When women win: what Australia’s close elections tell us about violence against women

Electorates that narrowly elect a woman to federal parliament see a fall in recorded sexual assault and in women’s hospital admissions for injury, new research shows. Our findings suggest two explanations. First, female MPs devote significantly more attention to gender-based issues in parliamentary speech and are more likely to support gender-related legislation. Second, electing a woman shifts community attitudes towards women’s rights. Female representation in parliament shapes policy priorities, social attitudes, and, ultimately, women’s safety.

More than one in five Australian women have experienced sexual violence since the age of 15 (ABS, 2023). Australian governments have committed to ending gender-based violence within a generation. Much of that effort rests on laws, services and policing. But who we elect, and whether or not that matters, has received far less attention.

The international evidence is mixed. Studies of close elections in Brazil and Mexico found that electing women reduces violence against women. In the United States and India, female representation has been linked to an increase in recorded crimes against women, as more women come forward willing to report.

Australia is a demanding place to test whether electing women makes a difference. Party discipline in federal parliament is unusually strong — between 1950 and 2019, MPs crossed the floor in only about 2.6 per cent of parliamentary divisions. If MPs simply follow the party line, their gender should make little difference to what happens in their electorates. Our research suggests otherwise.

The difficult part of answering a question like this is that electorates that choose women might already be different from those that choose men in ways that have nothing to do with the MP’s gender. To get around this, we looked only at close federal races between a woman and a man in Australia from 2010 to 2022. When a candidate wins or loses by a very small margin, the result is close to a coin toss (a method economists call regression discontinuity). When we compared electorates where a woman narrowly won with those where a man narrowly won, they looked much the same on everything we could measure: the winner’s party and political experience, local income and education levels, and, crucially, crime rates in the election year itself. That means any differences that show up afterwards are much more likely to be caused by who won.

We combined these election results with police records of sexual offences from six states and territories, women’s hospital admissions, post-election surveys and the full record of what MPs said in parliament and how they voted.

Within electorates that narrowly elected a woman, recorded sexual offences fell over the parliamentary term. There was no effect on other crimes: assaults, drug offences, homicide, robbery, public order and property crime were all unaffected, suggesting the fall was not the result of a general change in policing.

Recorded crime, however, has a well-known weakness: it reflects not only how much violence occurs, but also how much of it is reported. A fall in recorded sexual assault could instead, in principle, mean that fewer women came forward.

To address this, we turned to a second source – hospital admissions – as they do not depend on anyone contacting the police and are recorded by treating clinicians. In federal electorates where women won in close elections, women’s hospital admissions for injuries fell.

Admissions for mental health conditions, pregnancy and infections remained unchanged, suggesting a broader improvement in women’s health or access to care is unlikely to explain the findings. As these two records are generated in entirely different ways, their shared decline is unlikely to reflect changes in reporting. Instead, it points to a genuine reduction in violence.

How can a backbencher in a tightly disciplined party make women safer? We found two explanations that operate on different timescales: what women MPs do in parliament, and how their election shifts attitudes in the community.

Parliament. Female MPs behaved differently once in office. They devoted a larger share of their speeches – between 4 to 5 percent points – to gender-related issues, including gender-based violence. In divisions on family, domestic and sexual violence and related legislation, they were 12 to 25 percentage points more likely to vote yes, and 15 to 33 percentage points less likely to vote no. They were no more or less likely to be absent, so the gap reflects the positions they took rather than their attendance.

Women in parliament also drive efforts in putting gender issues, including violence, on the political agenda. They start a far larger share of the parliamentary speeches raising these issues than their share of seats would suggest, and by 2022-23 they were starting nearly two-thirds of them

Attitudes. Within months of the election, before any new law could plausibly take effect, constituents in electorates that had narrowly elected a woman expressed greater support for women’s rights. Support for legal access to abortion, for example, was around 8 percentage points higher. Strikingly, the shift was just as pronounced among men as among women.

If women MPs acted primarily as role models, we would expect the change to be concentrated among women and girls. A shift among men instead points to a signalling effect: electing a woman signals that the local political environment is becoming more responsive towards women’s interests, and people update their views in response.

Most existing evidence on women’s representation and violence comes from developing countries or presidential systems, and much of it relies on recorded crime alone. Australia is a wealthy parliamentary democracy with strong party discipline and no legislated gender quotas, which makes it a hard test.

The fact that the pattern holds here suggests representation can matter even where individual MPs have little room to act alone.

Debates about women’s representation are usually framed around fairness. Our findings suggest something more: who sits in parliament may also shape the safety of women in Australia – not only through the laws MPs champion but through the signals their election sends.

Paper details Do, Q., Mahmood, R., Mavisakalyan, A., & Tyers, L. (2026). Women in Power: Parliamentary Action, Social Attitudes, and Gender-Based Crime (No. 1722). GLO Discussion Paper. https://www.econstor.eu/handle/10419/337670