What disappears when Victoria’s family violence experience is reduced to a single statewide number?
Family violence is often represented through a single figure: the number of incidents recorded across Victoria. This provides an important indication of scale, but the statewide total combines incidents from communities with very different characteristics and experiences. It also brings together different forms of abuse, different victim–perpetrator relationships, and incidents that receive different levels of police-assessed risk.
Between 2020–21 and 2024–25, 482,333 family violence incidents were recorded across Victoria. Considered on its own, this figure shows the overall scale of recorded family violence, but it does not show how that total is distributed across the state or what is contained within it.
A closer examination reveals several distinct patterns. Recorded rates vary substantially between local government areas, meaning that the statewide figure does not reflect the experience of every community equally. The forms of abuse recorded have also changed at different rates over the five-year period. Looking at victim and perpetrator age groups reveals further differences in the relationships represented in the data. Finally, police-assessed risk provides another perspective by showing how the likelihood of future violence was assessed at the time incidents were recorded.
The following analysis therefore moves from the overall scale of recorded incidents to differences in place, form, people and risk. Together, these measures provide a more detailed view of what lies beneath Victoria’s statewide total.
The analysis uses the Victorian Crime Statistics Agency’s Family Violence Data Tables, covering the period from 1 July 2020 to 30 June 2025.
The five final visualisations draw on two published workbooks:
The Service Sector Data Tables 2025 were also reviewed during the data audit to understand the available family violence measures, but they were not required for the final visualisations.
The interactive map also uses a Victorian Government DataVic spatial dataset containing local government area boundaries. These geographic boundaries were used to display the Crime Statistics Agency’s local government area rates spatially and do not provide the family violence measures themselves.
The published Excel workbooks were imported directly into R, with the relevant tables selected and reshaped for analysis and visualisation. Values used in the final visualisations were therefore derived from the published source tables rather than manually entered or hard-coded into the analysis.
Data preparation included checking table structures, identifying missing and suppressed values, checking relevant records for consistency, converting measures to appropriate numeric formats, and using the published All Victoria values where statewide figures were available.
The source tables measure different aspects of recorded family violence and therefore use different units and definitions. The analysis distinguishes between recorded incidents, rates per 100,000 people, abuse-type records, unique perpetrator–victim pairs, and police risk assessments. These measures are not treated as directly interchangeable, and their specific meanings are explained when each visualisation is interpreted.
The data represent family violence recorded by police and included in the published datasets. They should not be interpreted as a measure of all family violence occurring in Victoria. Recorded figures can be influenced by whether incidents are reported, how incidents are identified and recorded, and other factors affecting police-recorded data.
The datasets also contain measures with different definitions. Abuse-type categories are not mutually exclusive, meaning that a single recorded incident may involve more than one form of abuse. The victim–perpetrator age data represent unique perpetrator–victim pairs rather than incident counts, while the risk data represent police assessments made at the time of recorded incidents. These distinctions are retained throughout the analysis and are considered when interpreting the visualisations.
The statewide number provides the clearest starting point for understanding the overall scale of recorded family violence in Victoria. Across the five-year period, 482,333 family violence incidents were recorded.
However, the trend is not simply a steady increase. Recorded incidents fell from 93,439 in 2020–21 to 90,540 in 2021–22, before increasing each year afterwards. By 2024–25, the number had reached 106,430 incidents, around 13.9% higher than in 2020–21.
The statewide total provides an important starting point, but it also hides variation underneath the number. It does not show where recorded rates are highest, which forms of abuse are changing, or who is involved. The following visualisations progressively examine these different dimensions.
Source: Victorian Crime Statistics Agency, Justice System Data Tables 2025.
The change after 2021–22 is clear: recorded incidents increased in every subsequent year, reaching the highest level in the series in 2024–25. This establishes the overall direction of recorded family violence across Victoria, but the statewide figure still treats the state as one combined population.
The next question is therefore where these recorded rates are highest. Looking across local government areas reveals differences that cannot be seen in the statewide total alone.
A statewide total can make differences between communities difficult to see. When the figures are adjusted for population, substantial variation emerges across local government areas. In 2024–25, East Gippsland recorded the highest rate, followed by areas including Latrobe, Mildura and Horsham, while several metropolitan LGAs recorded much lower rates.
These differences are important because raw incident counts alone would largely reflect the size of each local population. Expressing the figures as a rate per 100,000 people provides a more meaningful basis for comparing LGAs of different sizes. The map also allows the geographic pattern to be examined across all five years, making it possible to see whether higher-rate areas remain consistently elevated or change over time.
The distribution is therefore considerably more varied than the statewide figure suggests. A single Victorian rate cannot capture the differences between individual communities, particularly when some LGAs record rates several times higher than others.
Source: Victorian Crime Statistics Agency, Data by Local Government Area Data Tables 2025; Victorian Government DataVic, Vicmap Local Government Areas.
The geographic pattern shows that recorded family violence rates are not evenly distributed across Victoria. Some LGAs consistently record substantially higher rates than others, while the differences between areas remain visible across the five-year period. The 2024–25 distribution ranges from 606 recorded incidents per 100,000 people in Boroondara to 5,022 in East Gippsland, illustrating the scale of variation that is hidden within the statewide total.
At the same time, the map does not explain why these differences occur. Recorded rates can reflect a range of factors, including reporting and recording patterns, and should not be interpreted as a direct measure of the underlying prevalence of family violence in each community. What the data establish is that the experience represented in police-recorded figures varies considerably by location.
Location is only one part of the picture. The next section examines what forms of family violence are being recorded and whether different types of abuse have changed at the same rate over time.
The statewide incident total does not show what forms of abuse are being recorded within family violence incidents. To examine this dimension, the analysis uses the official “All Victoria” rows from Table 3 of the Victorian Crime Statistics Agency’s Data by Local Government Area Data Tables 2025. These statewide rows are used rather than summing individual LGAs because some LGA-level values are suppressed.
The categories represent different forms of recorded abuse, but they are not mutually exclusive. A single family violence incident can therefore appear in more than one abuse category. For this reason, the visualisation does not compare the categories as shares of all incidents. Instead, each category is indexed to its 2020–21 level of 100, allowing the relative change in each form to be compared over the five-year period.
This approach shifts the focus from how many incidents were recorded overall to how different forms of recorded abuse have changed relative to their own starting points.
Source: Victorian Crime Statistics Agency, Data by Local Government Area Data Tables 2025.
The differences between the categories become clear over the five-year period. Sexual abuse increased by approximately 22.7% between 2020–21 and 2024–25, the largest relative increase among the categories shown. Emotional abuse also increased substantially, by approximately 17.1%, while physical abuse increased by around 9.1% and other abuse by approximately 6.9%.
In comparison, verbal abuse increased by only around 2.7%, while economic abuse remained almost unchanged, decreasing slightly by approximately 0.4%. The overall rise in recorded family violence therefore does not represent a uniform change across all forms of abuse. Different categories have followed noticeably different trajectories.
These differences add another layer to the statewide picture. Knowing where recorded rates are higher tells us about place, while comparing abuse types shows what is changing within the overall total. The next section looks at the people represented in the data by examining victim and perpetrator age relationships.
Note: Abuse-type categories are not mutually exclusive, so a single recorded family violence incident may involve more than one form of abuse. The visualisation compares relative change from 2020–21 and should not be interpreted as the proportion of incidents involving each form of abuse.
The geographic and abuse-type analyses show where family violence is recorded and how different forms have changed, but they do not show the age relationships between the people involved. Table 5 provides another perspective by recording unique perpetrator–victim pairs across three victim age groups and three perpetrator age groups.
The analysis uses the official All Victoria rows because some LGA-level values are suppressed. Rather than comparing the raw number of pairs alone, the visualisation expresses each perpetrator age group as a share within its victim age group. This makes the age relationships easier to compare across children, adults and older victims, despite the large difference in the number of recorded pairs between these groups.
The resulting matrix allows the age composition of each victim group to be examined across the five years, showing whether the same perpetrator age patterns are present for children, adults and older victims.
Source: Victorian Crime Statistics Agency, Data by Local Government Area Data Tables 2025.
The age patterns are clearly different across the three victim groups. For adult victims aged 18–64, perpetrators aged 18–64 account for around 91–92% of unique perpetrator–victim pairs throughout the period, while perpetrators under 17 account for approximately 7% and those aged 65+ account for around 2%.
The pattern is different for child victims under 17. Perpetrators aged 18–64 still represent the largest share, but the proportion involving perpetrators under 17 is much higher, increasing from approximately 16% in 2020–21 to 17% in 2024–25. For elderly victims aged 65+, perpetrators aged 18–64 account for roughly 79–81%, while perpetrators aged 65+ represent approximately 15–17% of the recorded pairs.
These differences show that there is no single victim–perpetrator age pattern that represents all recorded family violence. The relationships vary substantially depending on the age group of the victim, and those differences would be difficult to see in the statewide incident total alone. Importantly, these measures represent unique perpetrator–victim pairs rather than incident counts, so they should not be interpreted as the number of family violence incidents.
Age provides another layer of the recorded data, but it does not indicate how police assessed the potential for further violence. The final visualisation therefore examines police-assessed risk across the same five-year period.
The final dimension considers how police assessed the potential for future violence within recorded family violence incidents. Victoria Police’s L17 risk assessment records a score from 0 to 16 and identifies whether the assessment indicated that future violence was likely.
This measure is different from the incident counts examined earlier. It reflects a police assessment made at the time of the recorded incident, rather than a measure of whether future violence subsequently occurred. The visualisation therefore focuses on the proportion of incidents with a recorded L17 score that were assessed as “future violence likely” across the five-year period.
Source: Victorian Crime Statistics Agency, Justice System Data Tables 2025.
The proportion assessed as “future violence likely” increased consistently across the period, from 73.2% in 2020–21 to 77.1% in 2024–25, an increase of 3.9 percentage points. By 2024–25, more than three-quarters of incidents with a recorded risk assessment were therefore classified in this category.
The increase provides an additional perspective on the statewide figures. It does not mean that future violence occurred in these incidents, nor does it predict what will happen to an individual victim or perpetrator. Instead, it indicates a change in the level of risk identified by police at the time the incidents were assessed.
Across the five visualisations, the statewide total can therefore be viewed from several different perspectives: where recorded rates vary, which forms of abuse are changing, how victim–perpetrator age relationships differ, and how police-assessed risk changes over time. Together, these dimensions provide a more detailed understanding of what lies beneath the overall number of recorded incidents.
The statewide figure provides an important measure of the scale of family violence recorded in Victoria, with 482,333 incidents recorded between 2020–21 and 2024–25. However, the analysis shows that this number becomes much more informative when examined across different dimensions.
Recorded rates vary substantially between local government areas, showing that family violence is not represented evenly across Victoria. The forms of abuse recorded have also changed at different rates, with sexual and emotional abuse increasing more strongly than categories such as verbal and economic abuse. The victim–perpetrator age analysis further shows that the relationships represented in the data differ across victim age groups, while police-assessed risk indicates that the share of risk-scored incidents classified as “future violence likely” increased from 73.2% to 77.1% over the period.
Together, these measures show that the statewide total captures the overall scale but not the variation within it. Place, form, people and risk each reveal information that is hidden by a single statewide number. At the same time, these are police-recorded data and should not be interpreted as a complete measure of family violence occurring in Victoria.
Ultimately, 482,333 incidents are not 482,333 identical records. They occur in different communities, involve different forms of abuse, represent different victim–perpetrator age relationships, and receive different police-assessed risk classifications. Looking beyond the total makes these differences visible and provides a more complete understanding of what the published data can — and cannot — tell us about family violence in Victoria.
Crime Statistics Agency Victoria. (2025). Data by local government area data tables 2025 [Data set]. Victorian Government. https://www.crimestatistics.vic.gov.au/family-violence-data/family-violence-data-tables
Crime Statistics Agency Victoria. (2025). Justice system data tables 2025 [Data set]. Victorian Government. https://www.crimestatistics.vic.gov.au/family-violence-data/family-violence-data-tables
Department of Transport and Planning. (2025). Vicmap Land Administration Themes – Local Government Area Polygon Aligned to Property (VLATLGAPOLYGON) [Spatial data set]. Victorian Government DataVic. https://discover.data.vic.gov.au/en_AU/dataset/vicmap-land-administration-themes-local-government-area-polygon-aligned-to-property
The author acknowledges the Victorian Crime Statistics Agency and the Victorian Government for providing the open datasets used in this assignment. The Victorian Government DataVic spatial dataset was used for the local government area boundaries in the interactive map.
Generative AI disclosure: ChatGPT (OpenAI) was used to refine written work, review spelling and grammar, suggest relevant references and online sources, and provide technical assistance with R and the R Markdown publishing workflow. The author reviewed and adapted all suggestions and is responsible for the final analysis, visualisations, interpretation and content of the assignment.