The latest figures on intimate partner femicide show much of a recent rise in men killing women has now been reversed, at least temporarily.

---

Prologue: Violence against women is a bad thing, and it’s still bad even when, as the article below points out, it used to be far worse. We should be trying hard to lower rates of violence, by finding good solutions and implementing them with urgency. As part of this, we should understand just what we’re dealing with – which is what this series of posts tries to do.

This is my second post on Australian male violence against women. Back in early 2024, I posted here on Troppo about what some community groups argued was an epidemic of male violence. 2022-23 had seen a rise in the number of women killed in intimate partner violence, and 2023-24 was already shaping up to be worse. As community groups voiced their concern, news media focused on the issue for several weeks. South Australia called a Royal Commission.

Some 18 months on, it's worth checking what has happened to those statistics.

The 2024-25 drop is not historic

The news seems good. After two years of rises, 2024-25 brought a huge 35% drop in the rate of intimate partner homicide against women. This fall is shown at the right-hand bottom corner of the graph atop this page. See an interactive version here.

In percentage terms, the 2024-25 drop is the biggest recorded single-year fall in intimate partner homicide against women in Australian history (the records go back to 1989-90). We're now back down near the all-time lows of 2020-2022 – a period when COVID lockdowns may have been making the figures look misleadingly good (though we'll probably never know for sure).

Did Australia celebrate that "historic fall"? Did community groups put out press releases congratulating Australia on slashing this most reliable and most awful of all domestic violence metrics? Did the media announce this biggest fall in the history of our figures?

No, we saw no celebration, no press releases – indeed, not even a media story that I could find.

And here's the thing: I would argue that this non-response was entirely appropriate – for two reasons.

• One reason is that we still have a long way to go in reducing violence against both women and men. Other advanced nations – notably Singapore – have pushed the rate much lower. We can do that too.
• But the second reason is that the 2024-25 fall was no more important than the rises that preceded it. Both seem likely to have been the sort of statistical variations you would expect for a number like this.

Never mind that the 2022-2024 rises grabbed the attention of community groups and the media. Never mind even that the 2020-2022 figures may have been artificially depressed by the COVID lockdowns. Both the rise and the fall are likely to be blips in a bigger long-term change – that long drop in female partner homicides.

For 2024-25, the figures we're looking at show around 30 intimate partner homicides of women. The obvious and correct response is that this is still too many deaths. But in statistical terms, we can make another, more descriptive social science observation: when your numbers are in this territory, the statistics will bounce around a lot. They will bounce around even more than they did when the homicide rate was three times higher, at the start of the 1990s.

The lesson here is that people should have been less eager to draw conclusions about the 2022-23 and 2023-24 rises. Due to the small statistical sizes, we need to take care in drawing conclusions about any short-term changes in homicide, and we need to take even more care when discussing a smaller sub-group of homicides.

This was a key point of last year's post. And it will be true next year too, whether the 2025-26 figures show a rise or a fall.

The hellscape impulse is rising

One lesson from this episode is that in the 2020s, we are eager to infuse our talk about gender violence with heavy moral significance. And that's appropriate. We're talking about people intentionally killing their partners, not only destroying those people's futures but blighting the futures of people around them.

Indeed, while I came to this issue as a philosophical moral realist, I've noticed that even people who would normally not describe themselves that way tend to suddenly become moral realists on this issue. Thankfully, almost everyone seems to think that intimate partner homicide against women is objectively a bad thing, regardless of whether they believe in moral objectivity the rest of the time.

What we struggle to do is stop that legitimate moral impulse from infecting our subsequent analysis of what's happening to Australian society.

The result is that we end up depicting Australia as a sort of growing hellscape of male-on-female homicide.

In fact, as the graph above shows, the opposite is the case. Intimate partner homicide against women has dropped by two-thirds in the past 35 years. That seems an achievement. Other countries have done well too, but we seem to have done better than most.

But that picture is hard to make out amid the blizzard of condemnation. Reality gets overshadowed. Our morality pollutes our epistemology. More and more people seem to want to declare an ineradicable hellscape.

I've mentioned that concern over an epidemic of violence manifested most quickly in the statements of community groups and the media. But the concern reached much wider, into academia and government.

Take the example of Rick Sarre, emeritus professor of law and criminal justice at the University of South Australia, and possessed of a masters degree in criminology. Writing in The Conversation on the same day as my Club Troppo post, Sarre referred to "the seemingly unrelenting number of murders perpetrated by men against their intimate partners". That might be considered a descriptive statement, a characterisation of the state of public worry. But Sarre also described the figures in normative terms, as "cause for mounting concern for all Australians". And he concluded that Australia needed a "transformative" reallocation of resources to the problem.

I don't mean to pick on Sarre. He was far from the only academic to paint such a picture. Indeed, a few days after my initial Club Troppo post, he was quoted in the Australian Financial Review making very sensible observations: he noted falling long-term homicide rates and observed that "long-term trends are often ignored in the rush to analyse short-term crime rate fluctuations". My worry is that Sarre's Conversation piece suggests even sensible people may struggle to resist the hellscape impulse.

The pressures can be intense. A prominent anti-violence activist underlined this for me by ringing me one evening soon after I made my April 2024 post. She wasn't subtle: I should take the whole post down, she said, because I simply didn't know what I was talking about. Sadly, she rang off before I could ask her how often this sort of tactic worked on other people.

One result of this sort of debate: most people end up in, at best, a sort of epistemic fog about certain issues. This site's own Nick Gruen reacted to my original 2024 post by disclosing in the comments that he had been wanting that sort of statistical analysis, "as I’ve asked myself the obvious questions about whether this thing is worse or better than before". Nick has degrees of various sorts in history, public policy, economics and law. At the time, if I had had to bet on anyone outside the criminal justice and social statistics fields understanding the long-term changes in Australian homicide figures, I would have bet on Nick. But even he didn't know what to think.

If Nick Gruen didn't know what was going on with intimate partner homicide in 2024, I don't have any faith that very many ordinary intelligent Australians understood it then, or understand it now.

The repercussions of misunderstanding

When the concern about an "epidemic" of violence suddenly broke out in 2024, the federal government had only recently published its National Plan to End Violence against Women and Children 2022–2032. Nevertheless, it found itself required by political logic to act, no matter what its more careful minds may have thought of the underlying realities. The government pledged more than $925 million in new funding over five years to "address men's violence towards women". This included up to $5000 per person to support those escaping violent relationships.

You might well think that this increase in government support seems like a good idea, so there's no point in complaining about how we got it. Heck, I think that $5000 payment seems a good idea, even though I also think that I don't know enough about the issue to be sure.

At the same time, though, I worry about the effects this sort of thing is having on both government and society.

• As a culture, it is possible that our hellscape impulses are suggesting to some people that violence is somehow inevitable – that our efforts do not make a difference. In fact, while we don't understand the reasons for the fall in intimate partner homicide, it certainly seems possible that our efforts are making a difference.


• We are effectively encouraging lobby groups to make public arguments that are not merely emotive but detached from facts. The events of the past few years don't really show that intimate partner homicide against women is any more of a problem than it was in 2020. But it may be easier and less painful to go with the flow.


• We are not in any way penalising media outlets for writing stories that suggest things which are untrue. We no longer even seem to encourage them to supply both sides of some stories. The only media outlet that I saw attempting rational analysis of the 2024 violence "epidemic" was the Australian Financial Review. (If anyone has other examples, please let me know and I'll mention them here.)


• More broadly, we are encouraging people in general to believe that on very many fronts our lives are growing inexorably worse, even as – at least on many of these fronts – our lives are actually growing better. You famously see this on Twitter, but you might reasonably worry that it is creeping into the wider culture. Again, fight the hellscape impulse.

Australia is much better than it has been, and Australia can be much better than it is now. To encourage future betterment, we should remind Australians of how much betterment has happened already.

---

Comments: As usual, yes, I'm an idiot about a lot of things. I really will be grateful if you can point out in the comments specifically where my idiocy lies, and detail the huge mistake(s) I'm making. (But maybe don't just ring me up, insist I remove this blog post, and then hang up.)

About the author: I studied criminology at the University of Adelaide and have dealt with statistics and their presentation in various roles for more than 30 years.

Twitter: @shorewalker1