Relative Cost of Government Infrastructure


The Australian population is projected to double in the next 4-5 decades. Major cities are projected to double in size, Melbourne and Sydney to 8m. This will require considerably more government infrastructure spending to maintain and increase liveability or quality of life and prosperity as the population grows. It is worthwhile developing preliminary views on the relative cost of this infrastructure and of housing for the projected extra 24m people. Some 9-10m dwellings will be needed to house 24m people over the 4-5 decades and possibly more if population continues to over 50m later.


There are three main options or choices for population distribution in the major cities of Melbourne, Sydney and Brisbane:

  1. high densification of old, low density, established inner-city areas
  2. new greenfield development of city fringes
  3. new greenfield cities of 1m in regions on High Speed Rail (HSR)

The question is where to place the emphasis, as 1 and 2 will both occur to some extent anyway. Option 3 is not authorised at present, but should be considered here as it would relieve the cities from growing by +4m to +8-9m, if 2 or more new cities of 1m were built in the regions on the HSR line around the major cities and connected to CBDs in less than an hour commuting time.

Government infrastructure

Infrastructure required for growing populations to 8m is mainly for water and sewage, gas, electricity, telecommunications, roads, public transport, hospitals and schools. Each is considered for each option.

Water and sewage

This infrastructure already exists in the inner-city suburbs of a major city. Some presently installed surplus capacity may be available to some of the 3m increase in this area of a city. While not enough, it would save some cost. Much of the infrastructure may be old and need replacement to serve the increase, so savings would be less. Building new replacement infrastructure in built-up areas is expensive and disruptive. The total cost of water and sewage may be somewhat less than in greenfield fringes. Supply of water may cost the same as greenfield, if its source is desalination.

Fringe greenfield development for 1m more requires the cost of new infrastructure. It is easier, cheaper and less disruptive to install, if before house building commences. It is difficult to graft-on separate potable, grey water and storm-water retaining system to the old inner-city single water system.

Two-three new cities around a major city, saving it 2-3m increase or more, are like fringe greenfield development with some advantages. They can be designed holistically from scratch to be smart, environmentally friendly, sustainable and almost self-sufficient. Three separated water systems can be installed from the beginning. The regional cities are generally closer to lower cost new dam water supplies, if available, and may not need expensive desalination. If they do, they are near the sea, which keeps down high pipeline costs. Less water is required for newly designed cities than inner and fringe city areas. The cost of water and sewage may be much less per head than inner-cities and city fringes.


Some gas mains may have to be enlarged, old pipes renewed and pressure increased for population growth in inner-cities. The cost would be less than new reticulation in fringe areas. New cities may be like the fringes, depending on proximity to the location of the gas source.


In Victoria, local off-shore supply of gas is plentiful for the present. About half of production is exported interstate. Supply from Bass Strait is projected to decline in the long term, so there may not be enough for doubling of the population. Gas may have to be imported from interstate. Fortunately, Exon has found a new large scale, deep, on-shore, gas field (not fracking) in Gippsland. If proved, Victoria would have sufficient local supply.

A new city on HSR in Gippsland would be especially well placed on top of the gas field. A large quantity of good quality water would be associated with this gas production. It would supply the new city, with plenty of water left over for irrigation, intensive horticulture and industry for the enlarged population. This is a bonus for Gippsland, which may not apply to other new cities on HSR.


Electricity distribution has been renewed extensively in major cities in recent times with higher capacity. It may need some enlargement for doubling the population, but maybe not the full cost of doubling the network. Supply would need to grow with population independent of population location. Larger cities are more efficient users. Fringes and new cities would be high cost to install. Electricity would need to be lower priced. There would be a plentiful supply of gas for base-load electricity in the long term to back-up renewable energy where necessary in Victoria and probably in NSW and Queensland. Gas will be piped from the Northern Territory. Most rooves of new buildings would have solar panels.


Inner-city roads are in place. There may be some extra cost in widening. Certainly, extra cost will be needed for additional freeways and associated demolition to serve autonomous electric vehicles and endeavour to contain exponential growth of congestion.

Fringes will need extensive new roads as they develop and for many autonomous vehicles which will proliferate because new traditional public transport will be less available in these areas as they grow.

New cities will need roads, but the area of the cities’ footprint would be less than fringes. Public transport in the form of trams to the HSR station and city centre would be provided. Trams would lessen the need for autonomous vehicles, reduce congestion and simplify the road system thereby reducing its cost. New city roads may cost as much than inner-cities, but less than city fringes. The new city roads would be more effective than fringes.

Public transport

Inner-city public transport systems are quite extensive, but not anywhere near as comprehensive as in London with its present population of 8m in a smaller area of high densification than each of Australia’s three major cities and its world rank for liveability is a low 53. It is unlikely that Australian cities’ public transport will ever equal London’s. The 20-minute areas of London are a consequence of densification.

The authorities here seem to rely on the introduction of autonomous electric vehicles to provide the necessary transport for 8m people. These can be increased incrementally and would probably be private investments, not a government expense. Their effectiveness is suspect, as they would be regarded as ‘public transport vehicles’, not privately-owned cars, and not be cared-for so well. They will be on the move 24/7 on freeways and highways, as they will not be parked or garaged. Congestion will be great. Three new concentric freeways around CBDs will be required to facilitate rapid pick-up. The extra cost to government of freeways may be passed on to the vehicles as road-use taxes.

Autonomous vehicles may become the main public transport in the fringes. They will have longer, more expensive trips to CBDs and require more freeways for quick access. Idle vehicles will add to congestion in inner-city and fringes as they continuously circulate. A margin will always be idle ready for pick-up.


New cities in the regions would have a public transport advantage with HSR over fringe areas. HSR would be quicker than from the fringes to the CBDs and less expensive for commuters. HSR would be a private enterprise project. HSR would be self-funding within the construction period through sale of dwellings built above tracks in trenches, not financed from higher fares over 40 years of operations. HSR fares would be low because it is self-funded. Also, governments would subsidise them from a small increase of city land tax for the unearned benefit of sustained liveability because of lower city populations as more people live in the regions on HSR lines. This would encourage new city development and commuters. Cost of commuting from new cities would be less than from fringes. It would be free from the irritation of congestion in inner-cities and fringes. Housing would be cheaper.


These are a Federal cost. NBN is updating the service and should be able to extend it readily. Present dense inner-city areas would be lower cost. New densification would replace old housing to be more easily accessible to NBN use.

Dispersed new fringe areas would be lower density and higher cost. New cities would be medium density. They would cost more than inner-cities and less than fringes for the same number of services.


New hospitals are costly to build, particularly so in inner-cities on much more expensive, cleared land. Congestion increases difficult of access, though there may be more public transport in well located sites.

Hospitals in the fringes would need to be many and smaller to serve the more dispersed population. Land and building cost would be high. Congestion and poor public transport would curtail ready access.

New regional hospitals would be lower cost to build and land is cheaper. Access would be good because of less congestion and more public transport to the central location of hospitals in smaller cities.


Inner-city schools are at a disadvantage in Melbourne as many were closed and sold-off after the Cain/Kerner era. New schools for population increase in built-up areas of cities would have to be high-rise with very limited school yards and playing fields. These would be high cost and low effectiveness.

Fringe areas have more land for schools, but accessibility by traditional public transport is limited. The cost would be less than inner, but their effectiveness would be constrained.

New city schools would have plenty of cheap land. They would have public transport access by trams. They would cost less than inner-city and fringes, and be more effective and convenient than both.


Although housing is not a large government cost, it is central to peoples’ interests. Densification in inner-cities would be higher cost than elsewhere because of the high cost of land and the need for demolition of the single dwellings on ¼ acre blocks and their replacement by 4 dwellings as planned by government. The luxury of the largest average sized houses in the world would be destroyed.

Fringe dwellings tend to be larger, but are limited to ⅛ acre of land, which is still expensive in the city. Densification will increase in the fringes. Open green spaces would be less high cost than inner-cities.

New cities would average medium density spread between high-rise/high density, medium density and low density further from the HSR station. Land is much cheaper than inner-cities and fringes. Construction costs are lower than in the major cities. Total housing costs would be much lower. There is not need for demolition as in inner-cities. More open green spaces would be lowest cost on cheap land.



Table 1 below gives a summary of relative government infrastructure costs and their effectiveness in inner-cities, city fringes and new regional cities on the HSR line. It indicates that well designed new cities on HSR should be adopted for lower government costs and for greater effectiveness of spending for tax payers and for users. Smart digital technology cannot substitute for loss of spatial liveability in the cities. Densification of dwellings causes densification of traffic.

  Inner-city City fringe New cities Priority
Water/sewage 2** 3** 1*** New cities
Gas 2*** 3*** 3*** Inner-city
Electricity 2*** 3*** 3*** Inner-city
Roads 2* 3** 2*** New cities
Public Transport 2* 3* 1*** New cities
Telecommunications 1** 3** 2** Inner-city
Hospitals 3** 3* 1*** New cities
Schools 3* 2** 1*** New cities
Housing 3* 3** 1*** New cities
Overall 2* 3** 1*** New cities

Table1                        Summary of Relative Cost of Government Infrastructure Costs

Notes:   1 Low cost; 2 Medium cost; 3 High cost

                * Low effectiveness; ** Medium effectiveness; *** High effectiveness

These are projected relative government infrastructure costs in today’s dollars. There are other real costs attributable to major cities and location of population increase that are not discussed above. The serious unintended consequences of densification are three-fold: there is a high cost of continual disruption, demolition and displacement of people; secondly, there is the great cost of loss of liveability as dwellings are downsized, even the average size halved, as evidenced in London, and the high cost of exponentially increasing congestion; and thirdly, there is the danger of growing reluctance to travel into CBDs because of the increasing cost of commuting, congestion and unproductive time lost, and the lack of connectivity of dispersed centres of innovation, which is a drag on innovation, as evidenced in China.

These great indirect costs over-arch, over-ride and determine consideration of population location and therefore infrastructure costs. There is a choice. It is not necessary to densify the major cities and lose liveability and innovation. These disturbing, huge, incalculable costs can be avoided by building new, lower cost, highly effective cities on HSR lines around major cities and distributing population increase into the regions within easy reach of city CBDs by public transport.

The new regional cities would be much healthier living with many walking and cycling opportunities, nearby country and sea, and clean air with less smog than major cities. They would be less stressful with little local congestion, a quick commute to jobs in the CBDs and high housing affordability.

The new cities and regional towns on HSR would be flexible in absorbing any extra population increase in later years should it occur, at little additional indirect cost, where 8m cities would not be flexible.     For further information on HSR, densification and new cities, please visit


Back To Top