Misinterpretation, damn misinterpretation and statistics.
Over the last few years, male violence against women has become a matter of public discussion in ways we have never seen before. Despite the horrors of it, it’s a hugely positive step because it means we are finally taking steps towards addressing it.
The difficulty with all this public discussion though, is that too much data is being misrepresented or misunderstood, and misinformation is almost as bad as no information.
When we talk about violence there are usually a few different contexts. Definitions matter when you’re talking about numbers, but get complicated when you’re talking about people and relationships, because there is almost always some crossover.
Domestic violence: any form of violence that occurs between people who have a familial or intimate relationship, or in the context of those relationships.
Intimate Partner Violence (IPV): definitions vary, in some contexts it only applied to cohabiting intimate partners, other definitions include any people engaged in a sexual relationship, current or ex. It can be physical, sexual, financial, emotional or all of the above.
Sexual violence: again, definitions vary, but usually for statistical purposes it means any form of sexual threat or assault involving any form of sexual violence. Depending on the source, it may or may not include stalking and harassment.
Physical violence: any form of physical threat or violence to anyone, whether they are known to the offender or not.
Counting Dead Women
Destroy the Joint have done outstanding work in bringing to glaring public notice the number of women killed by violence in Australia. But their data is too often misreported. At the time of writing, they have recorded 78 women killed so far this year, or nearly 2 women every week. Their Counting Dead Women project records all women killed by violence, as reported by the media. So, firstly, it is possible that more women have been killed and the murder wasn’t reported by the media; and secondly, as they state on their page, they are recording all women killed by violence, not just women killed by intimate partners. They estimate that about 75% of those women were killed by a man known to them, but this is based on media reports, not police or coroner’s investigations, so it’s an estimate, not a fact.
DTJ are very clear that they are only recording dead women only, not children (because reporting on child victims of crime is risky and legally complicated) or men. There are good reasons for this, but it can distort public perception that women are the primary victims of violence. They’re not. They are the overwhelming majority of victims of family violence, but most victims of male violence are actually other men.
There are several types of information on violence in Australia.
Crime statistics: supplied by state police forces, covering crimes recorded by police. This data excludes acts reported to police where police have determined that no offence was detected, and does not account for whether a court determined guilt or innocence. It also obviously does not include all the crimes not reported to police.
Homicide data: this is the most robust and reliable form of data we have (I've collated some of the more detailed ones here). There is no underreporting, and all incidents of homicide are thoroughly investigated by Coroner’s Courts as well as police, so information about the crimes is verified. However, homicide is relatively rare in Australia, so while this information can give an indications of lower levels of violence, it cannot be reliably extrapolated.
Survey data: where a person is asked a series of questions about the violence they have experienced, the one most commonly used in Australia is the Personal Safety Survey (PSS) conducted by the Australian Bureau of Statistics.
The Personal Safety Survey – an indication, not fact
The PSS is a useful indication of how Australians perceive the violence they have experienced, but it is not a list of empirical facts. The PSS statistics, like any statistics, need to be interpreted and understood in context.
The most important thing about the PSS is that it is self-reported data; that is, it’s not verified by any other source, it is a record of the respondents’ perception and memory of the violence they have experienced, and how they are willing to describe that in answer to a survey by a stranger. The response rate for the PSS is 57% and obviously there is no data on why the remaining 43% declined to participate.
There are a number of factors in this that can skew the data. Memory and perception are coloured by emotion, and even the best efforts of the highly trained staff at the ABS cannot always correct for that.
For example, most abusive men do not think of themselves as the abusers, in their perception, they are the victims. Conversely, abused women will often not think of themselves as the victim, they take responsibility for violence enacted against them, feel shame for it, and tend to minimise its extent and effect.
To explain how this might play out in skewing PSS data, take a scenario where Jennifer is in an abusive relationship with John. In one incident, John thinks Jennifer has said something demeaning to him, he responds by standing close to her and yelling, she pushes him away and tries to leave the room, he knocks her to the ground.
If John was asked to complete PSS survey, he might feel shame and refuse to participate, or he might gladly participate because he believes he had nothing to feel ashamed of. He might the report the incident as his experience of violence (she pushed him) and emotional abuse (she demeaned him). The PSS does not ask if he has committed violence against Jennifer.
If that scenario played out regularly in their house, Jennifer is quite likely to avoid answering the PSS survey because she is afraid that John would find out, or she is ashamed of the violence enacted against her and doesn’t want to tell anyone about it. If she did agree to participate, she might not report it as a violent incident because she believes she initiated the violence. She also might not admit or even remember he hit her, because she does not want to acknowledge to herself or anyone else that she is in a violent relationship.
Obviously this is not the case with all people who respond to the PSS and the questions are specific enough that some of this information is teased out, but it is a factor in understanding the data that comes out of the PSS, particularly when it is focussing on sexual violence or family violence, both of which are subjects frequently associated with shame and secrecy.
Not all abuse is violent and not all violence is abuse
Domestic or family violence is not a binary thing – where someone either is violent or they are not.
Mutually violent relationships do include physical violence, but rarely any other kind. Both partner will equally hit, kick, push or throw things at each other. It might be dangerous, make both partners unhappy, and can be traumatic for children of the relationship, but it’s not abusive, in that both parties to it are equally damaging and neither one is in control of the other.
Intimate terrorism, as it is called by Michael P. Johnson in his book A Typology of Domestic Violence, is very different. There is no mutual violence, one partner is violent for the sole purpose of creating fear and exerting control over the other. Intimate terrorism may not always be physically violent, but the fear of physical and sexual violence is almost always present. Violence in this context is a means to an end (control), it is not the result of a lack of control.
Emotional abuse is almost always an aspect of intimate terrorism, but is rarely part of uncontrolled violence. And it is most definitely not the same as hurting someone’s feelings. People can be deeply hurt by their partners, through thoughtlessness, anger, even infidelity or the end of a relationship, without suffering emotional abuse. It is a very specific and deliberate form of emotional damage, designed to destroy any feeling of independence or self-worth, and thereby make someone easy to control and manipulate.
Financial abuse is another form of control frequently misunderstood. It’s not the same as being irresponsible with money, or failing to take an equal share in household finances. It is again, deliberate, targeted behaviour designed to control the victim, make them feel powerless and ensure they do not have access to go anywhere or do anything without their abuser’s permission.
Sexual violence is probably the only form of violence that is always abusive, because it’s not something that can occur through thoughtlessness, by accident or by mutual agreement. It is only ever violation, and all the more so when it is at the hands of someone you love.
Statistics and facts
The complex interactions between all those forms of violence make it almost impossible to identify them statistically. The nature of abuse and control makes it very unlikely a victim could report objectively on what is happening to them, and quite likely that an abuser will misreport what is happening in a relationship in which they are abusive. The PSS does a very good job of asking detailed questions that could establish an indication of the prevalence of abuse, but at best it can only ever be an indication and is almost certainly underreporting the reality.
To be clear, this is not a criticism of the PSS or the staff who administer it, it’s just an explanation of what data they are reporting and the context in which we need to understand it.
Anecdotes are data
Because of all this, abuse may be the one thing we can only understand properly through storytelling rather than data. It’s too complex, too intangible and too emotionally fraught to make for reliable statistics. This is why Rosie Batty’s story was so compelling, and why the Hitting Home documentary was so important. The stories of women in fear of their lives because their abuser’s need for control has become life-threatening are the only way of explaining the inexplicable to someone who has never experienced violence in this way. Homicide data is the only time we can conduct thorough investigations and obtain reliable data, but but that time, it’s far too late.