GOVERNMENT TOOLS
CONSULTING
What Melbourne looks like if the growth boundary holds, and if it doesn't
Melbourne's urban growth boundary was drawn in 2002 to mark the point where the city would stop spreading outward. It is one of the questions in play at November's state election, and it is a version of an argument happening in every Australian capital.
The Age commissioned us to model what happens either way. We looked at where Melbourne's built-up area has been since 1883, and where it goes by 2101 if the boundary holds, and if it doesn't.
We are not going to tell you which city to choose. We can show you both, and what each one asks of us.
This is the short version. We will be publishing more from this project over the coming weeks, including maps and charts area by area across Melbourne. Follow us on LinkedIn to see them as they go up.
101 km²
The whole of Melbourne's built-up area in 1883.
2,702 km²
By 2101 if the growth boundary holds. The two scenarios barely differ before 2046, then separate.
3,207 km²
By 2101 if it doesn't, with the city still spreading outward.
Watch 218 years of Melbourne in one map
Our interactive map draws Greater Melbourne's built-up area from 1883 to 2101. Slide through the years to see the city spread, then switch between the two futures to compare them.
In the media
The Age published this work as a series through August, ahead of Melbourne Assembly in October. Johnny Barnard, who led the forecasts, went on 3AW to talk through what they show.
3AW radio. Listen to Johnny Barnard on what Melbourne looks like at 10 million.
The Age, 21 August 2026. Labor pledges to halt Melbourne's urban sprawl, but planners ask if we're heading to a megacity mess. The government's position, ahead of the November election.
The Age, 24 August 2026. These suburbs may never be the same as Melbourne's population soars to 10 million. Both scenarios, suburb by suburb.
The Age, 27 August 2026. How a 'halo of satellite cities' could absorb half of Victoria's swelling population. Whether decentralisation is a real alternative, and Johnny's case for decentralising within Melbourne instead.
The choice on the table
Melbourne is unusual in having drawn the line explicitly. The growth boundary is a political construct rather than a physical fact: a line on a map, drawn somewhat arbitrarily, that states an intention to stop the city expanding outward. Sydney, Brisbane, Perth and Adelaide have no direct equivalent, though all of them are having a version of the same argument. What Melbourne's line gives us is a clean way to model two versions of one city, which is harder to do where the constraint is implicit.
Melbourne is on track for just over 10 million people by 2101, up from 5.48 million today. That is about 4.5 million extra residents and close to two million extra homes. The same number arrives under either scenario. What changes is where they go.
The growth boundary loops around Pakenham, Cranbourne, Lilydale, Epping, Wallan, Melton and Werribee. Melbourne's established suburbs take around a million new homes either way. If it holds, they take roughly half a million more on top of that, homes that would otherwise have gone onto new land at the edge. If it doesn't, the built-up area keeps extending into farmland and green space at roughly the rate it has for the past two centuries.
Neither of these is a prediction. Getting from a population total to where people actually live is the harder half of the work, and it's what we've spent more than 30 years doing across Australian cities. It draws on what we know about land supply, housing markets, and how a place changes once the first estates or the first towers arrive.
The table below takes eight areas where the two scenarios pull hardest in opposite directions. The first five take more growth if the boundary holds. The last three take more if it doesn't. Every one of them grows either way.
Population growth to 2101, SA2 by SA2
| SA2 | 2026 population | Boundary holds | Boundary doesn't hold |
|---|---|---|---|
| Sunshine | 10,600 | +304% | +151% |
| Footscray | 25,500 | +129% | +91% |
| Box Hill | 27,800 | +105% | +92% |
| Brighton (Vic.) | 25,000 | +96% | +54% |
| Kew East | 6,700 | +72% | +16% |
| Greenvale - Bulla | 29,200 | +91% | +195% |
| Pearcedale - Tooradin | 8,100 | +31% | +659% |
| Kurunjang - Toolern Vale | 11,900 | +27% | +992% |
Percentages are growth on each area's own 2026 population, out to 2101. These are ABS Statistical Area Level 2 areas on 2021 boundaries, which share their names with suburbs but not always their edges. Figures are rounded, and they sit at the end of a 75-year scenario, so read them as a trajectory rather than an estimate.
The two columns cost different things. Holding the boundary puts about 490,000 homes into established suburbs that would otherwise have gone onto new land, which means activity centres absorbing 768,000 new homes rather than 477,000, and infill absorbing 648,000 rather than 448,000. Letting the boundary go costs farmland and green space, longer commutes, and infrastructure that has historically arrived years after the residents do.
Source: .id (informed decisions), Melbourne 2100 forecasts, August 2026.
What each scenario assumes
Births, deaths and migration tell you how many people are coming. They tell you nothing about where those people will live. That depends on what actually gets built, which is governed by land supply, development economics, infrastructure timing and planning approvals.
Both scenarios rest on an assumption that is not yet true, and it is worth being clear about both.
Containment assumes we build up at a rate we have not reached.
Victoria has committed to 70 per cent of new homes in established suburbs and 30 per cent in greenfield areas. Containment reaches 78 per cent inside the existing footprint. The current paradigm delivers 54 per cent, because new land keeps being released beyond the committed estates. Townhouse approvals are rising, which helps, but most are landing in the affordable growth areas on the edge rather than the established suburbs the strategy depends on, and apartment construction is running short of what the housing plan assumed. About a third of the containment task is infill spread through existing suburbs, which is less visible than towers but has to happen street by street.
Continued outward growth assumes infrastructure keeps up, which it has not.
The City of Melton takes about 200 new residents a week, and more than 80 per cent of residents leave the area each day for work. Electrification of the Melton line, promised in 2018, is still not delivered. Growth on the fringe is cheaper and faster to build, and that is a genuine advantage for affordability. The cost shows up later, in travel time and in services that arrive after the people do.
So we build these scenarios from what's being built, rather than from what the strategy sets out. That means tracking lot releases, approvals, how quickly estates fill and where apartment projects stall, then working out what those patterns do to a particular suburb over time. It's detailed, place-by-place work, and we do it for every region in the country.
Scenarios show you the range. Forecasts tell you what to plan for.
Where we sit
This was a custom exercise, commissioned by The Age, asking what happens to a whole city over 75 years under two policy decisions. It is built to make a choice visible, not to say what will happen, and nobody should plan a budget around a 2101 number.
For decisions being made now, you want the National Forecasting Program: annual, small area, nationally consistent, running 20 years ahead and updated as new evidence arrives. It is what councils, developers, utilities and service planners use, and it is the evidence base that made this Melbourne work possible.
What is driving change where you are
Method and assumptions
The team
The people who built it
Johnny Barnard led the work and Oliver Bowering led the modelling, drawing on the team that maintains our national evidence base region by region.
Johnny Barnard
Founder of the National Forecasting Program
Oliver Bowering
Senior Forecast Consultant
Richard Thornton
Head of Forecasting
Liza Ivanova
Principal Forecast Consultant
Chris Jones
Forecaster
Planning for growth in an area of your own?
Work with us
We forecast population, housing and demand for every area in Australia. If you are planning infrastructure, sizing future demand, timing capital expenditure, or working out where and when to open a new site, we can build the picture for the places you need to understand.
What happens next
The Age and the University of Melbourne are hosting Melbourne Assembly from 19 to 21 October 2026, a three-day forum on the future of the city. We will be there, and we will keep adding to this page as the series and the coverage continue.
For the analysis behind the headlines, subscribe to .id Insight.