Housing Needs Assessment
City of Boroondara

The Victorian Government set a housing target of 65,500 additional dwellings for the City of Boroondara by 2051, published in the final Plan for Victoria on 28 February 2025. The statewide total is 2,239,700 dwellings across 79 local government areas. A draft figure of 67,000 circulated in June 2024 and was superseded. This assessment uses 65,500.

Section 01 states what the target requires. This section tests what is being delivered. ABS Building Approvals is the only published series that measures it, and it is available monthly at council level, currently to June 2026.

A target of 65,500 on a base of 62,715 cannot be met with detached houses on existing lots, and the most recent complete year is the one in which the council approved the fewest attached and apartment dwellings of the three. Whether that is a market cycle, a response to interest rates, or an effect of the planning controls in force is not something an approvals series can answer. What it does establish is that the trend is running against the target rather than towards it.

Estimated resident population is the ABS official population measure and is the appropriate base for a demand case. Census counts are a point in time enumeration and are not the same series. The figures to 2023 come from the ABS hosted series and 2024 and 2025 from the Regional population release of 31 March 2026. Both are ASGS Edition 3 and use the same SA2 codes.

The council is now 1,062 people below its 2019 peak, having been 10,031 below it in 2022. Compound annual growth from 2001 to 2019 was 0.78 per cent. Extended to 2025 it is 0.56 per cent. Neither is a forecast.

Where the growth is coming from

This matters for a housing strategy in a way a single growth figure does not. Growth built almost entirely on one component is exposed to changes in that component, and net overseas migration is set by Commonwealth policy rather than by anything a council or the state controls. A demand case that assumes the last two years continue is assuming a policy setting, not a demographic trend. Natural increase of 60 across a municipality of 178,601 people also says something about who lives here now, and it is consistent with the ageing visible in the median ages in Section 05.

Components are for 2024 to 2025 only. Recovery is the share of the 2019 to 2021 decline regained by 2025, derived from the ABS series and not published by ABS. Four SA2s now sit above their 2019 level: Hawthorn East, Balwyn North, Ashburton (Vic.) and Glen Iris - East.

The council held 62,715 occupied private dwellings at the 2021 Census: 34,771 separate houses, 12,501 semi detached, row or terrace dwellings and 15,162 flats or apartments. Read at council level that is a 55 to 24 split between detached and apartment stock. Read at SA2 level it is two different places.

Of 62,715 occupied private dwellings, 24,290 were owned outright, 18,643 owned with a mortgage and 18,036 rented. Home ownership at 68 per cent sits well above the Victorian pattern, and the spread across SA2s is wide: Balwyn North at 78 per cent against Hawthorn - South at 50 per cent.

Median household income across the council is $2,376 a week. The SA2 range runs from $2,017 in Balwyn to $2,914 in Glen Iris - East, a spread of $897 a week. Median rent moves in the opposite direction to income in several SA2s. Balwyn North records the highest median rent in the council at $554 while Hawthorn - South records the lowest at $400, which reflects dwelling size and structure rather than relative affordability.

The Planning and Environment Act 1987 carries an objective to facilitate the provision of affordable housing, and section 3AA defines affordable housing by reference to very low, low and moderate income households. A Governor in Council Order published in the Government Gazette sets those income ranges and is remade every year. This assessment uses the Order made in 2026 and the Greater Melbourne table, which is the table that applies to Boroondara.

Annual gross household income. The final column applies the thirty per cent benchmark to the ceiling of each band and converts to a weekly rent.

Census income bands and gazetted thresholds do not share boundaries, so every count below is taken at the Census band edge nearest to and below the statutory ceiling. That is a whole band count, never an interpolation inside a band, and the cut used is stated beside each figure. It makes the numbers here conservative by construction.

Income distributions describe who lives in a place. They do not say whether that place has anything for them to rent. This section tests the rental stock directly: for each statutory band, how many of the council's 18,036 rented dwellings were let at or below the rent that band can afford.

For context, the council's entire social housing stock is 602 dwellings, so the great majority of what is affordable to the lowest income bands is private rental rather than social housing, and carries none of the security of tenure that social housing carries.

Two published counts and one disclosed parameter produce the number a housing strategy actually needs. The benchmark used here is the $1,488 a week ceiling, which is the low income ceiling for a couple and the moderate income ceiling for a single adult under the Order. Rent affordable at that income is $446 a week.

Applying the same income distribution and the same affordability rate to the 65,500 dwellings in the target adds a forward shortfall of 3,723 to 6,473 households. New dwellings generate new lower income households, and at the council's current rate only 18.8 per cent of new rentals would be affordable to them.

Where the shortfall actually sits

This reverses the intuition that the apartment SA2s should carry more. On this measure they are already carrying the council's affordable rental supply, and have been doing so while absorbing most of its growth. The SA2s where an affordable housing requirement would change the distribution are the ones that have so far avoided both.

The Census records dwellings by landlord type. Across the council 408 occupied private dwellings were rented from a state or territory housing authority and 194 from a community housing provider, 602 in total or 0.96 per cent of occupied stock.

The evidence above does not produce a strategy. It narrows the questions a strategy has to answer, and it makes each answer defensible against a published source. Four decisions follow directly from the figures.

A gap stated is more useful than a gap filled. The following are absent because a published government source was not retrieved for them, not because they are unimportant.

Every figure in this assessment carries a tag. Measured means the value came directly from a published dataset. Derived means it was calculated from measured values by the arithmetic stated. Disclosed means a parameter was set rather than retrieved.