Parkland, Melbourne
Public open space, Melbourne

There was a commentary by Fairfax journalist Michael Short in Saturday’s paper arguing we should make our cities greener (Urban jungle: radical ways to improve our city life).

As higher-density housing options are becoming more common, even on the fringes, we need to plan for community green spaces for both recreation and the growing of food.

The article cites examples of how other cities have retro-fitted more green space e.g. Paris’s Promenade Plantee and Manhattan’s High Line.

Few would argue with the desirability of making our cities greener. The article has some modest but worthwhile ideas, like encouraging “pavement parks” and “mini acacia” parks. (1)

It also has some bigger ideas that are more questionable. The most ambitious proposal is to deck over existing motorways to create new parkland as per the plan to cover 3.7 km of Hamburg’s A7 autobahn.

The proposed opportunity is Melbourne’s Tullamarine freeway. Sounds wonderful but would it make sense?

The cost would be astronomical. Consider that decking over Federation Square East is estimated to cost around $15,000 per sq. m.

The Tulla is 13 km long, but suppose only the innermost section, from Flemington Rd to Brunswick Rd, were decked. It’s 1.4 km and is about to be widened to take five lanes in each direction.

Assuming the decked area is 50 metres wide (it also has to cover emergency lanes, entry/exit lanes, and the central median), it would cost circa $1 billion to provide a seven-hectare platform. As a comparison, Melbourne’s Albert Park is 225 Ha i.e. 32 times larger.

It could cost more though because the beams necessary to span five traffic lanes would be very long and expensive compared to those required to cross the rail lines at Federation Square East.

Then there’s the additional cost of providing access, amenities, and soil. It probably wouldn’t be feasible to have big trees and some of the parkland would be sterilised by ventilation stacks.

The bigger issue though is that the case for providing additional open space in Australian cities is much less compelling than it is in places like Paris or Manhattan.

For one thing, we have a lot more private space. Only 16% of the housing stock in the Melbourne statistical urban area is apartments compared, for example, to circa 80% in metropolitan Paris. (2)

The rest is made up of dwellings with ground level outdoor open space i.e. detached, terrace and town houses. Arguably Australian apartments have more balcony space too.

Another point is that we already have much more parkland in absolute terms than either of those places, let alone when account is taken of their vastly higher population density (see exhibit).

There are some questions about how equitably parkland is distributed across Melbourne and, at least by our high standards, there are some locations that are under-supplied.

But by and large the key issue for us isn’t the quantity of available parkland; it’s the quality of public spaces more generally, including parkland. I think we need to focus on ideas like making existing public spaces better rather than spending a lot to provide more parks. (3)

There are two straightforward actions we could take. First, increase the number of trees in city streets, as I’ve discussed in detail before (see Why don’t we green the streets of Australia?).

Second, improve the design of existing parks and public spaces. Melbourne City Council’s excellent Urban Forest Strategy is a good example of what might be done: it proposes increasing public realm tree canopy cover within the municipality from 22% at present to 40% by 2040.

Michael Short provides another example; he’s written separately about a proposal to turn Melbourne’s Elizabeth St into a rainforest canal. It probably wouldn’t work in that location but the general idea is good.

The concept of spending huge amounts to retro-fit more parks in Australia’s low density cities – much less spending billions of dollars to create a deck over a freeway – is hard to justify when we have so many other pressing demands on public funds.

It makes more sense in Paris and Manhattan; yet they’re nevertheless highly sought-after locations despite their relative lack of open space. It seems they have other amenities – like their streets – that residents and visitors obviously value. (4) (5)


  1. The contention that we “need” to grow food in urban areas is dubious; see Should urban parkland be used for farming?
  2. Even in the municipality that takes in the bottom end of the Tulla, apartments only comprise 21% of the dwelling stock. In central Paris they’re 100%.
  3. How much value we should put on parkland – and hence how much we should provide – is an issue that warrants a lot more debate; at present we rely mostly on rules of thumb of uncertain provenance.
  4. Infrastructure like the High Line and the Promenade Plantee (the original partly elevated linear park) didn’t result solely from a pressing demand for more open space; they were in large part the result of opportunistic circumstance i.e. the availability of disused and publicly owned infrastructure. If Sydneysider’s were persuaded the Cahill Expressway had lost its usefulness I expect they’d love the idea of converting it into a magnificent harbourside pedestrian promenade.
  5. I wonder how attractive the High Line is as a recreational asset for residents given how intensively it’s used by visitors?