Growth: Suitability and Allocation Across an Area

Across a council or an area you draw: how suitable each lot is, how many homes the land could hold, and where a total would go. It's a first screen, not a housing capacity assessment, and not a statement of where growth should go. Open Growth in the sidebar. It needs an account, and reading an area takes a minute or two.

What it reads

Every lot on title, one polygon each, with its zone, overlays, buildings and estimated homes. Stacked titles (apartments, units) share their polygon.

• Roads, reserves and common property are left out and counted.
• Homes and people are ABS Mesh Block Counts 2021, shared across each block by address points, as for a site. Checked against Victoria in Future's 2021 dwellings for councils, they come within a few per cent.
• Walks go along Overture's streets, not in a straight line, to the nearest train station, tram stop, centre and shop.

Suitability

Two scores, side by side:

• Normalised: each factor scaled 0 to 1 and combined with weights you set, shown in six levels.
• Binary: suitable only if every test is true: a housing zone, over 70% of permitted floor area unbuilt, a lot over 500 m², within 800 m of a station and of shops, and no Heritage Overlay. Building age, part of the published screen, isn't known across an area, so it isn't tested.

The weights and the zone values are UrbanMind's starting points, not findings. Zone values: Residential Growth, Housing Choice and Transport, Mixed Use, Activity Centre, Capital City and Docklands 1.0; Commercial 1 0.9; Comprehensive Development, Priority Development and Urban Growth 0.8; General Residential 0.6; Township 0.4; Neighbourhood Residential 0.3; Low Density Residential 0.1.

What isn't known stays unknown. A factor that isn't known for a lot drops out and the others are reweighted. Under half the weight known, the lot has no score. There's no neutral value.

Sensitivity: each weight varied by up to half, at random, 30 times, to see how many of the top tenth of lots stay there. Under 60%, the screen says the answer is weak.

Taken out, and for a closer look

Taken out (Not suitable): zones not meant for housing as a main use; Flood, Road Flood and Public Acquisition Overlays; lots under 100 m²; lots under an owners corporation or with stacked titles, as redeveloping them needs every owner.

For a closer look, staying in: heritage, bushfire, flood (Land Subject to Inundation, Special Building), contamination, erosion, and lots over 2 ha outside the Urban Growth Zone, which are usually a school, club or institution. Heritage and large lots are left out of allocation unless you put them back, and their capacity is shown apart.

Capacity

More homes each lot's zone would allow, less the homes estimated there. Set against what Fill places on real lots:

• Townhouses and units: one home per 300 m² of lot.
• Apartments: on lots from 1,100 m², in zones that invite them, with at least 3 storeys: lot × 45% × storeys × 80% ÷ 85 m².
• Urban Growth Zone: 20 homes per net hectare, on 65% of the gross land.
• Low Density Residential: one home per 4,000 m².
• Heights a schedule or overlay sets are assumed (Activity Centre and Commercial 1: 6 storeys; Mixed Use, Comprehensive Development and Priority Development: 4; Capital City and Docklands: 12) and counted apart.
• Floor area already built: footprint × storeys from Overture's heights. A building with no height counts as one storey, which overstates what's unbuilt. In Monash, 81% of buildings have a height.

Capacity is what the current rules would allow, not what will happen, and it doesn't read overlay schedules across an area, as Fill and the Check do on a site.

Checked against Fill on 105 real lots across eight zones: the rule gives 886 homes, Fill places 854. Lot by lot the middle ratio is 1.00, half of lots fall between 0.76 and 1.50, and 61% are within a factor of 1.5. So read it added up, not lot by lot. On 12 small lots Fill placed nothing where the rule says more than one, as a lot's frontage can rule it out. Fill itself is held back on large lots by a development type's usual number of buildings, so it's a floor there.

Capacity isn't supply. Most lots that could redevelop won't, in any period. So allocation works with capacity × uptake, which you can change. It starts at 45%: ABS building approvals from July 2021 to August 2026, held for 25 years, as a share of capacity, the middle of 25 established metropolitan councils. They range from 9% to 156%, so treat it as a starting figure. Over 100% means homes are being approved faster than the rule says the land holds, as in Yarra, Merri-bek and Moonee Valley, so the rule understates inner areas, where apartments go on smaller lots and higher than it assumes. It's untested outside metropolitan Melbourne.

Allocation

Choose a total: the council's Plan for Victoria target to 2051, Victoria in Future's growth to 2036, or your own. A target and a projection answer different questions over different years.

Homes go to the most suitable lots first, each up to its capacity × uptake, until the total is met. Never over capacity. What can't be placed is a shortfall, and that is a finding: at this uptake, the current zones hold that much of the total. Where a target splits greenfield and other land, each part goes on its own land.

You can exclude a lot or set its homes. The rest re-allocates, and what you set counts towards the total.

A lot's homes are its share at this uptake, so a lot may show 0.3 of a home. Read them added up across an area, not as a forecast for one lot.

What to do with it

• Totals by zone and by SA2.
• Make a scenario from the lots in view with homes allocated, to Fill, Check and analyse.
• Download by lot, as GeoJSON or CSV, in EPSG:4326.

Limits

• It isn't a housing capacity assessment under Planning Practice Note 90 or a substitute for one.
• It doesn't test market feasibility, infrastructure or who will redevelop when.
• Envision Tomorrow's own suitability method couldn't be confirmed from its documents, so this follows the greyfield method as AHDAP Envision implements it.
• Jobs aren't allocated: there's no open small-area jobs projection.

How well does it hold up?

Homes: estimated homes add up to between 101% and 103% of Victoria in Future's 2021 dwellings in all 31 metropolitan councils.

Against what got built: does a higher score go with more homes approved, measured in places that had no say in setting it? The weights weren't fitted to this. For 347 metropolitan SA2s: the mean score of their lots against ABS building approvals per existing home, July 2021 to August 2026. (A rank correlation runs from -1 to 1, with 0 being chance.)

What didn't hold up: the two factors about the lot itself ran the wrong way at this scale, in established areas. Share of permitted floor area unbuilt: -0.09. Lot size: -0.23. Taking both out leaves the result at 0.34. So this test gives no support to the unbuilt floor area stand-in. Walks and zones carry the result: zone 0.34, shops 0.33, centres 0.29, density 0.29, trams 0.20, stations 0.17.

How to read it:

• Approvals are by SA2, so they can't say which lot in a neighbourhood goes. That is what those two factors are for, and it remains untested.
• Past approvals reflect old zoning, and today's footprints already show some of what was built, which counts against the unbuilt factor.
• Passing doesn't mean the score says where growth should go, and failing wouldn't make it wrong.
• Land value ÷ capital improved value, the published measure, would need council rates data.

Weights: across all 31 councils, about 89% of the top tenth of lots stay there when each weight changes by up to half (lowest 83%).

Against targets: at 100% uptake, current zones hold each council's target in 20 of 31 metropolitan councils. The tightest are Yarra, Merri-bek, Moonee Valley and Darebin, where the target is more than twice the capacity. Together with the note on uptake above, that says more about the rule's limits in inner areas than about those places.

Sources