In fact the authors, on page 21, actually refute that belief:

Our research provides an additional reason to question whether arresting offenders for domestic violence is a deterrent. We find that the coefficient representing the effect of arrest on re-offending is in a predicted direction, but it is small and not statistically significant. Thus, our results, along with at leas some of the experimental research, challenge the efficacy of mandatory arrest [emphasis added].

There is no support in the above study to encourage law enforcement arrest policies nor mandate arrest for all domestic violence incidents. In fact from page 21 to page 26 the authors discuss many other alternatives to arrest. What this study actually concludes on page 24 is quite the opposite of Klein’s interpretations.

The result suggests that the reporting of partner violence to the police is a deterrent, even if the police do not make an arrest [emphasis added.]

The next study Klein cites Creating a Structured Decision-Making Model for Police Intervention in Intimate Partner Violence is both interesting and reveals data that both Klein and its author seem to ignore. Rather than supporting Klein’s and perhaps the author, Madeline Wordes, belief, a careful reading of this study demonstrates what most of us work have been working in the criminal justice system for years readily acknowledge.

(1)The reports documents that the highest reabuse rates are found for incidents where there is no evidence [no problem cause?] that an assault had taken place (p. 26).

The suggestion here is that approximately one in five (21%) arrests are being made where there is no observable criminal behavior and that the arrest may exacerbate the “family dispute” and does little to nothing to resolve it (p. 26).

This demonstrates that officers made arrests in these incidents that lack traditional “probable cause” only because of a department policy that encourages arrest for all domestic violence incidents even for incidents that do no result in physical contact (p. 10).”

(2) The report also suggests that arrest does not deter people who are “heavily involved in the criminal justice system (p. 26). The report documents that, “Suspects having a prior conviction for domestic violence also shows a slightly higher recidivism rate than average (p. 33).”

I find the above study to be the most interesting and informative of all the reports. In fact this study, more than the any other, documents the dangers inherent in using arrest as the default position for all domestic violence incidents.

The next report cited is, Beyond Arrest: The Portland, Oregon Domestic Violence Experiment, Final Report and it reports on page 17 that “police response also significantly increases the likelihood that victims’ will secure protective orders.” Klein has written extensively about the failure of these orders to protect victims from repeat abuse and an ever growing number U.S. Department of Justice sponsored studies support Klein’s position.

In fact on page of this study the authors reports that:

However, the smaller number of women who were revictimized in the treatment group reported on average as many incidents of revictimization as the larger group of revictimized women in the control group. And

Police reports of further victimization and reports of alleged reoffending during the six-month follow-up period were associated significantly more [emphasis added] often with the treatment group than the control group…”

Most troubling on page 107 of this report the authors report:

Dugan et al., (1997) have found an empirical link between the increased availability of domestic violence services and reduction in fatal partner violence

However, Dugan et al ., (2003) report, Do Domestic Violence Services Save Lives? actual warns:

Policies and services designed to help victims of domestic violence appear to have two possible and opposing effects: either they decrease the abuse and risk of homicides, or they have the unintended consequences of increasing them [emphasis added].

The fact is that there is little to no empirical evidence-based data in any of the reports that Klein cites to support his position that arrest should be used in all domestic violence incidents or that arrest deters repeat abuse (p. 16.) In fact there are now an ever growing number of studies published by the National Institute of Justice that warn us that intimate partner violence is a complex and multifaceted enigma and interveners need to provide interventions that address the context and circumstances of specific incidents and provide individual assessments that are based on empirical evidence-based data and not ideology. All one-size-fits-all interventions have proven to be problematic.

An ever growing number of researchers warn about the gullibility of accepting any “one-size-fits-all” intervention. Empirical evidence-based data in studies cited by Klein clearly document that arrest for all domestic violence incidents can be troubling double edged swords. Most troubling in the study by Madeline Wordes where the data suggests that approximately one of every five arrests are made without any traditional standards concerning problem cause or any actual evidence that an assault actually took place.  

Most of the other reports that Klein cites to support his positions only report about victim satisfaction or empowerment. While some of the data in these studies are interesting and helpful these studies provide little to no support for the Klein positions.

In my column Exploring law enforcments response to intimate partner violence I provide evidence that seems to demonstrates the majority of domestic disputes law enforcement respond to are verbal disagreements and not violent physical assaults. Hence, I find it almost impossible to agree with Klein when he suggests that arrest should be the law enforcement default position in all domestic violence incidents. And concerning the claim that arrest deters repeat abuse, most if not all these studies screen out the most serious incidents and some actually report higher repeat abuse for those arrested and convicted.

What is both troubling and ironic about the studies Klein cites in support of his position is that, if read in their entirety, most provide little to no empirical evidence-based data to support his positions.

It is my belief that the greatest single problem with many contemporary law enforcement domestic violence interventions is that many are based on hypotheses, ideology or opinion rather than empirical evidence-based data. Too often subjective beliefs are replacing objective facts. An unbiased and pragmatic read of the studies provided by Klein actually lead us in the opposite direction that arrest should be the default position for all domestic violence incidents and that arrest deters repeat abuse.

None of the six experiments sponsored by the U.S. Department of Justice, labeled the Spouse Assault Replication Program (SARP) conclude that arrest deters repeat abuse for all domestic violence incidents. The fact is that everyone who claims that the SARP demonstrates that “arrest is best” have either not read the SARP or have chosen to ignore the fact that all of the serious incidents were screened out.

Hence, it should be obvious that there is no empirical evidence-based data in any of the SARP studies that support the position that “arrest deters repeat abuse” for all domestic violence incidents.

Most troubling to me is that, at least do date, I have yet to meet a single researcher, public policy maker or domestic violence intervener who, like myself, has read all of the SARP studies in their entirety. If you have read the all the SARP studies in their entirety I invite you to contact me and tell me where you find any empirical evidence-based data that can demonstrate arrest deters repeat abuse for all domestic violence incidents.

Despite the complete absence of any data in the SARP concerning serious incidents some researchers labeled the results of the SARP as, “among the most influential results ever generated by social science (p. 3). Apparently, what these people choose to ignore is that the authors of the SARP write:

Thus, the full potential of SARP to answer questions about the specific deterrent effect of arrest and the safety of victims has not been realized [emphasis added] (p. 2).

My 21 years in law enforcement, my 15 years of research, my own personal experiences and most of the studies Klein cites to support his position, have lead me, once again, to conclude that law enforcement should make arrests for serious, injurious and repeat domestic violence incidents. While arrest, as the data here documents, may sometimes exacerbate rather than mitigate the problem, under certain egregious circumstance law enforcement can do no better.

However, the belief that arrest should be mandated, preferred, encouraged or be the default position in all domestic violence incidents is not only disingenuous, as the studies cited by Klein document , in some instances it can be dangerous.

It is my opinion, paradoxically supported here by the studies Klein cites to support his positions that arrest for all domestic violence incidents, regardless of context or circumstances, too often disempowers some victim, marginalizes and minimizes the context of serious and injurious interventions and can escalate rather than reduce the abuse.

And if we are sincere about reducing intimate partner abuse, regardless of gender or sexual orientation, we must end marginalizing, minimizing and excusing the coercive behavior and physical aggression that women initiate. 

Because it has been my lifelong belief that complaints about what others are doing is wrong without provide an opinion of what is right is to often the cowards way out of a difficult dilemma.


(1)The authority, without a mandate, preference, encouragement or default position of  arrest for all domestic violence incidents should continue. It is important that officers are provided the right to make arrests in witnessed or un-witnessed incidents.
(2) A mandated, preferred, encouraged or default position for arrest should remain in place for incidents that result in injury or for families with a history of chronic violence inside or outside the family or any history of criminal behavior.
(3)A mandated, preferred, encouraged or default position for arrest should remain in place for all felony cases. Felony cases most often include injuries and/or the use of weapons. Departments should not, as the Berkley, CA police department has, provide policy that turns a state misdemeanor into a city felony.
(4)Officers should always check their computer or with the dispatcher if they do not have a computer to ascertain if there have been prior calls at the same address or for the same offender of a victim. Repeat offenses regardless of severity need to be acted on. Officers should always check for outstanding warrants for any criminal behavior.
(5)Officer discretion – a position supported everywhere else in the criminal justice system -should be allowed for victim preference. The exception is where families, victims or offenders have records of chronic behavior or repeat calls to law enforcement. This does not mean that officers should or must adhere to victim preference, only that victim preference should be considered within the complete context and circumstance of individual incidents.
(6)Studies document that more often than not the offender has left by the time law enforcement arrives. In serious, injurious or repeat incidents officers should apply for arrest warrants where requested and/or when applicable.
(7)In minor incidents where there no injuries or there are no witnesses, officers should record all of the pertinent information and request a court date whereby both parties involved would be compelled to appear before a judge or a magistrate who can then determine the proper criminal or civil intervention for that family.
A list of criminal justice and social service agencies should always be provided by all relevant interveners. Social service agencies should be notified about the law enforcement response and those agencies should provide those families with a list of their resources.
(8)Law enforcement agencies should have computer programs in place that can document calls from chronically abusive high-risk couples. It has been found that a small number of chronically aggressive high-risk intimate partners produce a high-rate of repeat offenders. Follow-up interviews in studies document that about 8 percent of victims reported a total number of incidents that represented more than 82 percent of the 9,000 reported incidents. The vast majority of research documents that while domestic violence crosses all socioeconomic educations strata of society, most serious domestic violence is committed by classic high-rate offenders upon classic-high risk victims.
(9)Less than ½ of 1% of domestic violence incidents are lethal and there no effective tools to distinguish between lethal and non-lethal behavior, once arrested abusers should to be identified, using dangerousness assessment guides, for heightened risks of violent behavior. However, while these assessment tools can not effectively predict lethality they are very effective in recognizing risk factors concerning severe or chronic abuse and law enforcement should make use of them.
   (10) While this is last perhaps it should be number one. Female initiation is almost universally ignored because many interveners claim it amounts to blaming the victim. Education concerning female offending is important, if for no other reason than a growing number of studies that document the single best predictor for females being abused is their initiation of an abusive event. Often that initiation is coercive behavior or minor assault, slapping, etc. Recognizing the dangers of female initiation is not victim blaming. Victim safety should be paramount and the truth be told.

Copyright 2009 Richard L. Davis rldavis@post.harvard.edu



 Retirees Site