The 'gendered violence' narrative: rhetoric, errors, cherry-picked statistics and circular references
Tuesday, July 18, 2017
One in Three Campaign

Abstract

This paper examines bias in an online article published by the Australian ABC’s triple j HACK, as a case study of the way the ‘gendered violence’ narrative is largely built upon unsupported opinions, errors, cherry-picked statistics, circular references and ignoring of robust data.

The triple j HACK article in question reported there is only anecdotal evidence to back up the claim that men are either too ashamed, too stoic, or too chivalrous to report being hit by their female partner. This was subsequently amended to instead report there is mixed evidence to back up this claim, with some studies showing men are more likely to report violence, and others showing they're less likely.

The data cited in support of the amended report were found in an unpublished, unreferenced, non-peer-reviewed seminar presentation by Dr Michael Flood. Dr Flood’s presentation cited two overseas studies up to 44 years old with small or biased samples; an Australian study co-authored by himself that cited no data to support his opinion; an Irish study that actually showed the opposite of his claims; and an Australian crime victimisation survey where the majority of perpetrators were male, making its data irrelevant.

triple j HACK ignored or was ignorant of four large-scale representative community surveys from across the Western world, providing conclusive evidence that men are less likely than women to report domestic violence against them to the police.

Critique

On the 19th June 2017, the ABC’s triple j HACK posted an anonymous article on its website titled, “What about men?: Challenging the MRA claim of a domestic violence conspiracy.” The error-ridden piece lacked balance by failing to interview anyone who supports the claim that one in three victims of family violence is male. Instead it focused entirely upon the views of its detractors.

Perhaps in response to complaints about these errors, triple j HACK amended the article on 23rd June. This amendment ended up being a wonderful case study of how the ‘gendered violence’ narrative is largely built upon unsupported rhetoric, errors, cherry-picked statistics, circular references and ignoring of robust data.

The original version of the triple j HACK article reads as follows: 

Some MRAs argue the one in three figure actually underestimates the number of male victims of domestic violence, because men are either too ashamed, too stoic, or too chivalrous to report being hit by their female partner.

Hack has heard from men who said they were too afraid to report violence by their female partner, or thought the police wouldn't take them seriously.

But apart from these anecdotal reports, there's no other evidence to back up this claim, and no easy way of measuring this potential statistical bias [my emphasis].

The amended version reads:

... But apart from these anecdotal reports, there's only mixed evidence to back up this claim, with some studies showing men are more likely to report violence, and others showing they're less likely [my emphasis].

The “mixed evidence” provided by triple j HACK links to an unpublished, non-peer-reviewed seminar presentation by Dr Michael Flood, ARC Future Fellow, Faculty of Law at QUT (who appeared as an “expert” on the recent triple j HACK TV program HACK Live: Is Male Privilege Bullsh!t?). “He Hits, She Hits: Assessing debates regarding men’s and women’s experiences of domestic violence” was presented at a seminar held at the Australian Domestic and Family Violence Clearinghouse, UNSW, Sydney, in December 2012.

Let’s take a look at the evidence provided by Dr Flood in his seminar presentation. Page 14 contains a section titled “Gendered patterns in reporting of victimisation”. It reads as follows:

Let’s focus now on the reporting of subjection to violence. It has been argued that men are likely to under-estimate and under-report their subjection to domestic violence by women, because admitting such victimisation and vulnerability is emasculating (George 1994: 149). 

There is mixed evidence regarding whether male victims of domestic violence are more or less likely than female victims to report their experience. 

Men’s Health Australia cite some studies suggesting that men are less likely than women to report domestic violence against them to police. However there is other evidence that men are more likely to report than women, or no more likely to do so.

A fundamental limitation here is that there has been little research which examines women’s and men’s reporting of domestic violence. We know little about whether the factors shaping under-reporting are similar or different for women and men.

In the studies cited by Men’s Health Australia, lower proportions of men than women report to police the physical aggression they have experienced. It is possible that men in these studies were less likely than women to report these incidents because they did not see them as serious or threatening. 

If men are less likely than women to go to the police when they have experienced physical aggression by a partner, it may be because this violence is less severe, less threatening, and even minor or trivial.

Let us analyse these claims in detail. Firstly, what are the studies that Men’s Health Australia cite, suggesting that men are less likely than women to report domestic violence against them to police? Although Dr Flood is a prominent Australian academic in the field of domestic violence, he appears to have been unable to provide a citation, but it is likely he is referring to the online article titled “What about the men? White Ribbon, men and violence: A response to Dr Michael Flood by Men’s Health Australia” which found (page 13):

The large-scale South Australian Interpersonal Violence and Abuse Survey found that “females (22.0%) were more likely to report the [domestic violence] incident(s) to the police than males (7.5%)” (Dal Grande et al 2001: 10). Likewise “The 2004 [Canadian] General Social Survey (GSS) on victimisation found that fewer than 3 in 10 (28%) victims of spousal violence reported the abuse to the police (36% of female victims and 17% of male victims)” (Statistics Canada 2009: 24). Also, the 2008-09 Scottish Crime and Justice Survey: Partner Abuse found that... “Around one in five (21%) who experienced partner abuse in the last 12 months said the police did come to know about the most recent / only incident. Again, this figure was significantly higher among women (35% compared with eight per cent of men)” (Macleod et al 2009: 29-30).

That’s three large-scale representative community surveys conducted across the English-speaking world all finding that men are significantly less likely than women to report domestic violence against them to police. Let’s take a look now at the other evidence presented by Dr Flood that men are more likely to report to police than women, or no more likely to do so.

Firstly he cites Allen (2011: 249), who “describes various studies finding that men are more likely to report than women. For example, an Irish study found that three quarters of the men, compared to about half of the women, had told someone about the abuse within a year.” Dr Flood appears to have been unable to provide references to his citations at the end of his seminar presentation, however It is likely he was referring to Mary Allen’s paper “Is there gender symmetry in intimate partner violence?” from the journal of Child and Family Social Work.

Page 249 of Allen’s paper contains the following section under the title, “Why do men disappear in official statistics?”:

One of the commonest explanations for the clear disparity between the avowed gender symmetry of domestic violence and the invisibility of battered men in police and hospital statistics is by reference to the social stigma of admitting to being abused by one’s female partner. As Steinmetz (1977/78: 503) notes ‘the stigma... which is embarrassing for beaten wives, is doubly so for beaten husbands’. Dobash et al. (1992, p. 76) counter this suggestion, citing Schwartz’s (1987) analysis of the 1973/82 US National Crime Survey Data, which found that 67.2% of men and 56.8% of women called the police after an assault by their partners. They also cite Kincaid’s (1982) study of family court cases in Ontario which found that while there were 17 times as many female as male victims of domestic violence, only 22% of the women pressed charges in contrast to 40% of the men, and men were less likely to drop the charges (p. 91).

Taft et al. (2001, p. 500), using Australian data, state categorically that there is no ‘empirical evidence that men are more likely than women to under-report to police, hospitals or to seek help’. Watson & Parsons’ (2005, p. 77) Irish prevalence study found that men were more likely to tell someone about the abuse they were experiencing: ‘about half of the women, compared to three quarters of the men had told someone within a year.’ Data such as this clearly challenges the common perception that men are too ashamed to report violence by their partners.

Neither the Dobash et al. or Kincaid books appear to be available online so it’s difficult to check the veracity of Allen’s claims. However, the Dobash et al. data is some 44 years old and relies on a small US sample set of some 1743 (1641 female, 102 male) respondents. It is also quite possible that the 10.4% difference in police reporting between men and women is not statistically significant. The Kincaid data only tangentially supports Flood’s claims. It is at least 35 years old and relies on a small biased Canadian sample of family court cases (not the general community) where litigants would have had added incentives (i.e. property and children) to press charges against their ex-partners. Nevertheless, without being able to properly check the cited sources, we should take this at face value as some data supporting Dr Flood’s claims, if limited.

The next paragraph is where things get really interesting. Allen claims that Taft et al., using Australian data, state categorically that there is no ‘empirical evidence that men are more likely than women to under-report to police, hospitals or to seek help’. Guess who the ‘et al.’ refers to as the co-authors of this citation? No other than Dr Michael Flood himself along with Angela Taft and Kelsey Hegarty. Their article “Are men and women equally violence to intimate partners” (Australian and New Zealand Journal of Public Health) does indeed make the claim cited by Allen (page 5). However, unlike the dozens of other claims made in the article, this particular claim contains no citation to provide evidence of its veracity. It appears therefore to be a statement of opinion, rather than a statement of fact. There is certainly no “use of Australian data” to support it, as claimed by Allen.

Let’s turn to Watson & Parsons’ Irish prevalence study. Allen claims it found that men were more likely to tell someone about the abuse they were experiencing, citing the quote that “about half of the women, compared to three quarters of the men had told someone within a year.” Watson & Parsons’ “Domestic Abuse of Women and Men in Ireland: Report on the National Study of Domestic Abuse” is a large-scale community survey. It actually found that “women are more likely than men to report [severe abuse] to the Gardaí [Irish Police]. Over a quarter of women reported their experience to the Gardaí compared to about one man in 20” (page 26). The actual figures (page 76) are that 5% of men and 29% of women who had experienced severe abuse had reported it to the police (a statistically significant difference). The report does contain the statistic that only about half of the women, compared to about three quarters of the men, had told someone within a year (page 77), but this refers to the speed of reporting, not the numbers of men and women who ever reported abuse. Allen appears to have not only cherry-picked this “statistic”, ignoring the real data, but also misrepresented it by claiming it “found that men were more likely to tell someone about the abuse they were experiencing”. [When referencing victims revealing their experience of abuse to anyone - not specifically police - the report found that “Women and men were about equally likely to have told someone about the abuse, with two thirds revealing their experiences to someone”].

It appears Dr Flood either didn’t bother to check Allen’s source data, or he wilfully ignored her errors. In any case, her errors were repeated by him when he claimed that the Irish study found “that men are more likely to report than women”.*

Let’s turn now to Dr Flood’s claim that, in the Australian Crime Victimisation Survey 2008-09, men and women were equally likely to report a domestic assault to police (Grech 2011: 9). Once again, as Dr Flood neglected to provide references for his seminar presentation, we are left to conclude that he is referring to Katrina Grech and Melissa Burgess’ “Trends and patterns in domestic violence assaults: 2001 to 2010”, published by the NSW Bureau of Crime Statistics and Research.

This report did indeed find “There is no difference between men and women in their willingness to report a domestic assault to police” (page 9). However, “New South Wales police define domestic assault in line with the Crimes (Domestic and Personal Violence) Act 2007. Under this legislation a person has a domestic relationship with another if they are a spouse/partner, ex spouse/ex partner, boy/girlfriend (including ex), parent/guardian (including step/foster), child (including step/foster), sibling or other member of family (including kin). Domestic violence can be considered as an associated factor for any criminal incident where the victim-offender relationship meets any of these criteria.”

Therefore, the domestic assaults being reported to NSW Police by male victims include assaults by partners (81.8% of these were female) and also non-partners (only 17% of these were female). In total, more than half of the perpetrators of domestic assault against men reported to NSW Police in this report were male. The issue at hand is men’s reluctance to report victimisation to the police because of the shame and stigma of admitting that a woman has been assaulting and/or abusing them. Data with primarily male perpetrators are largely irrelevant without a gender breakdown, which the report fails to provide.

Let’s move on to Dr Flood’s final ‘argument’, his claim that it is possible that the men in the large representative community surveys cited by Men’s Health Australia are less likely than women to report violent incidents because they do not see them as serious or threatening, or because the violence is less severe, less threatening, and even minor or trivial.

This is of course possible, however it is complete conjecture as there is little evidence to support Dr Flood’s claim from within the studies themselves. In the South Australian study, respondents who did not leave their partner as the result of the violence were asked their reasons for staying in the abusive relationship. 28 per cent of males and 20.8 per cent of females answered “violence not serious enough” – slightly higher for males but not greatly so (page 144). Respondents who had left or stayed apart because of the violence and abuse were asked their reasons for leaving. 50 per cent of males and 64.1 per cent of females answered “continuation of violence/abuse” – once again slightly higher for females but not greatly so (page 140).

In the Scottish survey, respondents were asked their reasons for not informing the police about the violence. Unfortunately there is no gender breakdown of these figures (page 31), but 26 per cent said it was a private, personal or family matter; 25 per cent said they did not report it because they dealt with the matter themselves; 23 per cent felt it was too trivial / not worth reporting; and fewer than one in ten mentioned a range of other reasons including the police not being interested (9 per cent), it being too much trouble (7 per cent), being frightened of making matters worse and because the police could have done nothing about it (both 5 per cent).

The large-scale community survey from Ireland misrepresented by Mary Allen and Dr Flood, found that male victims of severe abuse were less likely than female victims to have told the police – making a lie of any claim that the violence against men was minor or trivial.

If, like Dr Flood, we are going to hypothesise without direct evidence, it is also possible that the men in these studies were less likely than women to report these incidents because of well-documented factors (for example, Tilbrook et al 2010) such as:

Furthermore, as part of their evaluation of the 2012 family violence amendments, the Australian Institute of Family Studies found in their Experiences of Separated Parents Study that males made up 51.7% of parents who reported experiencing the highest levels of severity of fear (9 or 10 on a 10-point scale) since separation, 60.5% of parents who reported experiencing the most severe control, and 57.6% of parents who reported experiencing the most severe coercion. It appears extremely unlikely therefore that men are less likely than women to report violence because they do not see it as serious or threatening, or because it is less severe, less threatening, minor or trivial.

To summarise, triple j HACK first of all claimed “there's no other evidence to back up” the claim that “men are either too ashamed, too stoic, or too chivalrous to report being hit by their female partner”. After apparently receiving complaints from their readers, they changed their story to there being “only mixed evidence to back up this claim, with some studies showing men are more likely to report violence, and others showing they're less likely”. With all of the ABC’s resources, the only data they could find to support this claim was an unpublished, unreferenced, non-peer-reviewed seminar presentation by Dr Michael Flood.

Despite being one of Australia’s “leading experts” in the area, the only evidence Dr Flood was able to find to support his claim were two very old overseas studies with small or biased samples; an Australian study co-authored by himself that cited no data to support his opinion; an Irish study that actually showed the opposite of his claims; and an Australian crime victimisation survey where the majority of perpetrators were male, making its data irrelevant.

Against this, we have four large-scale representative community surveys from South Australia, Canada, Scotland and Ireland, providing conclusive evidence that not only are men less likely than women to report domestic violence to the police, but that there is little evidence to support the conjecture that men do not report domestic violence because they don’t see it as serious or threatening, or because the violence against them is less severe, less threatening, minor or trivial.

Greg Andresen
Senior Researcher
One in Three Campaign
18 July 2017

* UPDATE 20 July 2017: Dr Michael Flood has updated the text of his seminar available online, correcting the account of the Irish study to acknowledge that it supports the claim that male victims of DV are less likely to report their victimisation to police than female victims. You can download the amended version of his seminar here and the original version here.

Article originally appeared on One in Three Campaign (http://www.oneinthree.com.au/).
See website for complete article licensing information.