Introduction

I work in retail in Melbourne, which made me interested in understanding how retail theft has changed across Victoria. I wanted to find out whether retail theft was changing differently from other recorded offences and how much the figures had changed over time.

Using these charts: Hover over points to see their figures. In Chart 3, search for a council or click a point to highlight it while keeping the other councils visible.

Chart 1: Retail Theft Rose While Other Offence Rates Barely Changed

Recorded retail theft offences increased from 23,276 in 2017 to 43,124 in 2026. The population-adjusted retail theft rate rose by 62.3%, while the combined rate for other offences rose by only 0.1%.

Source: Crime Statistics Agency (2026a), Recorded offences – year ending June 2026, statewide Table 01.

Chart notes: Both lines show recorded offences per 100,000 people, indexed to 2017 = 100. The dotted line shows the starting point. All other recorded offences excludes retail theft (B43). Recorded offences can be affected by reporting and recording practices.

Chart description: Between 2017 and 2020, retail theft rates fluctuated around their starting level. They fell in 2021 and 2022, then increased sharply between 2023 and 2025. In 2026, the retail theft rate index was 162.3, compared with 100.1 for other recorded offences.

Chart 2: Retail Theft Had the Largest Rate Increase Among Theft Categories

Retail theft had the fastest rate growth among the seven B4 theft subgroups (+62.3%). However, theft from a motor vehicle recorded more offences in 2026 (77,348) than retail theft (43,124).

The dots show the percentage change in offence rates, while their size represents the number of recorded offences in 2026.

Source: Crime Statistics Agency (2026a), Recorded offences – year ending June 2026, statewide Table 01.

Chart notes: Each category’s position represents the percentage change in its recorded offence rate per 100,000 people between 2017 and 2026. Bubble area represents the 2026 offence count, which is also shown beside each category. The dotted vertical line means no change. Fare evasion is shown as a separate hollow symbol because it recorded only 35 offences in 2026, compared with 142 in 2017. Its symbol is therefore not scaled by area. Changes based on very small counts need careful interpretation.

Chart description: Retail theft had the greatest increase in its population-adjusted offence rate among the seven B4 theft subgroups. Theft from a motor vehicle recorded more offences in 2026, while bicycle theft, handling stolen goods and fare evasion showed lower rates than in 2017.

Chart 3: Retail Theft Rates Rose Unevenly Across Victorian Councils

Port Phillip, Yarra and Greater Shepparton had three of the largest qualifying increases in retail theft rates. Councils that started with similar rates did not necessarily experience similar increases.

Search for your council or click a point to highlight it against the rest. Each bubble’s position shows its starting rate and change by 2026.

Bubble area represents offences recorded in 2026. Grey points are metropolitan councils and orange points are regional councils.

43 of 79 councils meet the threshold. Of these, 30 are metropolitan and 13 are regional.

Above Victoria’s rate-change benchmark: 14 of 30 metropolitan councils and 8 of 13 regional councils.

These results describe qualifying councils. They do not establish that being metropolitan or regional caused an increase.

Sources: Crime Statistics Agency (2026b), Recorded offences by LGA – year ending June 2026, Table 02; Victorian Government (2025), Know Your Council comparison dashboard: https://www.vic.gov.au/know-your-council-comparison-dashboard

Chart notes: The horizontal axis shows the 2017–19 baseline rate, while the vertical axis shows its change by 2026. Bubble areas are proportional to recorded offences in 2026. Grey points represent metropolitan councils and orange points represent regional councils. The dotted and dashed lines mark Victoria’s corresponding rates. Search for a council to highlight its point without removing the others.

Chart description: Port Phillip, Yarra and Greater Shepparton had some of the largest qualifying increases in retail theft rates. Councils with similar starting rates did not necessarily experience similar changes. Melbourne (†) should be interpreted carefully because its residential population does not include many commuters and visitors.

Chart 4: Licensed Premises and Supermarkets Took a Larger Share of Recorded Retail Theft

Licensed premises increased their share of location-recorded retail theft from 6.0% in 2017 to 11.4% in 2026. Supermarkets rose from 19.2% to 23.0%, while department stores fell from 9.0% to 6.0%.

Switch between Share (%) and Offence counts. Grey dots show 2017 and solid blue dots show 2026. Teal lines show increases in percentage share; muted rose lines show decreases. Bold labels and thicker teal lines emphasise the two headline categories.

Sources: Crime Statistics Agency (2026a), Recorded offences – year ending June 2026, statewide Table 02; Crime Statistics Agency (2025), location-type classifications.

Definition: “Licensed Premises” is a CSA location classification for premises recorded as licensed. The location table does not separate pubs, bottle shops or other venue types, so the specific mix cannot be determined here.

Chart notes: Teal connecting lines mean a category’s percentage share increased; muted rose lines mean it decreased. All eight displayed categories had more recorded offences in 2026. A smaller share does not mean fewer offences. Detailed coverage and denominator checks appear in Technical Notes.

Chart description: Supermarkets and licensed premises took a larger proportion of location-recorded retail theft in 2026, while department stores and shopping complexes took a smaller proportion. The absolute number of recorded offences increased in all eight categories.

Chart 5: Retail Theft Had a Lower Unsolved Share Than Most Other Theft Types in 2025

In 2025, 39.5% of recorded retail theft offences were marked Unsolved (16,573 of 41,987), compared with higher shares in four of the six other theft subgroups. Yet 16,573 retail theft offences carried this status—nearly twice the 8,656 recorded for 2017. This comparison describes police-recorded statuses, not convictions or final outcomes.

Use the status selector to switch investigation status; blue = retail theft.

Sources: Crime Statistics Agency (2026a), Recorded offences – year ending June 2026, statewide Table 04; Crime Statistics Agency (2018), section 2.5: Investigation status, which describes investigation status as a point-in-time record.

Chart notes: All categories use the year ending June 2025. Rows remain in descending order of the default Unsolved percentage when the status is changed. Four of six other B4 theft subgroups had a higher recorded Unsolved share than retail theft. Fare evasion uses a hollow marker because only 43 offences were recorded in 2025; percentages for this subgroup are sensitive to small changes. Handling stolen goods and fare evasion show particularly low Unsolved shares and high Arrest/Summons shares. These offences can be detected or recorded under circumstances where an alleged offender has already been identified, unlike some theft offences reported after the event; the dataset does not establish the reason for the difference. Investigation status is a snapshot at extraction, not a final outcome. Unsolved does not mean a case was permanently closed, and Arrest/Summons does not mean a conviction. The chart does not measure police effectiveness.

Chart description: The recorded Unsolved share for retail theft was 39.5% in 2025, compared with 66.1% for motor vehicle theft, 75.2% for other theft, 87.3% for bicycle theft, and 92.7% for theft from motor vehicles. Handling stolen goods (2.0%) and fare evasion (0.0%, 43 offences in total) had smaller recorded Unsolved shares. Retail theft nevertheless had 16,573 offences recorded as Unsolved, nearly twice the 8,656 in 2017.

What this means

Working in retail in Melbourne made me interested in whether the increase in recorded retail theft was part of a wider crime trend. The figures show that retail theft rose much faster than other recorded offences, but that its increase differed across councils and types of premises. In 2025, retail theft had a smaller Unsolved share than four of the six other theft subgroups, although the number recorded as Unsolved was nearly double its 2017 level. These are police-recorded offences and statuses, not proof of what caused the changes; for me, they show why the growing volume of retail theft matters even when its percentage share does not change dramatically.

Technical Notes

Definitions and period. Retail theft is offence subgroup B43 (Steal from a retail store). Every year refers to the 12 months ending June, not the calendar year. Charts 1 and 2 compare 2017 with 2026. Chart 1 indexes the recorded offence rate per 100,000 residents to 2017 = 100; this differs from growth in raw counts. Recorded crime depends on offences coming to police attention and how they are classified or reported.

Council comparison. Chart 3 uses the average of 2017–2019 for a less volatile local baseline, then compares this with 2026. It includes only LGAs with all three baseline years present, at least 100 retail theft offences on average per baseline year, and at least 100 in 2026. This yields 43 of 79 LGAs; conclusions are restricted to those included. The metropolitan group combines the Metro and Interface council comparison groups from the Victorian Government; other LGAs are labelled regional. Rates use estimated resident population, which does not include the number of shoppers, commuters and visitors present at a particular time. This is particularly important for Melbourne and tourist or shopping precincts. These area types are descriptions, not explanations of an increase.

Location classification and denominators. Chart 4 compares the eight largest location groups in the 2026 location table. The eight groups represent 97.5% of its retail theft offences; the remaining 2.5% fall within other categories. Percentages use all Table 02 location records for each year: 23,192 in 2017 and 42,984 in 2026. The corresponding statewide offence totals in Table 01 are 23,276 and 43,124, leaving differences of 84 and 140 respectively between the tables. The source totals do not tell us why those records are absent from the location breakdown, so no cause is assumed. A declining share of a location type can occur even when its offence count rises. The category Licensed Premises does not identify the mixture of venue types within it.

Investigation status. Chart 5 compares theft subgroups within the same 2025 offence year, reducing differences in time available to progress investigations across categories. A recorded status is a point-in-time snapshot and may subsequently change (Crime Statistics Agency, 2018). Unsolved does not necessarily mean permanently closed or ignored; Arrest/Summons is police action, not proof of guilt or conviction. Fare evasion had just 43 recorded offences in 2025 and is shown with a hollow dot. Some offence types may be detected when a suspected offender is already identifiable, so differences in status shares cannot alone measure police effectiveness. A 2017–2025 comparison of Unsolved counts (8,656 versus 16,573) describes volume and does not establish the causes of change.

References

Anthropic. (2026). Claude (Opus 5.5) [Large language model]. https://claude.ai/

Crime Statistics Agency. (2018). Crime statistics Victoria: Year ending September 2018 [Statistical report]. https://files.crimestatistics.vic.gov.au/2021-07/Crime%20Statistics%20Victoria%20year%20ending%20September%202018.pdf

Crime Statistics Agency. (2025). Crime Statistics Agency data tables – Location type classifications [Data set]. DataVic. https://discover.data.vic.gov.au/dataset/crime-statistics-agency-data-tables-location-type-classifications

Crime Statistics Agency. (2026a). Recorded offences – year ending June 2026 [Data set]. Victorian Government. https://discover.data.vic.gov.au/dataset/data-tables-recorded-offences/resource/35098a5b-d906-446f-a7fb-eaf07270ce45

Crime Statistics Agency. (2026b). Recorded offences by LGA – year ending June 2026 [Data set]. Victorian Government. https://discover.data.vic.gov.au/dataset/data-tables-recorded-offences/resource/06bb24c8-8f62-4144-a746-99ef5492e4f9

Cheng, J., & Sievert, C. (2025). crosstalk: Inter-widget interactivity for HTML widgets [R package]. https://rstudio.github.io/crosstalk/

R Core Team. (2026). R: A language and environment for statistical computing [Computer software]. R Foundation for Statistical Computing. https://www.R-project.org/

Sievert, C. (2020). Interactive web-based data visualization with R, plotly, and shiny. Chapman and Hall/CRC. https://plotly-r.com/

Victorian Government. (2025). Know Your Council comparison dashboard. https://www.vic.gov.au/know-your-council-comparison-dashboard

Wickham, H. (2016). ggplot2: Elegant graphics for data analysis. Springer. https://ggplot2.tidyverse.org/

Wickham, H., & Bryan, J. (2026). readxl: Read Excel files [R package]. https://readxl.tidyverse.org/

Wickham, H., Pedersen, T. L., & Seidel, D. (2025). scales: Scale functions for visualization [R package]. https://scales.r-lib.org/

Xie, Y. (2015). Dynamic documents with R and knitr (2nd ed.). Chapman and Hall/CRC. https://yihui.org/knitr/

Xie, Y., Allaire, J. J., & Grolemund, G. (2018). R Markdown: The definitive guide. Chapman and Hall/CRC. https://yihui.org/rmarkdown/

Acknowledgement

I used Claude (Anthropic, 2026) to get suggestions for improving my written work, checking spelling and grammar, and making my explanations clearer. I also used it to suggest relevant references and online sources, which I reviewed against the original sources. I used it for technical guidance with R programming, visualisation code, troubleshooting errors, and publishing my interactive report online. I checked the data, calculations, and visualisations to ensure the final report was accurate.