Content uploaded by Michael Johnson
Author content
All content in this area was uploaded by Michael Johnson on Nov 07, 2017
Content may be subject to copyright.
This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
Author's personal copy
Gender and types of intimate partner violence: A response to an anti-feminist
literature review
Michael P. Johnson ⁎
The Pennsylvania State University, USA
abstractarticle info
Available online 12 April 2011
Keywords:
Domestic violence
Feminism
Intimate partner violence
Coercive control
Family violence
Partner abuse
This article presents a feminist perspective on domestic violence that is rooted in an explication of the
differences among three major types of intimate partner violence (Johnson, 2008). Theory and research from
this perspective is then reviewed to rebut recent attacks on feminist scholarship and policy regarding intimate
partner violence.
© 2011 Elsevier Ltd. All rights reserved.
Contents
1. Introduction .............................................................. 290
2. A feminist perspective on domestic violence .............................................. 290
2.1. A feminist perspective on types of intimate partner violence ................................... 290
2.1.1. Intimate terrorism ................................................... 290
2.1.2. Violent resistance ................................................... 290
2.1.3. Situational couple violence ............................................... 290
2.2. A feminist perspective on sampling biases ............................................ 290
3. The anti-feminist backlash ....................................................... 291
3.1. Misrepresentations of the general feminist analysis of domestic violence . . ............................ 291
3.1.1. General allegation #1: feminists say that only men do it ................................. 291
3.1.2. General allegation #2: feminists say violent men are evil, violent women are good ..................... 291
3.1.3. General allegation #3: feminists say that the only cause of intimate partner violence is the patriarchy ........... 291
3.1.4. General allegation #4: the feminist mindset has a lock on a variety of institutions, especially the law ............ 292
3.2. Misrepresentations of specific work: the case of Kelly and Johnson ................................ 292
3.2.1. Specific allegation #1: Kelly and Johnson ignore the data on gender symmetry ...................... 292
3.2.2. Specific allegation #2: Kelly and Johnson misrepresent female situational couple violence ................. 292
3.2.3. Specific allegation #3: Kelly and Johnson deny female intimate terrorism ......................... 292
3.3. Dutton et al.'s review of research that allegedly contradicts the feminist analysis .......................... 292
3.3.1. Reality check for the gender paradigm .......................................... 292
3.3.2. Old wine in new bottles ................................................ 293
3.3.3. The myth of equivalent methodological bias ....................................... 293
3.3.4. Shelter to general population extrapolation ....................................... 293
3.3.5. Non-selective sample studies .............................................. 293
4. Conclusion .............................................................. 295
References ................................................................. 295
Aggression and Violent Behavior 16 (2011) 289–296
⁎1155 Oneida St., State College, PA 16801, USA. Tel.: +1 814 237 8061.
E-mail address: mpj@psu.edu.
1359-1789/$ –see front matter © 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.avb.2011.04.006
Contents lists available at ScienceDirect
Aggression and Violent Behavior
Author's personal copy
1. Introduction
The most recent of a series of anti-feminist attacks from Dutton,
Hamel, and their colleagues is “The gender paradigm in family court
processes: Re-balancing the scales of justice from biased social
science”(Dutton, Hamel, & Aaronson, 2010), an ironic title, given the
panoply of biases with which it itself is riddled. In this particular
article they claim to expose two recent papers (Jaffe, Johnston, Crooks,
& Bala, 2008; Kelly & Johnson, 2008) as biased and unsupported by
research evidence. Responding to this particular attack is useful in
itself, but their article also serves as a good example of the substance
and tactics of their more general anti-feminist critique. In the process
of responding here to their allegations about feminist theory and
research, I hope to accomplish two goals. First, I will present a feminist
perspective on domestic violence that is rooted in an explication of
the differences among the major types of intimate partner violence
(Johnson, 2008). Second, theory and research from this perspective
will be used to rebut the Dutton et al. claims about what they call “the
gender paradigm,”which includes my own work.
2. A feminist perspective on domestic violence
It is probably useful to begin by saying that there is more than one
feminist understanding of the nature of domestic violence, more than
one “gender paradigm,”just as there are multiple feminist perspec-
tives on anything. What I will present here is my feminist perspective
on the nature of intimate partner violence, a perspective formed
primarily from a wide reading of over thirty years of research on
“domestic violence,”and informed by feminist perspectives from my
home discipline of sociology.
2.1. A feminist perspective on types of intimate partner violence
The core proposition of this perspective is simple: there is more
than one type of intimate partner violence, and the major types differ
dramatically in almost all respects (Johnson, 2008). The typology that
I began developing in the early 1990s is organized around the concept
of coercive controlling violence, a pattern of behaviors identified by
feminists working in the battered women's movement as the type of
intimate partner violence that was reported by women coming to
shelters to seek help (Pence & Paymar, 1993). There are three major
types.
2.1.1. Intimate terrorism
This is the pattern of violent coercive control that comes to mind
for most people when they hear the term “domestic violence”.
Although it probably represents a small part of all of the violence that
takes place between partners in intimate relationships, it is the type of
violence that predominates among the cases that come to the
attention of law enforcement, shelters and other public agencies,
and that therefore has been the prototype of domestic violence for the
battered women's movement (see almost any shelter Web site). It
involves the combination of physical and/or sexual violence with a
variety of non-violent control tactics, such as economic abuse,
emotional abuse, the use of children, threats and intimidation,
invocation of male privilege, constant monitoring, blaming the victim,
threats to report to immigration authorities, or threats to “out”a
person to work or family.
Although this is the type of violence initially identified by the
battered women's movement as characteristic of the male violence
encountered in shelters and law enforcement, it is not exclusively
male-perpetrated, having been identified in lesbian couples (Renzetti,
1992) and among some women who terrorize their male partners
(Cook, 1997; Hines & Douglas, 2010). The data are clear, however, that
the primary perpetrators in heterosexual couples are men (Graham-
Kevan & Archer, 2003; Johnson, 2006a, 2008). It is also clear from the
research of Holtzworth-Munroe, Meehan, Herron, Rehman, & Stuart
(2000) and from a major literature review (Sugarman & Frankel,
1996) that misogyny and gender traditionalism play an important
role in heterosexual intimate terrorism.
1
2.1.2. Violent resistance
Many victims of intimate terrorism do respond with violence of
their own. For some, this is an instinctive reaction to being attacked,
and it happens at the first blow—almost without thought. For others, it
doesn't happen until it seems that the assaults will continue forever if
something isn't done to stop them. For most women in heterosexual
relationships, the size difference between them and their male
partner ensures that violent resistance won't help, and may make
things worse, so they turn to other means of coping. For a few,
eventually it seems that the only way out is to kill their partner.
2.1.3. Situational couple violence
This is violence that is not part of a general pattern of coercive
control, but rather occurs when couple conflicts become arguments
that turn to aggression that becomes violent. It is by far the most
common form of intimate partner violence, and also the most
variable. Somewhere around 40% of the cases identified in general
surveys involve only one relatively minor incident, but many cases do
involve chronic and/or serious, even life-threatening, violence. In
contrast to intimate terrorism, situational couple violence does not
involve an attempt on the part of one partner to gain general control
over the other, and unlike intimate terrorism and violent resistance it
is roughly gender-symmetric in terms of perpetration. The violence is
situationally-provoked, as the tensions or emotions of a particular
encounter lead one or both of the partners to resort to violence.
2.2. A feminist perspective on sampling biases
Here is another simple proposition: all of our major sampling
methods are biased, with the result that they yield samples that differ
dramatically in the representation of the major types of intimate
partner violence. So-called random sample surveys are biased because
of high rates of non-response, beginning with non-response to the
brief screening interview for eligibility that often precedes the request
for a full interview. Response rates often do not reflect that initial
refusal to answer even the screening questions. For example, the
National Family Violence Surveys that report an 82% response rate
actually have a 60% response rate if non-response to the screening
questions is included (Johnson, 1995). Because intimate terrorism and
violent resistance have low base rates to begin with, and because
perpetrators and victims of intimate terrorism are highly likely to
refuse to respond to surveys –perpetrators because they do not wish
to implicate themselves, victims because they fear reprisals from their
partner –the violence in general surveys is heavily dominated by
situational couple violence.
Agency studies are biased not by non-response as much as by the
nature of the sampling frame itself. Because only serious or chronic
violence tends to come to the attention of law enforcement, shelters,
hospitals, and other such agencies, the violence in agency data or in
surveys conducted in these settings is heavily biased in the direction
of intimate terrorism and violent resistance. Similar biases are found
in help lines, voluntary on-line databases, and other sources of
information that involve safe self-reporting, but the general point here
290 M.P. Johnson / Aggression and Violent Behavior 16 (2011) 289–296
1
Although the Sugarman and Frankel meta-analysis found a strong relationship
between gender traditionalism and male intimate partner violence (d = .54), more
telling is an important interaction effect that they do not include in their major
conclusions. The relationship between gender traditionalism and intimate partner
violence is quite strong in samples that are likely to be dominated by intimate
terrorism (d= .80) and tiny for samples that are likely to be dominated by situational
couple violence (d = −.14).
Author's personal copy
is that the sampling frame of every study in a specific institutional
setting has a specific set of processes that shape the balance of types of
violence that enter it.
The biases of these major approaches to sampling in intimate
partner violence research are the major source of the seemingly
contradictory data that continue to maintain the gender symmetry
debate. Those who believe in gender symmetry cite hundreds of
general survey studies that show that women perpetrate intimate
partner violence at least as often as men. On the other side, believers
in male perpetration of intimate partner violence cite hundreds of
agency studies that show that men are the primary perpetrators.
Studies with mixed samples that give access to all three major types of
intimate partner violence, and that make distinctions among the
types, find that intimate terrorism and violent resistance are heavily
gendered, and that situational couple violence is perpetrated about
equally by men and women—and it is this pattern, combined with
sampling biases, that explains the dramatic differences among various
studies with regard to the issue of gender symmetry. Surveys,
dominated by situational couple violence, show rough gender
symmetry in perpetration. Agency studies, dominated by intimate
terrorism and violent resistance, show a pattern of (primarily) male
violent coercive control and female resistance.
3. The anti-feminist backlash
The Dutton et al. (2010) paper to which I am responding
exemplifies all of the general strategies of recent attacks on the
progress of the battered women's movement and on the research that
confirms that the feminist analysis of “domestic violence”(intimate
terrorism) is largely correct.
2
In this particular paper they begin with
an attack on their own caricature of the feminist analysis (which they
call “the gender paradigm”), then move on to a rebuttal of what they
allege has been said in two papers published in a 2008 issue of Family
Court Review (Jaffe et al., 2008; Kelly & Johnson, 2008). I will focus
here on their general misrepresentations of feminism and on their
attack on the paper I wrote with Joan Kelly.
3.1. Misrepresentations of the general feminist analysis of domestic
violence
3
3.1.1. General allegation #1: feminists say that only men do it
The misrepresentations begin in the first paragraph of the paper:
“This view [what Dutton et al. label as “the gender paradigm”] holds
that consequential IPV is an exclusively male-perpetrated crime
against female victims and children”(p. 2).
4
This caricature continues
at the beginning of the “analysis”section: “…IPV is viewed as mainly
male perpetrated against female victims”(p.3). Perhaps the authors
are confusing the early rhetoric of the battered women's movement
with contemporary feminist analyses of intimate partner violence. It is
important to remember that feminist analyses began with an
understanding of what was happening to women victims who came
to the attention of law enforcement and who contacted shelters and
help lines in the 1970s. The pattern of violent coercive control that
dominated those cases came to be labeled by the movement as
“domestic violence,”creating considerable confusion because family
sociologists continued to use that term for any violence between
intimate partners, not just the coercive controlling violence that I have
labeled as intimate terrorism.
More “recent”feminist analyses (since the early 1990s) have
stressed the differences among the three major types of intimate
partner violence that I have just discussed above. First, it should be
noted that this framework does not dismiss the existence of female-
perpetrated intimate terrorism. Second, violent resistance, which
involves mostly women, is acknowledged to lead sometimes to very
serious violence, including homicide. Third, I and others have always
noted that situational couple violence (a) is far and away the most
common form of intimate partner violence, (b) is perpetrated about
equally by men and women, and (c) can be extremely consequential.
Of course, we feminists, and all other family sociologists for that
matter, have always noted that male violence is more likely to
produce injuries and fear than is women's violence. That observation
of a relationship between gender and the consequences of violence
certainly does not translate into Dutton et al.'s cartoonish version of
the feminist analysis as arguing that all consequential intimate
partner violence is male-perpetrated.
3.1.2. General allegation #2: feminists say violent men are evil, violent
women are good
Here is how Dutton et al. present this pair of ideas, which they
attribute to the so-called gender paradigm: “Men are presented as
intentionally perpetrating domestic violence in order to maintain
power and control in family relationships. In contrast, female violence
is rationalized as the result of external circumstances—primarily as a
reaction to male oppression”(p. 3). Of course, the point of the analysis
I presented above is that most intimate partner violence does not
involve an attempt on the part of either partner to exercise coercive
control—it is situational couple violence. The feminist analysis does
argue that in heterosexual relationships most of the intimate
terrorists are men, but also that most of the violent men are not
intimate terrorists.
As with the men, most women are involved in situational couple
violence—no control issues, no reaction to male oppression, just
arguments that escalate for a variety of reasons that differ from couple
to couple. The feminist analysis also recognizes the existence of a few
female intimate terrorists, and they certainly are not characterized as
reacting to male oppression. The authors have purposely or
inadvertently expanded the relatively small number of women who
are involved in violent resistance into a group that encompasses the
entire feminist analysis of women's violence.
3.1.3. General allegation #3: feminists say that the only cause of intimate
partner violence is the patriarchy
Here's how Dutton et al. put it: “Various empirically demonstrated
etiological contributions to IPV (e.g., learning, attachment, and
personality) are ignored, as are correlates of IPV perpetration such
as alcohol abuse, depression, reported interpersonal dominance
between partners (regardless of gender), and dyadic communication
deficits”(p. 3). Well, let's start where one might expect the feminist
analysis to be most single-minded—intimate terrorism. For more than
a decade (Johnson & Ferraro, 2000) my feminist colleagues and I have
incorporated Holtzworth-Munroe's work into our analysis of intimate
terrorism, work that centers on matters of personality in general, and
attachment in particular. And my work with Alison Cares demon-
strates the relationship between violence in one's childhood home
and male intimate terrorism (Johnson, 2008; Johnson & Cares, 2004),
a central tenet of the learning approach to understanding intimate
partner violence. As I have noted above, and in a number of published
pieces, substance abuse and couple communication issues are central
to any analysis of situational couple violence (Johnson, 2006b, 2007),
and my analyses of situational couple violence have always
emphasized the extreme variability of its causes.
The authors go on in this section to cite my concept of violent
resistance as a major example of how the feminist analysis even
attributes women's violence to men's coercive control. They neglect to
291M.P. Johnson / Aggression and Violent Behavior 16 (2011) 289–296
2
I will focus on the research critique here, but the attacks also include attempts to
undermine funding for services for battered women by arguing that they discriminate
on the basis of gender (Dragiewicz, 2008; Dragiewicz & Lindgren, 2009).
3
There have been a number of other responses to these attacks on the feminist
analysis (e.g., DeKeseredy & Dragiewicz, 2007; Dragiewicz & Lindgren, 2009; Frieze,
2000; Holtzworth-Munroe, 2005; Johnson, 2005a, 2005b, 2005c; Kimmel, 2002).
4
Page numbers without a reference all refer to Dutton et al., 2010.
Author's personal copy
point out that violent resistance accounts for only a small part of
women's intimate partner violence in my typology, in which most of
women's violence is situational couple violence with much the same
causes as men's situational couple violence.
3.1.4. General allegation #4: the feminist mindset has a lock on a variety
of institutions, especially the law
Certainly the feminist analysis has had an important impact on
major institutions (Buzawa, 2003; Dobash & Dobash, 1992), in-
stitutions that until the 1970s had generally assumed that all intimate
partner violence was situational couple violence, a private matter for
the couple to work out themselves. The feminist analysis produced a
major shift in perspective, one in which most, if not all, domestic
violence was seen as intimate terrorism. In presenting that out-of-
date kernel of truth, however, the authors seem to have missed the
major paradigm shift that has been sweeping those institutions in the
last ten years: differentiation is the new catchword in the courts and
law enforcement (see the bibliographies in Jaffe et al., 2008; Kelly &
Johnson, 2008).This oversight is puzzling, because two major
examples of that paradigm shift are the very articles that they criticize
in Dutton et al. as examples of the death grip of the feminist focus on
patriarchy (Jaffe et al., 2008; Kelly & Johnson, 2008). Both of those
articles are rooted in the idea that not all intimate partner violence is
intimate terrorism, and that the courts and other institutions need to
use all of the assessment tools at their disposal to identify what type of
intimate partner violence is involved in each particular case in order
to decide on an appropriate course of action.
3.2. Misrepresentations of specific work: the case of Kelly and Johnson
3.2.1. Specific allegation #1: Kelly and Johnson ignore the data on gender
symmetry
The authors accuse Joan Kelly and me of purposely ignoring the
data on gender symmetry: “By thus omitting significant similarities in
the actual incidence of male- and female-perpetrated domestic
violence, victim advocates and allied researchers [this is in the section
on my typology] present truncated, empirically skewed and data-
poor, sometimes emotionally charged, stereotypic interpretations of
IPV….”(p. 6). What we actually say is the following (Kelly & Johnson,
2008, pp. 486–487): “Situational Couple Violence is initiated at similar
rates by men and women, as measured by large survey studies and
community samples…. Overall, these and other survey data support
claims that women both initiate violence and participate in mutual
violence and that, particularly in teenage and young adult samples,
women perpetrate violence against their partners more frequently
than do the men.”We had already stated clearly that situational
couple violence is the most common type of intimate partner violence.
3.2.2. Specific allegation #2: Kelly and Johnson misrepresent female
situational couple violence
“For the most part Johnson…relegates female IPV to the category
‘situational couple violence’…. Female-instigated, conflict-engen-
dered SCV is cast as an understandable reaction to male SCV…”
(p. 6). There are no quotes to back up this distortion of what we
allegedly said, and there are two important misrepresentations in this
two-part characterization. The first distortion is somewhat subtle in
that the authors simply chose not to note that for the most part we
relegate all IPV, female and male, to situational couple violence. The
second distortion is more direct: far from excusing women's
situational couple violence, we do not distinguish between men's
and women's situational couple violence in terms of causes (Kelly &
Johnson, 2008, p. 485): “Situational Couple Violence results from
situations or arguments between partners that escalate on occasion
into physical violence. One or both partners appear to have poor
ability to manage their conflicts and/or poor control of anger.”We
never (nor have I anywhere else) characterized women's situational
couple violence as an “understandable reaction”to anything.
3.2.3. Specific allegation #3: Kelly and Johnson deny female intimate
terrorism
Here Dutton et al. first damn with faint praise: “As Johnson
occasionally acknowledges (Johnson [2006a], fn 2), most but not all
severe IPV is perpetrated by men”(p. 7). Of course, the choice of a
footnote seems to imply that I hide even those few times that I
acknowledge that women sometimes perpetrate intimate terrorism.
The truth is that there is not a single piece among the dozens of my
papers and my book in which I did not acknowledge that there are
female intimate terrorists, beginning with the earliest published
paper on the typology, in which I was still using the term “patriarchal”
terrorism and was focused almost entirely on men's violence
(Johnson, 1995, p. 292): “Although it is indisputable that some men
are terrorized by their female partners (I have worked with some at
my local shelter), [the argument] that men are terrorized as
frequently as women produces a dangerous distortion of reality.”
Not satisfied with this subtle distortion, Dutton et al. follow with a
more blatant lie, alleging that in my work “…by definition, male
violence is internally caused by the conscious intent to dominate
women”(p. 7). No—I have repeatedly stated that most male violence
is situational couple violence.
3.3. Dutton et al.'s review of research that allegedly contradicts the
feminist analysis
3.3.1. Reality check for the gender paradigm
Throughout much of this section Dutton et al. “refute”our
argument that intimate terrorism is primarily male-perpetrated by
citing survey data on gender symmetry. As Kelly and I noted, research
conducted by me and my colleagues, and by other scholars (Ansara &
Hindin, 2010; Graham-Kevan & Archer, 2003; Johnson, 2006a;
Johnson, Leone, & Xu, 2008) has demonstrated that the violence in
survey research is so heavily dominated by situational couple violence
that it provides no information regarding the gender symmetry or
asymmetry of intimate terrorism.
The authors also claim that there is no evidence that men are the
major perpetrators of intimate terrorism. My colleagues and I have
presented such evidence from Irene Frieze's Pittsburgh study
(Johnson, 2001, 2006a) and from data on previous marriages from
the U.S. National Violence Against Women Survey (Johnson et al.,
2008). Graham-Kevan and Archer (2003) have presented confirming
data from their study in Britain—different researchers, different
measures, and different populations. And of course there are decades
of research from law enforcement, shelters, hospitals and other
agencies (in which the data are dominated by intimate terrorism) that
show that men are the primary perpetrators of intimate terrorism in
heterosexual relationships.
Next, Dutton et al. cite Graham-Kevan and Archer (2003) as
evidence for the gender symmetry of intimate terrorism, and later
they argue in a particularly nasty way that Kelly and I have
misinterpreted the Graham-Kevan and Archer data: “[Kelly and
Johnson] misrepresent Graham-Kevan and Archer's (2003) findings…
and cherry pick and distort the data….”(p. 13). The alleged
“distortion”to which they refer is that we do not point out that,
although 87% of the intimate terrorism in Graham-Kevan and Archer's
data is male-perpetrated, most of these cases come from their shelter
sample. Here is how Dutton et al. summarize the gender and intimate
terrorism findings in the Graham-Kevan and Archer study: “… all
[non-shelter] groups, including a group of men court-mandated for
spouse assault treatment, exhibited gender symmetry in incidence of
[intimate terrorism]….”(p. 13, my emphasis). Here are the actual data
for all of the groups. In the shelter sample (n = 68), there were 36
male and one female intimate terrorists. Among students (n = 56),
292 M.P. Johnson / Aggression and Violent Behavior 16 (2011) 289–296
Author's personal copy
there were seven male and two female intimate terrorists. Among
prisoners (n= 105), there were two male and five female intimate
terrorists. For the batterer intervention program (n= 10), there were
no male or female intimate terrorists. Do they really consider this to
be evidence for the gender symmetry of intimate terrorism?
Finally, they cite Laroche's study (2005) using Canadian General
Social Survey data. The problem with the Laroche study (much as I
admire it) is that he made the mistake of using the cutoffs for intimate
terrorism that Janel Leone and I had mistakenly adopted in our earlier
research on current relationships in the NVAW (Johnson & Leone,
2005). Our mistake was relying on a cluster analysis in a survey
sample of current relationships, a sample that essentially includes no
intimate terrorism. When we corrected our analysis by applying the
cluster analysis to previous marriages (where there is a reasonable
number of cases of intimate terrorism), we obtained a different cutoff
and found that 83% of the intimate terrorism in previous relationships
is male-perpetrated, and for current marriages only 6/10 of one
percent of respondents report intimate terrorism Johnson et al., 2008.
3.3.2. Old wine in new bottles
The gist of this section is that this is “the typology of one researcher
(Johnson), whose linking of gender and [intimate terrorism] …has
been disconfirmed by relevant empirical research finding female as
well as male [intimate terrorists]”(p.8). Of course, it is not I, but
Dutton et al. who conflate type of violence and gender in their
caricature of my work. I have never said there were no female
intimate terrorists, only that in heterosexual relationships intimate
terrorism is primarily male-perpetrated. In fact, my own studies cited
above include female intimate terrorists (3% of the intimate terrorists
in the Pittsburgh data, 17% among NVAW previous marriages).
Dutton et al. then go on to present largely irrelevant data on the
distribution of bilateral, male-initiated, and female-initiated violence
in survey data—data which make no distinctions among types. Of
course, those data are dominated by situational couple violence,
which is gender-symmetric.
3.3.3. The myth of equivalent methodological bias
In this section Dutton et al. claim that there is no evidence that so-
called representative sample surveys are heavily biased in the
direction of finding primarily situational couple violence and very
little intimate terrorism. Well, in the general sample of the Pittsburgh
study,themaleviolencewas89%situationalcoupleviolence
(Johnson, 2006a), among current marriages in the NVAW it was 85%
situational couple violence (Johnson et al., 2008). For Graham-Kevan
and Archer's British general sample, the male violence was 75%
situational couple violence. For Ansara and Hindin's (2010) Canadian
data, the male violence among current partners was 81% situational
couple violence. The point is two-fold: (a) general survey data
without distinctions will show patterns that are characteristic of
situational couple violence, and (b) therefore we need to develop
standard operationalizations of types that will allow us to make
distinctions in various types of samples.
3.3.4. Shelter to general population extrapolation
“…Johnson based his typology solely on self-reports from women
in shelters [and] erroneously generalized his findings to the
distribution of IPV in the broader community”(p. 9). Wrong! The
Pittsburgh sample is self-reports from women, but from shelters,
courts, and the general community. The differences between intimate
terrorism and situational couple violence have also been documented
in a U.S. national sample (Johnson et al., 2008; Johnson & Leone,
2005), in a Chicago health service sample (Leone, Johnson, & Cohan,
2007; Leone, Johnson, Cohan, & Lloyd, 2004), in a large multi-city
study in the U.S. (Frye, Manganello, Campbell, Walton-Moss, & Wilt,
2006), in two studies of different Canadian General Social Surveys
(Ansara & Hindin, 2010; Laroche, 2005), and in a multi-sample study
in Britain (Graham-Kevan & Archer, 2003).
As far as I know, I have never argued that “…community sample
methodology yields data as selective and biased as that collected from
shelter houses”(p. 9). Again, Dutton et al. offer no quotes, but in the
2006 paper that they cite for this statement I actually said, “Finally, let
me nail down my sampling argument, that general survey samples tap
primarily situational couple violence, whereas agency samples give
access primarily to intimate terrorism…. [M]ost couples who
experience violence, including those in our audiences, are involved
in situational couple violence…. [O]ne can err by assuming that the
patterns observed in agency samples describe all partner violence”
(Johnson, 2006a, pp. 1010–1011). In the same article, here is how I
spoke about the issue of extrapolation: “If my arguments regarding
the biases of various types of sampling strategies are correct, it is
almost impossible to develop precise estimates of the incidence of the
various types of violence. I come to the conclusion that most partner
violence is situational couple violence in the following way, based on
figures in my 1995 article. First, accepting my evidence that almost all
of the partner violence in general surveys is situational couple
violence, we can use the figures from the National Family Violence
Surveys to estimate the incidence of situational couple violence.
Second, extrapolating from agency data in two states that keep
excellent shelter statistics, we can develop an estimate of the
incidence of intimate terrorism. Those figures, which may be found
in the 1995 article, suggest that there is probably 3 times as much
situational couple violence as intimate terrorism, which would mean
that 75% of women experiencing violence from their male partners are
experiencing situational couple violence”(Johnson, 2006a, p. 1016,
fn. 11).
3.3.5. Non-selective sample studies
In the next, long section the authors cite studies that are alleged to
show that (a) women are as violent as men, (b) women are as
controlling as men, and (c) women are as likely to be intimate
terrorists as men. I intend to try to go through this section study by
study, but before I go there, let me simply tell you what the three
major problems are with the uses to which Dutton et al. put the
studies that they cite. First, almost all of the studies they cite are
general surveys, studies that have been demonstrated to represent
situational couple violence, a type of violence that we all agree is
roughly gender-symmetric in terms of perpetration.
Second, a number of the studies they cite show that men and
women are equally controlling or dominant in relationships. That, of
course, is irrelevant to the issue of whether men and women are
equally likely to use a combination of violence and non-violent control
tactics to attempt to take complete control over their partner—
intimate terrorism.
Third, a few of the studies they cite are alleged to show that
women are as likely as men to combine violence and control, i.e., to be
intimate terrorists. There are two responses to these studies. On the
one hand, some of them involve general surveys that show similar
correlations between control motive and violence for men and
women. With respect to these, I need to remind Dutton, his
colleagues, and the other anti-feminists, that this typology is a
typology—for a reason. I believe that what is involved here is not a
simple correlation between control motive and violence, for either
men or women. What we have is a small group of people, men and
women but mostly men, who attempt to take total control over their
partners with a combination of violence and other control tactics, i.e.,
violent coercive control. These are the intimate terrorists. Among other
couples, we will sometimes see a relationship between control motive
and situational couple violence because those with a stronger need for
control will be more likely to escalate arguments when they feel they
are not winning. And the studies that Dutton et al. cite are general
293M.P. Johnson / Aggression and Violent Behavior 16 (2011) 289–296
Author's personal copy
surveys that are dominated by situational couple violence, the other
couples.
In one important case Dutton et al. do misrepresent a study's
findings regarding control and violence. They say that Felson and
Outlaw (2007) find that “…the relationship between use of control/
jealousy and physical violence exists equally for both male and female
respondents….”(p. 11). What Felson and Outlaw actually found was a
strong interaction effect in which there is no relationship between
control and violence for current partners (as I would predict because
survey data on current partners include no intimate terrorism), but a
strong relationship between control and violence only for men for
previous partners (as I would predict for a sample that includes
intimate terrorism). As my colleagues and I have recently argued, this
is actually compelling evidence for our feminist analysis of intimate
partner violence (Johnson et al., 2008).
Now for the specific studies.
5
Dutton et al. start with four studies
that find no gender differences in emotional abuse and control. These
are general survey studies—situational couple violence. Then come
two articles that show similar correlations between dominant
personality, the need to control, and physical violence for men and
women. One is a general survey, the other a convenience sample of
undergraduates—situational couple violence. Next are two studies
that allegedly show that males and females are equally likely to
combine physical violence with emotionally abusive and controlling
behaviors. As far as I can tell, neither of the studies presents evidence
on that matter. Perhaps Dutton et al. are under the mistaken
impression that similar correlations between those variables indicate
a similar likelihood of combining them. It does not. But more
importantly, one of the studies is of high school students, the other
of college students—situational couple violence. Then comes Straus's
big study of dating university couples, showing that dominance scores
for men and women are equal and that female dominance predicts
female physical aggression. Another general survey—situational
couple violence.
Dutton et al. now turn to some studies of what they call clinical
populations. Here, depending on the institutional setting, we may find
some intimate terrorism. The first study they cite seems to me to
provide support for the feminist analysis. Judge for yourself: “In the
pioneering studies by Stacey, Hazlewood, and Shupe (1994), on men
arrested for domestic violence and mandated to batterer intervention
programs, one-third of the physical violence was perpetrated by the
female partner (legally deemed the ‘victim’), and rates of male-
perpetrated emotional abuse and control were significantly higher
than female partner rates in only about half of the 13 categories”
(p. 11). This sounds to me like some situational couple violence and
some intimate terrorism/violent resistance (the higher emotional
abuse and control among the men would suggest that they were the
intimate terrorists, although it is impossible to know without looking
at patterns rather than average differences in individual categories).
At the very least, these findings are hardly incompatible with the
feminist analysis.
The second study cited in this category provides no data on the
types, but finds “equivalent rates of injury-causing physical violence
among couples dual-arrested for domestic violence, with men more
likely to engage in isolation behaviors and women somewhat more
likely to engage in verbal abuse”(p. 11). One of the most important
points of my typological approach is that all of our samples are biased
in terms of types and we need to ask what those biases are. These
were dual arrests!
The final study in this trio is a 2008 conference presentation
(Hamel, Graham-Kevan, & Prospero, 2008) that finds “comparable
levels of controlling and emotionally abusive behaviors by male and
female clients court-mandated to batterer intervention programs
across California”(p. 11 of Dutton et al.). It is possible that they have
both male and female intimate terrorists in this sample, and I would
like to know how many there were of each. Comparable levels do not
address the question of comparable numbers. In our own study of
intimate terrorism among previous spouses in the NVAW (Johnson
et al., 2008) we found that most of the intimate terrorism was male-
perpetrated, but that control tactics for the much smaller number of
female intimate terrorists were in many ways similar to those used by
the men.
The next study cited is the Felson and Outlaw (2007) study
discussed above, that actually shows strong support for our position
that intimate terrorism is mostly male-perpetrated (see also our more
appropriate analysis of the same data: Johnson et al., 2008). Then
comes Laroche, also discussed above (Section 3.3.1).
Next Dutton et al. cite a series of findings from general surveys—
situational couple violence. They seem to believe, somehow, that this
evidence that women often initiate violence against a non-violent
partner is evidence that “the IPV profiles in these surveys involved
significant female-perpetrated ‘abusive-controlling’violence [inti-
mate terrorism]….”(p. 13). Of course, there is no such evidence, as
Kelly and I noted: “Overall, these and other survey data support claims
that women both initiate violence and participate in mutual violence
and that, particularly in teenage and young adult samples, women
perpetrate violence against their partners more frequently than do the
men. Based on knowledge available, this gender symmetry is
associated primarily with Situational Couple Violence and not
Coercive Controlling Violence”(Kelly & Johnson, 2008, p. 487).
Next comes the allegedly relevant finding that in one study women
in lesbian relationships report more IPV perpetration by a lesbian
partner than by prior heterosexual partners. My initial reaction was to
be amazed that Dutton et al. somehow thought this was relevant to
questionsabout the relative frequencyof heterosexual male and female
intimate terrorism. But if it were relevant, one might want to know that
this was a sample of “abusive lesbian relationships”(Lie, Schilit, Bush, &
Montagne, 1991, p. 123). Given this dramatic sampling bias (100%
abusive relationships), how could this study possibly have found the
same level of violence among prior heterosexual partners? Did Dutton
et al. somehow not notice this? The title of this paper is “Lesbians in
currently aggressive relationships.”
As a lead-in to the discussion of Graham-Kevan and Archer (2003)
in this section of their paper, Dutton et al. misrepresent us again: “In
identifying [intimate terrorism] as an exclusively male pattern of
domination over female intimates, J. B. Kelly and Johnson…. [in a]
typology based exclusively on shelter sample data….”(p. 13). I guess
they think that if they say it often enough, it will be true. Of course, we
do not identify intimate terrorism as exclusively male, and the
typology is not based exclusively on shelter data. As I have argued
above (Section 3.3.1), Graham-Kevan and Archer (2003) essentially
supports the gender asymmetry of intimate terrorism.
The authors end this section with discussions of more findings
from two general surveys, so I'll note again that such surveys are
dominated by situational couple violence. One set of findings is from
Stets and Straus's (1989) analysis of National Family Violence Survey
data, an analysis in which they show that “repeat, severe violence
against a non-violent intimate is symmetrical by gender”(p. 14).
Dutton et al. make the mistake of thinking that this type of violence is
necessarily intimate terrorism. It is not; it is simply repeat, severe,
non-reciprocal violence; there is no evidence regarding control.
The other evidence is a reference to a piece in which Graham-
Kevan and Archer make the mistake discussed above (Section 3.3.1),
using a cluster analysis in a sample that includes little or no intimate
terrorism. When this piece was presented at meetings, I contacted
Graham-Kevan as follows: “Ifinally found time to read your ‘non-
selected sample’paper carefully. As usual, I love the care with which
you do your work. I skimmed it when you first sent it to me and, as
294 M.P. Johnson / Aggression and Violent Behavior 16 (2011) 289–296
5
To keep my reference list manageable, in general I will not provide citations for
these papers. I will proceed through them in order and the citations can be found in
the Dutton et al. paper.
Author's personal copy
you might expect, I was troubled that you found gender symmetry for
IT and VR, and I think I know why that happened. My interpretation of
your results has general implications that I think are very important
both for DV research and for the uses of cluster analysis in general.
Here's what I came to think as I read your findings. Does this study
really get at IT or merely at the most controlling cases of SCV? I have
always argued that general samples will include very little IT. I expect
that you had a very low response rate (you don't report how many
emails were sent), which would exaggerate that bias. If you have a
sample that in fact is mostly SCV (we had the same problem with
NVAW data), then your cluster analysis will identify relatively high
control, but will still include lots of (maybe even mostly) SCV. We
really need to get away from the cluster analysis approach because it
is almost entirely dependent on the nature of the sample (as you note
in your paper, when you point out that Frieze's sample has that group
of male ITs from shelters and courts). The patterns that you find with
regard to the difference between IT and SCV for other variables also
didn't look as strong as the patterns we've found in other work. So, I
thought, ‘I'll bet their high control group isn't as high as it is in their
studies with other, selected samples that would include real IT.’And I
went to your 2003 JIV paper. Look at the numbers for the control types
below. The first column is your 2-cluster highs, the second your 3-
cluster highs, and the third your 2003 highs. [Here I gave her the
numbers, which showed way higher control for the 2003 intimate
terrorist group.] I wonder what would happen if you re-ran your
unselected data using a cutoff derived from the 2003 paper. Take your
2003 cluster analyses, crosstab it against the total control score, and
choose a cutoff that comes closest to replicating the high cluster (I've
done this in some of my papers, but I've never had the same measures
across the various samples I've used so that I could use the same
criterion to identify high control across samples). Now use that cutoff
to identify IT, VR, SCV, and MVC in your unselected sample. The
numbers in each type will change dramatically, and I would predict
that you'll see a more dramatic differentiation between IT and SCV,
including a shift in the gender symmetry. I'm really dying to find what
happens if you do this. Willing to try it?”Graham-Kevan said she'd do
it and get back to me, but that never happened. I'd still like to see the
re-analysis. When my colleagues and I used a similar strategy to re-
analyze the NVAW data (Johnson et al., 2008), we found only 34 male
intimate terrorists among current partners of almost 5000 married
respondents (Johnson & Leone, 2005).
4. Conclusion
So, what's up with these authors? Why the comic book caricatures
of the feminist analysis? Why the gross misrepresentations of what
Joan Kelly and I wrote in our 2008 article? Why the single-minded
focus on alleged evidence that women are as bad as men? In their
determination to see what they want to see, they seem to have missed
the obvious implications of the call for differentiation among types of
intimate partner violence that is the heart of the Kelly & Johnson and
Jaffe et al. articles. So, let me finish with a summary of the implications
of the Kelly and Johnson paper and of this review.
Most intimate partner violence, both men's and women's, does not
fit the power and control model of intimate terrorism. It is situational
couple violence and must be treated accordingly. And even if it were
to turn out to be the case that men and women were equally likely to
be intimate terrorists (I of course do not believe that and the evidence
does not indicate that), it would not affect what Joan Kelly and I
suggested in our paper. The dramatic differences among intimate
terrorism, violent resistance, and situational couple violence make it
essential that the family courts make these distinctions in order to do
the right thing with respect to the adults involved, and to serve the
best interests of the children. In fact, the final paragraph of the Kelly
and Johnson paper serves well as the correct interpretation of the
literature cited by Dutton et al. in their attack on the feminist analysis
of intimate partner violence: “Current research provides considerable
support for differentiating among types of intimate partner violence,
and such differentiations should provide benefits to those required to
make recommendations and decisions about custody and parenting
plans, treatment programs, and legal sanctions”(Kelly & Johnson,
2008, p. 495).
References
Ansara, D. L., & Hindin, M. J. (2010). Exploring gender differences in the patterns of
intimate partner violence in Canada: A latent class approach. Journal of
Epidemiology and Community Health,64, 849–854.
Buzawa, E. S. (2003). Domestic Violence: The Criminal Justice Response. Thousand Oaks,
CA: Sage.
Cook, P. W. (1997). Abused Men: The Hidden Side of Domestic Violence. Westport, CT:
Praeger Publishers/Greenwood Publishing Group, Inc.
DeKeseredy, W. S., & Dragiewicz, M. (2007). Understanding the complexities of
feminist perspectives on woman abuse: A commentary on Donald G. Dutton's
Rethinking Domestic Violence. Violence Against Women,13(8), 874–884.
Dobash, R. E., & Dobash, R. P. (1992). Women, Violence and Social Change. New York:
Routledge.
Dragiewicz, M. (2008). Patriarchy reasserted: Fathers' rights and anti-VAWA activism.
Feminist Criminology,3(2), 121–144.
Dragiewicz, M., & Lindgren, Y. (2009). The gendered nature of domestic violence:
Statistical data for lawyers considering equal protection analysi s. American
University Journal of Gender, Social Policy & the Law,17(2), 229–268.
Dutton, D. G., Hamel, J., & Aaronson, J. (2010). The gender paradigm in family court
processes: Re-balancing the scales of justice from biased social science. Journal of
Child Custody,7(1), 1–31.
Felson, R. B., & Outlaw, M. C. (2007). The control motive and marital violence. Violence
and Victims,22(4), 387–407.
Frieze, I. H. (2000). Violence in close relationships-development of a research area:
Comment on Archer (2000). Psychological Bulletin,126(5), 681–684.
Frye, V., Manganello, J., Campbell, J. C., Walton-Moss, B., & Wilt, S. (2006). The
distribution of and factors associated with intimate terrorism and situational
couple violence among a population-based sample of urban women in the United
States. Journal of Interpersonal Violence,21(10), 1286–1313.
Graham-Kevan, N., & Archer, J. (2003). Intimate terrorism and common couple
violence: A test of Johnson's predictions in four British samples. Journal of
Interpersonal Violence,18(11), 1247–1270.
Hamel, J., Graham-Kevan, N., & Prospero, M. (2008). Controlling and abusive tactics:
Preliminary findings from a gender-inclusive questionnaire. Paper presented at the
12th International Family Violence Research Conf erence, Portsmouth, New
Hampshire.
Hines, D. A., & Douglas, E. M. (2010). Intimate terrorism by women towards men: Does
it exist? Journal of Aggression, Conflict, and Peace Research,2(3), 36–56.
Holtzworth-Munroe, A. (2005). Male versus female intimate partner violence: Putting
controversial findings into context. Journal of Marriage and Family,67(5),
1120–1125.
Holtzworth-Munroe, A., Meehan, J. C., Herron, K., Rehman, U., & Stuart, G. L. (2000).
Testing the Holtzworth-Munroe and Stuart (1994) batterer typology. Journal of
Consulting and Clinical Psychology,68(6), 1000–1019.
Jaffe, P. G., Johnston, J. R., Crooks, C. V., & Bala, N. (2008). Custody disputes involving
allegations of domestic violence: Toward a differentiated approach to parenting
plans. Family Court Review,46(3), 500–522.
Johnson, M. P. (1995). Patriarchal terrorism and common couple violence: Two forms of
violence against women. Journal of Marriage and the Family,57(2), 283–294.
Johnson, M. P. (2001). Conflict and control: Symmetry and asymmetry in domestic
violence. In A. Booth, A. C. Crouter, & M. Clements (Eds.), Couples in Conflict
(pp. 95–104). Mahwah, NJ: Lawrence Erlbaum.
Johnson, M. P. (2005a). Apples and oranges in child custody disputes: Intimate
terrorism vs. situational couple violence. Journal of Child Custody,2(4), 43–52.
Johnson, M. P. (2005b). A brief reply to Dutton. Journal of Child Custody,2(4), 65–67.
Johnson, M. P. (2005c). Domestic violence: It's not about gender—Or is it? Journal of
Marriage and Family,67(5), 1126–1130.
Johnson, M. P. (2006a). Conflict and control: Gender symmetry and asymmetry in
domestic violence. Violence Against Women,12(11), 1–16.
Johnson, M. P. (2006b). Gendered communication and intimate partner violence. In B. J.
Dow, & J. T. Wood (Eds.), The Sage Handbook of Gender and Communication
(pp. 71–87). Thousand Oaks, CA: Sage.
Johnson, M. P. (2007). Domestic violence: The intersection of gender and control. In L. L.
O'Toole, J. R. Schiffman, & M. K. Edwards (Eds.), Gender Violence: Interdisciplinary
Perspectives (pp. 257–268). (2nd ed.). New York: New York University Press.
Johnson, M. P. (2008). A Typology of Domestic Violence: Intimate Terrorism, Violent
Resistance, and Situational Couple Violence. Boston: Northeastern University Press.
Johnson, M. P., & Cares, A. (2004, November). Effects and non-effects of childhood
experiences of family violence on adult partner violence. Paper presented at the
National Council on Family Relations annual meeting, Orlando, FL.
Johnson, M. P., & Ferraro, K. J. (2000). Research on domestic violence in the 1990s:
Making distinctions. Journal of Marriage and the Family,62(4), 948–963.
Johnson, M. P., & Leone, J. M. (2005). The differential effects of intimate terrorism and
situational couple violence: Findings from the National Violence Against Women
Survey. Journal of Family Issues,26(3), 322–349.
295M.P. Johnson / Aggression and Violent Behavior 16 (2011) 289–296
Author's personal copy
Johnson, M. P., Leone, J. M., & Xu, Y. (2008, November). Gender, intimate terrorism, and
situational couple violence in general survey data: The gender debate ited–again.
Paper presented at the National Council on Family Relations annual meeting, Little
Rock, AR.
Kelly, J. B., & Johnson, M. P. (2008). Differentiation among types of intimate partner
violence: Research update and implications for interventions. Family Court Review,
46(3), 476–499.
Kimmel, M. S. (2002). “Gender symmetry”in domestic violence: A substantive and
methodological research review. Violence Against Women,8(11), 1332–1363.
Laroche, D. (2005). Aspects of the context and consequences of domestic violence:
Situational couple violence and intimate terrorism in Canada in 1999. Retrieved July
22, 2009, from http://www.stat.gouv.qc.ca/bul/conditions_vie/AspectViolen_an.pdf
Leone, J. M., Johnson, M. P., & Cohan, C. M. (2007). Victim help-seeking: Differences
between intimate terrorism and situational couple violence. Family Relations,
56(5), 427–439.
Leone, J. M., Johnson, M. P., Cohan, C. M., & Lloyd, S. (2004). Consequences of male
partner violence for low-income, ethnic women. Journal of Marriage and Family,
66(2), 471–489.
Lie, G. -y., Schilit, R., Bush, J., & Montagne, M. (1991). Lesbians in currently aggressive
relationships: How frequently do they report aggressive past relationships?
Violence and Victims,6(2), 121–135.
Pence, E., & Paymar, M. (1993). Education Groups for Men Who Batter: The Duluth Model.
New York: Springer.
Renzetti, C. M. (1992). Violent Betrayal: Partner Abuse in Lesbian Relationships. Thousand
Oaks, CA: Sage.
Stacey, W. A., Hazlewood, L. R., & Shupe, A. (1994). The Violent Couple. Westport, CT: Praeger.
Stets, J. E., & Straus, M. A. (1989). The marriage license as a hitting license: A comparison
of assaults in dating, cohabiting, and married couples. Journal of Family Violence,
4(2), 161–180.
Sugarman, D. B., & Frankel, S. L. (1996). Patriarchal ideology and wife-assault: A meta-
analytic review. Journal of Family Violence,11(1), 13–40.
296 M.P. Johnson / Aggression and Violent Behavior 16 (2011) 289–296
