Why I made this

One evening on the tram, a man I’d never met sat near me, told me I smelled good and kept asking me questions I didn’t want to answer. When I got off, he followed me. I called my brother, and the man only left when my brother arrived. I still think about that night. Getting off the tram didn’t make me safe. It was only when someone else showed up that it ended.

Recently, news of a woman being shot on a Melbourne tram brought that night back and made me ask a bigger question: are women actually safe in Victoria? I expected the data to be about strangers like the man on my tram. Instead, it pointed somewhere closer to home. In the year to June 2026, 2,160 victims of stalking were targeted by a former partner, almost twice as many as by someone unrelated to them (1,143). For many women, the danger doesn’t come from a stranger, and it doesn’t end when they leave. It follows them.

The story

Leaving a violent relationship does not end the danger. Per person, sexual offences by a former partner rose +325% between 2017 and 2026, assaults by a former partner +61% and stalking by a former partner +28%, while assaults by a current partner changed -13%. The violence is concentrated in regional Victoria, is rising fastest for older women, and is rarely a one-off.

What needs to change

Getting off the tram didn’t end my fear; someone had to come. For women leaving violent partners, the same is true. Leaving isn’t the end, and support has to keep showing up after she goes: safety planning and risk assessment that continue long after separation, stalking by an ex-partner treated as the warning sign it is, and services that reach older women and regional Victoria, where rates are highest.

Chart descriptions (text alternatives)

1. After she leaves. Line chart of victim reports per 100,000 Victorians, 2017–2026, by relationship to the alleged offender, with buttons for sexual offences, assault and stalking. Former-partner lines (pink) rise for all three offences; current-partner lines (dark grey) are flat or falling. For sexual offences, former partners overtook current partners in 2023.

2. Where. Choropleth map of 79 Victorian councils shaded by family incidents per 100,000 residents, with a year selector (2022–2026). Rates are highest in Gippsland and the north-west (East Gippsland, Latrobe, Mildura) and lowest in Melbourne’s inner east and bayside.

3. Who. Dot plot for the latest year (offences along the x-axis) comparing the share of victims who are female and the share of alleged offenders who are male for seven offence types. Sexual offences and stalking are the only offences where most victims are female; in each case, over 80% of alleged offenders are male.

4. Age. Dumbbell chart of female victim rates per 100,000 women by age group, comparing 2017–2019 with 2024–2026, with buttons for each offence. Women aged 18–24 face the highest assault rate, but rates for women aged 55 and over are rising fastest (assault, women 65–74: +52%).

5. Repetition. Grouped bar chart with a one-year/ten-year toggle. In one year, 85% of adult men recorded for violence against women have a single incident; over ten years only 52% do, and the 13% with five or more incidents account for 37% of incidents.

Data notes and limitations

• Police-recorded data only; most family and sexual violence is never reported. Trends reflect reporting and recording as well as offending.
• Offenders are alleged, not proven. Victoria Police records sex, not gender.
• Chart 1 relationship data are not split by victim sex; CSA linked-data research (In Fact 13) confirms the same former-partner rise for women harmed by men.
• Charts 1, 2 and 4 use rates so population growth does not inflate trends. Populations were derived from the CSA’s published rates (reports ÷ rate × 100,000). Chart 4 compares three-year averages to avoid cherry-picking end years.
• Chart 5 reproduces published figures from linked unit-record data that is not publicly available.
• Years are financial years ending 30 June. Affected family member = person recorded as harmed; other party = person alleged to have used violence; LGA = council.

Design notes

Dashboard principles (Few, 2006): one screen, no scrolling; the key chart is largest and top-left; one plot type per job; colour used with restraint. One colour rule throughout: pink = the story’s focus, grey = context (semantic consistency; Segel & Heer, 2010). Interactions (Yi et al., 2007): filter (offence buttons, map year), explore (map pan/zoom with Vic and Melb view buttons, legend toggles), reconfigure (1-year vs 10-year), abstract/elaborate (hover and popups). Accessibility: takeaway titles, text descriptions and data tables on this page, fonts ≥ 12px, a red–green-free palette, and colour never the only cue.

References

Crime Statistics Agency. (2024). Characteristics of men’s violence against women and girls in police-recorded crime (In Fact No. 13). https://www.crimestatistics.vic.gov.au/research-and-evaluation/publications/characteristics-of-mens-violence-against-women-and-girls-in

Crime Statistics Agency. (2026). Data tables: Victim reports; Alleged offender incidents; LGA family incidents, year ending June 2026 [Data sets]. https://www.crimestatistics.vic.gov.au/crime-statistics/latest-victorian-crime-data/download-data

Few, S. (2006). Information dashboard design: The effective visual communication of data. O’Reilly.

Mayer, M., & Teucher, A. (2024). ozmaps: Australia maps [R package]. https://CRAN.R-project.org/package=ozmaps

Segel, E., & Heer, J. (2010). Narrative visualization: Telling stories with data. IEEE Transactions on Visualization and Computer Graphics, 16(6), 1139–1148. https://doi.org/10.1109/TVCG.2010.179

Yi, J. S., Kang, Y. A., Stasko, J., & Jacko, J. A. (2007). Toward a deeper understanding of the role of interaction in information visualization. IEEE Transactions on Visualization and Computer Graphics, 13(6), 1224–1231. https://doi.org/10.1109/TVCG.2007.70515

Acknowledgements

CSA data are licensed under Creative Commons Attribution 4.0 International. Boundaries: Australian Bureau of Statistics via {ozmaps}.