On 28 November 2026, Victoria will elect a new parliament. As an international student at RMIT, I won’t be eligible to vote. I’m one of an estimated 712,000 adults living in Victoria who are not Australian citizens.
That represents roughly one in seven adults across the state. Non-citizens live, work and study in the same communities as eligible voters, relying on the same public transport, healthcare and other services. Yet many cannot participate in choosing the state government that makes decisions affecting their lives.
This report examines where Victoria’s non-citizen population lives, who these residents are, and how their distribution across electoral districts reveals a gap between the people who live in a community and those eligible to vote for its representative.
Non-citizens are not evenly distributed across Victoria. The map shows a state divided into very different experiences of electoral eligibility. The darker the district, the greater the share of adults who can’t vote.
In Melbourne, highlighted in teal, around 42% of adults are non-citizens — more than two in every five people living in the district. In many regional districts, the share is closer to one in twenty.
Outside the city centre, the map reveals several clusters: Laverton (29%), Tarneit (27%) and Point Cook (25%) in the west; Oakleigh (28%) and Dandenong (27%) in the south-east; and Box Hill (25%) in the inner east.
These districts are geographically spread out. They have different communities, housing markets, workplaces and connections to universities. Yet they share something that a conventional electoral map doesn’t immediately reveal: a substantial part of their adult population cannot participate in choosing their state representative. To understand what this means for their communities, it helps to look beyond where these residents live and ask who they are.
As an international student, I initially wondered whether these patterns were mainly driven by students like me. The next chart breaks down non-citizen adults by tertiary student status to see how much of the picture they account for, and it tells a more complicated story.
In Melbourne, tertiary students make up around 16% of all adults and nearly four in ten non-citizen adults. Oakleigh, near Monash University’s Clayton campus, and Box Hill also stand out.
But even in these districts, students are not the majority of people who can’t vote. Across Victoria, only about one in six non-citizen adults is a tertiary student, meaning the population unable to vote extends well beyond university campuses.
These residents may include workers, partners and families, although student status alone cannot tell us how long someone has been here or how deeply they are connected to their community.
There is also an important limitation to this picture. The Census was conducted in August 2021, when Australia’s international borders were closed during the COVID-19 pandemic. Many international students were unable to enter the country, so the student figures reflect an unusual period rather than a typical year.
Still, the broader point holds: people who can’t vote are not a single group of temporary students. They are a diverse part of the population, concentrated in particular districts. And where these residents are concentrated matters because electoral boundaries are based on enrolled voters, not the total number of people living in each district.
Victoria’s electoral boundaries are designed around enrolled voters, rather than the total number of adults living in each district. They are not designed to give every district the same number of adult residents.
The difference becomes clear when the two measures are placed side by side.
The number of enrolled voters, marked in orange, barely moves: districts have around 46,000 to 55,000 enrolled voters. The number of adult residents, however, ranges from around 50,000 to more than 109,000.
In Sunbury, around 92% of adults are enrolled to vote. In Melbourne, the figure is closer to half.
The districts with the largest adult populations are also among those where non-citizens make up a substantial share of residents. Where many adults are ineligible to enrol, the total population can grow without a corresponding increase in the enrolled voters.
This means a Member of Parliament (MP) in Melbourne may represent roughly twice as many adult residents as an MP in a more typical district, including the citizens who live alongside those who cannot vote.
The distinction is important: electoral boundaries aim to balance the number of enrolled voters, not the number of residents. As a result, equal-sized electorates do not necessarily mean equal-sized communities.
Imagine following 100 adults in two different parts of Victoria through the 2022 state election. In districts where fewer than 10% of adults are non-citizens, around 79 had a vote counted. In Melbourne, that number was just 39.
At first glance, the difference might suggest that people in Melbourne are less engaged with elections. But the breakdown tells another story.
The shares of adults who were enrolled but did not vote, or whose ballots were informal, are much more similar — around 10 and 4 out of every 100 adults, respectively. The largest difference came before election day: adults who are not enrolled, many of whom are ineligible because they are not Australian citizens, account for a substantial part of the difference.
In other words, the lower number of votes counted per 100 adults does not necessarily mean eligible voters in these districts are less willing to participate. A large part of the difference reflects who is eligible to vote in the first place.
The Melbourne district is also included in the group where more than 20% of adults are non-citizens, so the two categories overlap. Even so, the comparison highlights how the composition of a district can affect how many votes are counted relative to the population living there.
(A vote count tells us who participated. Looking at the whole population helps explain who was able to participate at all.)
The statewide patterns are not the whole story. Every district has its own combination of residents, enrolled voters and electoral participation. Search for your district or select any dot below to see how it compares with the Victorian average across all four measures.
Most districts sit relatively close to the Victorian average, but a few stand out. Melbourne, Oakleigh, Laverton and Box Hill have appeared repeatedly throughout this report, with higher shares of non-citizen adults and lower enrolment and vote-counted figures relative to their adult populations.
Their positions aren’t just isolated statistics. They connect the patterns on the first map to the differences in representation and participation explored throughout this report.
Your district might look very different from mine. But comparing them makes one thing clear: where you live can shape how many people around you are eligible to participate in an election, even before a single vote is cast.
On election night, the result in my district will be decided by votes cast by a fraction of the adults who live there. That is not a mistake in the counting. It reflects the rules that determine who is eligible to vote and how electoral boundaries are drawn.
Citizenship determines eligibility for Victoria’s state election, but it does not determine whether someone lives in a community, works in it or depends on its public services. Across the state, people who cannot vote are concentrated in particular districts, creating a gap between the number of adults living in a community and the electorate used to determine its boundaries.
The data cannot tell us whether these residents feel represented, or whether their interests receive less attention from elected officials. But it does reveal a structural difference: an MP may represent a similar number of enrolled voters to a colleague while serving a much larger adult population.
I began this story thinking about an election I couldn’t vote in. The data led me to a broader question: when we talk about who a district represents, should we consider only the people eligible to vote, or everyone who calls that place home?
On 28 November, I’ll be watching the results like everyone else. I live here, too. I just won’t have a vote in deciding them.
Anthropic. (2026). Claude (Sonnet 5.5 version). https://claude.ai
Australian Bureau of Statistics. (2022a). Australian Statistical Geography Standard (ASGS) Edition 3: State electoral divisions 2022. https://www.abs.gov.au/statistics/standards/australian-statistical-geography-standard-asgs-edition-3/jul2021-jun2026/access-and-downloads/digital-boundary-files
Australian Bureau of Statistics. (2022b). Census of Population and Housing, 2021. TableBuilder. Retrieved October 20, 2026, from https://www.abs.gov.au/statistics/microdata-tablebuilder/tablebuilder
Brideau, L. (2024, May 16). Spatial R: Using the sf package. University of Virginia Library. https://library.virginia.edu/data/articles/spatial-r-using-sf-package
Cheng, J., Schloerke, B., Karambelkar, B., Xie, Y., & Aden-Buie, G. (n.d.-a). Choropleths. Leaflet for R. Retrieved October 22, 2026, from https://rstudio.github.io/leaflet/articles/choropleths.html
Cheng, J., Schloerke, B., Karambelkar, B., Xie, Y., & Aden-Buie, G. (n.d.-b). Using basemaps. Leaflet for R. Retrieved October 23, 2026, from https://rstudio.github.io/leaflet/articles/basemaps.html
Pebesma, E. (n.d.). Reading, writing and converting simple features. sf: Simple Features for R. Retrieved October 22, 2026, from https://r-spatial.github.io/sf/articles/sf2.html
Sievert, C. (2020). Interactive web-based data visualization with R, plotly, and shiny. Chapman and Hall/CRC. https://plotly-r.com/client-side-linking.html
Teucher, A. (n.d.). rmapshaper basics. The Comprehensive R Archive Network. Retrieved October 22, 2026, from https://cran.r-project.org/web/packages/rmapshaper/vignettes/rmapshaper.html
Victorian Electoral Commission. (2022). 2022 State election results. https://www.vec.vic.gov.au/results/state-election-results/2022-state-election-results
The online guides listed in the References helped me learn how to read spatial files, simplify boundaries, build the choropleth map and add linked highlighting in plotly.
I used Claude (Anthropic, 2026) as a supporting tool throughout this project. It helped me troubleshoot R code, clarify data-processing steps, and improve the wording and formatting of my report. I am responsible for the analysis, design decisions, interpretation of the data and final content of this report.