Kenneth Davidson claimed in The Age yesterday that Melbourne has 15 years’ supply of outer suburban land zoned for urban development at the world’s lowest residential densities of 12.5 to 15 houses per hectare.

Lowest in the world? I think that’s possibly a little harsh when Melbourne is compared with the outer suburbs of US cities. However what I’m really interested in looking at is what Melbourne’s supposed “lowest residential densities” actually look like. What does 15 dwellings per hectare mean on the ground?

An ideal case study is the new mixed use development planned for Toolern, near Melton. According to the Precinct Structure Plan, when fully developed it is expected to cover 24 sq km, house an estimated 55,000 residents and host businesses that provide 28,000 jobs.

This is an enormous project, covering an area around a fifth larger than the entire inner city municipality of Yarra. It is equivalent in area to a 2.8-kilometre radius circle – if the centre were Melbourne Town Hall, it would extend to Richmond Station in the east, Alexandra Parade in the north, Bolte Bridge in the west and Albert Park in the south.

The minimum average density set down for Toolern is 15 dwellings (per net developable hectare), the same as the target minimum for the growth areas set out in Melbourne @ 5 Million and its predecessor, Melbourne 2030.

This is a substantial improvement on the historical pattern and in fact is not that much different from older, inner suburban municipalities like Darebin, Moreland and Moonee Valley. For example, the gross density of Toolern is expected to be 2,300 persons per square kilometre, similar to Moonee Valley’s current 2,400 persons per square kilometre.

A maximum 54% of dwellings are to be conventional housing, with 46% at higher densities. Of the latter, 36% are to be medium density (25 dwellings per hectare) and 11% high density (35 dwellings per hectare).

The Structure Plan envisages a diverse range of housing. Dwelling types will include multi-storey apartments, terrace housing, apartments and studios above garages, semidetached housing, detached houses and mixed-use buildings (shop-top apartment and live/work units).

Even the conventional subdivisions in Toolern are likely to be a far cry from the quarter-acre block (1,000 sq m) stereotype. A leading property company reports that the median lot size in Melbourne’s growth areas fell eight percent over the last year and is now 513 sq m (parts of inner city Yarra, such as Alphington, have an average lot size in excess of 600 sq m).

It is heartening that the increase in density on the fringe isn’t primarily determined by regulation, but rather is driven organically by the market. Consumers are responding to rising prices by downsizing – mainly smaller lots, but dwelling sizes are also showing signs they may be stabilising.

Density is only one aspect of a “good” development – there are many other variables. The Precinct Structure Plan espouses the rhetoric expected of any new development in 2010, fringe or otherwise – walkability, permeability, transit-oriented development, employment self-containment, water sensitive urban design, accessible and vibrant mixed-use activity centres and greater housing choice, diversity and affordability.

Toolern does not mandate a minimum average density of 20 to 25 lots per hectare as proposed for Melbourne’s growth areas by Mr Davidson. But 15 dwellings per hectare is not bad. It is a reasonable compromise between the realities of the market in 2010 and the ambitions of strategic planning to improve Melbourne’s economic, social and environmental performance. It’s consistent with the market and it does not preclude the possibility of a higher average density if the market should move further in that direction over the development period.

I can’t finish without taking issue with Mr Davidson over his claim that a recent study by Curtin University put the upfront infrastructure outlay per dwelling for fringe developments at $136,000 compared with $50,500 a dwelling in the inner city.

I’m disappointed that the report he refers to is still being used to support these sorts of claims. As I pointed out long ago (here, as well as here, here and here) that report’s development cost estimates are based on an earlier unpublished paper that simply summarised 22 previous projects, some of which date back as far as 1972 and cover the USA and Canada as well as various locations in Australia.  There is no up-to-date estimate of the comparative costs of development across the various regions of Melbourne but the indications are that any “gap” has already disappeared. In any event the inner city is not a relevant comparison for the fringe as the sorts of households that will settle in Toolern cannot afford to live in the centre and probably don’t want to.