Home‎ > ‎Announcements:‎ > ‎

Tracking DV Offenders' Recidivism by Re arrest Rates

posted Feb 25, 2010, 4:11 PM by Vivien Bliss   [ updated Jun 3, 2010, 12:58 PM ]
Tracking recidivism via re arrest rates for domestic violence crimes is NOT appropriate.  This is because most DV perpetrators that don't have an overlap of other forms of rehearsed violence are not likely to be arrested again for a similar crime.  Domestic violence perpetrators maintain a focus on their family based on a belief system that supports the notion that they have a right to have power over their family to control them.  For those of us who have worked in the field for a while, this is common knowledge.  ENDING DOMESTIC VIOLENCE, at a minimum, takes a coordinated effort between prosecuters, judges, defense attorneys, police, and probation departments, with leadership from shelter based programs in collaboration with batterer intervention providers.  So, when you are asked, Does It Work?, you know the answer.  We know how to end domestic abuse & violence.  The challenge is:  We all have to stay committed and work together to hold offenders accountable & listen to the voices of survivors.
The following is an excerpt of an article regarding the appropriateness of  domestic violence offender evaluations.  This was written by Doug Bartholomew, located in Bellevue, Washington.  (for the full article, please contact the author).


The only reason that we could possibly have which would justify having a domestic violence intervention system is to reduce the risk of reoffense to current and potential victims. Everything we do or don’t do needs to be measured against that standard. If it doesn’t reduce reoffenses – and God forbid that it increase the risk – then we shouldn’t do it. Period. It’s that simple. Really.

Let’s begin with this hypothetical situation;

A man drinks too much beer at the Seahawks game and as he’s walking to his car stops to urinate in an alley. Unbeknownst to him, a group of school girls on a field trip walk past the alley and, as he turns to see what all the laughter is about, they see his penis.

1. Did he break the law by being drunk in public?

2. Did he break the law by urinating in public?

3. Did he break the law by exposing his genitals to underage girls?

Clearly the answer to all three questions is; yes! When he goes to court he would most likely be found guilty, since, in fact, he did those things.

Now comes the question of what to do with him. That requires knowing what the "problem" is.

1. Is he an alcoholic or alcohol abuser?

2. Is he a criminal?

3. Is he a sexual deviant?

Clearly the answer to these questions is; "I don’t know! Let’s have him evaluated."

There is a pathology of alcoholism and a pathology of alcohol abuse and there are tests and measures and definitions which can be used to determine whether or not the individual meets those criteria. There is also a socially acceptable level of drinking and if

Doug Bartholomew and Associates Inc. 1603 116th Ave NE, #120, Bellevue, WA, 98004 Phone (425-635-0188, FAX 425-451-8184, email doug@dbandainc.com 2

the person doesn’t meet the criteria for having the alcohol bug and their drinking is within socially acceptable limits, then they are okay. He broke the law, but he doesn’t need treatment.

Similarly, we have socially acceptable levels of sexuality and urination. There are a number of diagnosable sexual disorders, but, given that we have socially acceptable levels of sexuality and urination, an evaluation would be needed to determine whether or not the person had a diagnosable sexual disorder or whether their sexual behavior fell within acceptable social levels.

After decades of educating the public and the legal system about substance abuse, mental illness and sexual deviancy, the use of evaluations to determine the nature of intervention for individual charged with such crimes has become a common and reasonable expectation.

Consequently it makes sense that in domestic violence cases where an individual has assaulted or violated a partner in a manner which is domestic violence according to RCW 10.99.020 the courts would also ask for an evaluation to determine

1. Whether or not the assailant "is an abuser", i.e., does he have the disease of abuse.

2. Whether or not they "need treatment."

Unfortunately the same process does not lend itself well to the domestic violence situation for a number of reasons.



There is no socially acceptable level of domestic violence. At least I hope not.

Do we really want to have one? I want there to be a socially acceptable level of sexual behavior. Sexuality is a source of great pleasure, and personal fulfillment. It makes relationships better and can help us like ourselves more. Most of us see nothing wrong with socially acceptable levels of drinking, for example wine with dinner. Handled appropriately it, too, can be a source of pleasure and relaxation.

But do we want to define a level of assault or abuse which is acceptable? Does it lead to pleasure? Does it relax us? Does assault make relationships better? Is it a "right" to assault someone? Do we want to fight to defend our right to hit our loved ones?

Let’s take the example of our football fan. If he had been drinking but was below the legally drunk level, then that level of drinking is "okay". For many people beer, if handled appropriately, makes a football game more pleasurable.

Do we really want to say that about

any of the forms of domestic violence defined by law? Do we want to take the position that some forms of abuse are "okay"? Doug Bartholomew and Associates Inc. 1603 116th Ave NE, #120, Bellevue, WA, 98004 Phone (425-635-0188, FAX 425-451-8184, email doug@dbandainc.com 3

That is precisely what the statement of the evaluator and the court would be – and precisely how it would be heard by the public and by future offenders – if there was an assault in which the assailant was evaluated and it was determined that, although the act was abuse and was illegal, that no intervention was needed because that assault was socially acceptable.

In my clinical experience clients who come from communities or jurisdictions which practice selective enforcement and allow some offenders to get out of intervention because they were "evaluated" are very hard to treat. They all know someone who knows someone who committed the same act they did and got away with it because of an evaluation which determined that what they did was "okay". It doesn’t matter in the least what the court intended. That is what is heard.

And that is what they say when I try to treat them in group. In my community, some domestic violence is okay.

Evaluations which say that this act of violence or that act of violence doesn’t warrant treatment are a statement to that offender and to that community that that act of abuse is okay and can be done with impunity. It establishes a socially acceptable level of abuse, no matter how hard the court might try to explain it away otherwise.

As anyone who has ever raised children can tell you, whatever you forgive, you permit, and whatever you permit you endorse, and whatever you endorse, you encourage.

This is born out in the research on what happens if you arrest someone and don’t put them in treatment.

In fact, the above adage has scientific backing. In the two studies noted below it was found, based on thousands of arrests in several different cities over long periods of time that if you arrest someone and don’t put them in treatment you are increasing the risk of more violence.


Effect of arrest without treatment on reoffense rates (what happens if they are arrested and not treated)

Schmidt and Sherman (1993)

"arrest (alone) increased domestic violence recidivisim among suspects in Omaha, Charlotte, and Milwaukee…over the course of one year those arrests (only) doubled the rate of violence by the same suspects."

Dunford, Huzinga, and Elliott (1990)

"…arresting (alone) and the immediate period of custody associated with arrest was not a deterrent to continued domestic violence." (parenthetical notes mine)





A musician is one who makes music.

A singer is one who sings.

A swimmer is one who swims.

An offender is one who…

For over twenty years there have been many large, sophisticated research studies performed attempting to find out what the profile of an offender is (see below for some of the most significant ones). Study after study found that the population of offenders consists of many subgroups. These subgroups, unfortunately, are, in general terms, about the same proportions you would find in the general population.

Even more significant, no study has found that more than half of the offenders have

any clinically significant, measurable pathologies.

In other words, battering is a common behavior in our society and many offenders are perfectly normal individuals doing something that their society tells them is OK. Not monsters, not deviants, not diagnosable people.

Hamberger, et al (1996), in a study of 833 court ordered offenders, found that 40% showed no elevations on any of the clinical scales of the MCMI. Their "Cluster 3" offender, as in Holtzworth-Monroe and Stuart (1994), was socially desirable, was isolated to family-only violence, experienced relatively little child abuse or witnessed parental violence, was not distinguishable from the general population on depression (Beck), or anger (Novoco), or police records,

but nevertheless committed domestic assaults on an average of once per month. While this is less than the other subtypes, this is still an unacceptable rate. Their violence was restricted to loved ones. Thus if a person is socially desirable with a relatively clean record and no measurable pathology the best that can be said is that they are likely to reoffend approximately monthly.

White and Gondolf (2000) found that 56% of offenders had no discernable traits which distinguished them from the general population.

There is no predictive value of the MCMI-II on subsequent violence (Jacobsen, et al, 1996). Lohr, Hamberger, and Bonge (1988) found that 39% of their sample of 196 violent men was undetectable using the MCMI-II.

Holtzworth-Monroe and Stuart (1994) identified three groups of offenders by personality type, one of which was "non-pathological", i.e., no diagnosable or detectable disorder. They were typified as being "…low severity of violence, low generality of violence, low criminal involvement, low-to-moderate depression and alcohol abuse, and moderate

Doug Bartholomew and Associates Inc. 1603 116th Ave NE, #120, Bellevue, WA, 98004 Phone (425-635-0188, FAX 425-451-8184, email doug@dbandainc.com 5

levels of anger." However these non-pathological offenders only differed from the pathological offenders in the

type of violence performed (it tends to be isolated to loved ones and family members) and in the type of ancillary services needed, such as substance abuse, not by an absence of reoffense.

In one of the largest and most recent pieces of research "Evaluating batterer counseling programs: a difficult task showing some effects and implications" by Ed Gondolf, he found in his research on reoffenders, that;


…their personality profiles did not distinguish them from men who reassaulted only one time, and men who did not reassault (Gondolf and White, 2001). Their patterns of violence, based on coding of the women‟s narratives, also did not distinguish these particularly dangerous men (Gondolf and Beeman, 2003)."

"In summary, the repeat reassaulters did not appear as a distinct "batterer type." The most distinguishing factor was the lack of response to these men. There partners were less likely to take action, possibly out of fear or subjection; and further arrests, protection orders and treatment write not as likely, as a laboratory study found with more antisocial men (Jacobson and Gottman,. 1998."

Sonkin, (2003) added as frightening note to this issue when he found that

"the personality disorder most likely to actually kill his spouse is dependent and passive-aggressive, not the profile predicted by these scales."

This was substantiated by Dutton, D. & Kerry, G, (1999) where

it was also found that contrary to common sense, men who kill their wives are more likely to be passive-aggressive than antisocial Study

Percent of offenders who were "normal" or indiscernible from the general population; who would normally be seen as "low risk" or "no risk"

Lohr, Hamberger, and Bonge (1988)


Hamberger, et al (1996)


White and Gondolf (2000)