Monash University Publishing | Contacts Page
Monash University Publishing: Advancing knowledge

Melbourne 2030: Planning Rhetoric Versus Urban Reality



Bob Birrell, Kevin O’Connor, Virginia Rapson and Ernest Healy

Infill already provides more than one third of the new housing in Melbourne. Contrary to the claims of the Melbourne 2030 planners, the scale of infill is likely to increase rather than decline. This is because the regulations governing infill (ResCode) do not prevent its spread (as was originally promised by the Bracks Government). A few municipalities have taken the statutory opportunity to tighten the infill regulations but, even where this has occurred, infill continues. Demand for infill, whether as detached houses or semi-detached dwellings, will be high, especially from households unable to afford a separate house in middle and inner suburbia. The chapter analyses the impact that infill is having on the established traditional garden style of Melbourne. Infill is destructive of the existing suburban streetscape, both of its tree and shrub canopy and its low density built heritage. In its place it leaves a more congested environment dominated by obtrusive dwellings.

Melbourne suburban dwelling style is distinctive for its emphasis on low density houses in garden settings which can often be traced back to the original plantings. While the attachment to canopy trees and shrubs can be linked to British ideals, there is an element of adaptation to the local climate. It is characteristic of Australian suburban landscaping – at least until recently – that the house design should be low slung (as is characteristic of the bungalow) and set unobtrusively within the canopy trees that feature in most gardens. These trees and shrubs are located not just in the front of the house (facilitated by generous setbacks of 12 metres or more), but also to the sides and in the backyards of each house. The result is that an observer looking down a typical street in a suburb more than ten years old will see trees and shrubs rather than roofs or walls.

The suburban ambience of Melbourne is central to most people’s sense of place. They are the custodians and the contributors (through their gardens) to its continued existence. It is intrinsically attractive. It provides privacy, a secure site for raising children and it provides a source of recreation – gardening. Most residents like it this way. When politicians talk about Melbourne as great place to live, there is no doubt that most residents of Melbourne have the above attributes in mind. In reality, this ‘liveability’ is under threat from infill. According to the planning rhetoric, Melbourne 2030 will help to preserve existing suburbia by concentrating development in activity centres. However, as we have shown, this is unlikely to be the case. As this chapter details, the Bracks Government has not acted to arrest the scale of infill, as it implied it would when it first came to office. Thus an unintended consequence of Melbourne 2030’s containment strategy is that it will promote more infill.


It is currently estimated that the contribution of infill to Melbourne’s total new dwelling construction is around 35 per cent (Department of Sustainability and Environment 2004 p. 17). This figure does not include the new apartments in development precincts like Docklands. Furthermore, there has been no decrease in this share since the days of the Kennett Government. There has been remarkably little analysis or even discussion of the infill phenomenon in recent years. Yet it is obviously of enormous significance in Melbourne’s housing market. For this reason the chapter begins with an analysis of the scale, location and market dynamics of infill before examining the Bracks Government’s ResCode regulations and their impact. The final section details why infill is such a threat to Melbourne’s suburban ambience.


As Figure 5.1 shows, the construction of medium and high density dwellings in Melbourne took off in the Kennett Liberal Government era of 1992 to1999. The total number of dwelling approvals grew dramatically during the 1990s, from 14,326 in 1990–91 to 39,240 by 1999–2000. The proportion of these which were classified ‘other residential’ dwellings also grew from 12 per cent to more than 30 per cent during the decade. That level has been maintained since the Bracks Government came to power in 1999. Other residential dwellings refer to all those dwellings which are not a detached house. They include semi-detached townhouses, terrace and row houses, as well as flats and apartments.

Figure 5.1 Number of dwellings approved and percentage of these which were other residential dwellings, Melbourne, 1990–91 to 2003–04
Australian Bureau of Statistics, Building Approvals by Capital City Subdivision (includes new dwellings and conversions), downloaded from Ausstats

These other residential dwellings were located across established suburban zones, as shown in Table 5.1. The total number expanded from 7,183 in 1996–97 to 12,651 in 1999–2000, by which time almost ten per cent of other residential construction was attributable to the Docklands project. By 1999–2000, 47 per cent of other residential dwelling approvals were for Inner Melbourne. Nevertheless, as shown in the table, there was a significant increase in the number of building approvals for other residential dwellings across the inner and middle areas of Melbourne, with activity more or less evenly shared amongst suburban areas.

Table 5.1 Building approvals for other residential dwellings by region, number and per cent of Melbourne, 1996–97 to 2003–04 See table in technical note at the end of chapter for the composition of the regions.Australian Bureau of Statistics, Building Approvals (data include conversions and refurbishments)

In the discussion in this chapter, infill refers to two or more new medium-density dwellings constructed on sites formerly occupied by detached houses or on vacant lots in established areas where the lots were originally intended for detached houses. It is not limited to the Australian Bureau of Statistics’ other residential dwelling category depicted in Figure 5.1 and Table 5.1. This category significantly understates the scale of infill because it does not include separate houses constructed as part of dual occupancy on a conventional detached housing site. If the dwellings constructed on these sites do not have adjoining walls, they are classified as detached houses and are included as such in Figure 5.1. An estimation of the scale of this undercount for Perth suggested that between 31 and 32 per cent of recent dwelling approvals were for medium and high density dwellings rather than the 20 per cent reported as other residential dwellings. One Melbourne study estimated that, in the case of Boroondara, the proportion of dwelling approvals which are medium density increased from around 50 per cent to nearly 70 per cent when the broader definition is used. For a discussion of this issue, see the technical note at the end of this chapter.



The controls over infill have changed during the 1990s. They began with the document Victorian Code for Residential Development – Subdivision and Single Dwellings (otherwise known as VicCode1), which dates to the last days of the Kirner Labor Government in 1992 (Lewis 1999 p. 128). VicCode1 provided for smaller lot sizes, zero lot line buildings (where the dwelling wall is built on the boundary) , footpaths down one side of the street only and other measures to reduce infrastructure costs and assist with urban consolidation. VicCode1 was aimed at making housing more affordable to the less well off and removing the ‘stultifying effects on dwelling design’ of the previous regulations (Eccles and Bryant 1999 pp. 28–29).

VicCode1 overrode the Victorian Building Regulations 1983 (VBRs) which had imposed a standard set of minimum siting controls on earlier subdivisions. Each municipal council had been required to make a local law which specified which set of controls they had chosen from this standard set. Different controls could apply to different parts of the municipality and the controls could be in excess of the minimum required in the VBRs. Many of the councils had adopted the ‘quarter acre’ block and minimum frontages of nine metres (30 feet) and side setbacks of 1.5 metres for all or most of their municipalities (Eccles and Bryant 1999 p. 29).

The Victorian Code for Residential Development – Multi Dwellings (known as VicCode2) was incorporated in 1993 but was subsequently replaced by The Good Design Guide for Medium Density Housing in 1995. These provisions affected all municipalities. They established the rules for medium-density dwelling proposals. (See Lewis 1999 p. xx, and Buxton and Tieman 2004 pp. 3–7, and Eccles and Bryant 1999 pp. 53–55 for further detail.)

VicCode2 and The Good Design Guide ‘had the effect of overriding municipal codes and policies in relation to medium density developments...’ (Eccles and Bryant 1999 p. 54). The Good Design Guide allowed densities greater than 1:200m2 within seven kilometres of the GPO (which includes, for example, the inner areas of Stonnington, Boroondara, Moreland and Moonee Valley), and elsewhere on sites larger than 2000m2. The Good Design Guide also provided a table of street setbacks (the distance between the street boundary and the wall-face of the dwelling – eaves, porches and verandahs could project into the setback). Where the adjacent dwellings were set back nine metres (the minimum setback for dwellings in the middle and outer suburbs under the VBRs in the 1980s), minimum frontages of only six metres were now permissible. New dwellings fronting the long side of a corner site of less than 1200m2 could have setbacks of just three metres. In all areas beyond the seven kilometre radius, site coverage could be up to 60 per cent (Department of Infrastructure 1997).

These codes created a permissive environment which left residents and municipal councils interested in controlling infill with little room to object to dual occupancy or other forms of medium density proposals which involved the redevelopment of existing suburban sites. Even buildings allegedly protected by heritage overlays were vulnerable because of the permissive way in which demolition permits could be issued (as a result of the Kennett Government privatisation of the permit process) (Lewis 1999 pp. 165-170).


Infill projects cannot occur in the absence of demand for the product, or in the absence of suppliers (investors, builders and sometimes owners of the dwellings in question) prepared to put their money on the line. On the demand side, as Chapter Four makes clear, in the 1990s there was a large cohort of young people aged 20–30 in the housing market place. They were the main group which took up the flats and apartments constructed at the time. But even if they had wanted to purchase a detached house, the prospect of doing so, especially if they wanted a house located in the inner suburbs or in the more affluent locations in the middle-eastern and southern suburbs, receded with the housing price boom beginning in the mid-1990s. According to one study, by 1992 a ‘representative’ home buyer (in terms of family income) would have been able to afford a house in a suburb adjacent to the South Eastern (Monash) freeway some 22 kilometres from the city. From the mid-1990s this prospect receded further as home values escalated, particularly in the suburbs in question (Productivity Commission 2004 p. 36).

The consequence is that for many of those wishing to purchase a dwelling, particularly a new dwelling, in existing suburban areas, infill (often flowing from a dual occupancy development) became the most feasible option.

On the supply side, the phenomenon of the small property investor has been much discussed. There has been a quite extraordinary engagement of Australian taxpayers in residential investment where rental income is involved. As has been widely canvassed, this investment has been propelled by financial deregulation and thus the ready availability of loan funds, low interest rates, and favourable tax treatment of property investors (negative gearing and low capital gains tax). These investors are a major source of the supply of other residential dwellings, as well as houses constructed under dual occupancy provisions.

In addition there are the investors looking to subdivide and sell their product into the marketplace. These include owners who invest in developing a dual occupancy dwelling on their own property. Unfortunately there is no definitive study of this phenomenon. Our discussions with municipal officers and others acquainted with the grass roots of the residential infill industry indicate that there are large numbers of builders and investors who purchase existing detached houses with the intention of redeveloping the property, either by replacing the existing house with two new dwellings or by renovation of the existing house and the addition of a new dwelling in the backyard. The initiators of infill range from the current home owner, to small builders who make a living from this industry, to developers (who may not be builders) who undertake several such projects in a year. These players can be described as opportunistic in the sense that the location of their investment depends on where properties come up for sale and their judgement of whether the market price for the detached house in question will allow a profitable redevelopment. The tag ‘opportunistic infill’ derives from its apparently unsystematic nature and the fact that is does not derive from any planned process such as containment near public transport routes. Illustrations of the phenomenon are shown for parts of Monash and Greater Dandenong in Figures 5.4 and 5.5 below.


In Opposition, the current Labor Government had taken note of the protests surrounding residential infill in established suburbs. It indicated that ‘[i]nstead of allowing medium density dwellings to be built anywhere in residential zones, it would be directed to the “right places”’. After the Bracks Labor Government took office, the Planning Minister, John Thwaites, made it clear that, ‘the right places were around public transport and activity centres’ (Buxton and Tieman 2004 p. 7).

The Labor Government changed the rules governing infill by establishing a new planning code entitled ResCode, which came into effect on the 24 August 2001. The Government claimed that the key aspect of ResCode was ‘respect for neighbourhood character’ and that ‘the standards to achieve these goals have been significantly strengthened compared with The Good Design Guide and VicCode1’ (Department of Infrastructure, citation date 2005).

ResCode specifies that neighbourhood character should be taken into account by municipalities when assessing permit applications and that the proposed building should ‘respond’ to the neighbourhood character in which it is located. To this end those seeking a planning permit for infill must first prepare a neighbourhood and site description. If infill is already prominent within the neighbourhood, there will be no grounds to object to further development. However, if this is not the case, the proponent must ensure that the dwellings are consistent with neighbourhood character. One might imagine that a proposal to insert medium-density development in a street which is overwhelmingly composed of detached houses of a low-slung bungalow style, most with gardens which feature canopy trees and shrubs, would be regarded as changing the neighbourhood character. However, in the world of the state planning fraternity, this is not the case. As long as the new medium-density dwellings (or extra house in the case of dual occupancy) meet the minimum setbacks and standards for site occupancy required in ResCode, and the building style is consistent with that of the nearby houses, the development will usually be approved by the municipality, or by VTAC if there is an appeal.

Some middle and inner suburban councils have added an extra hurdle for developers in the form of heritage overlays. The locations of these overlays for Boroondara and Stonnington are shown in Figure 5.2. Municipalities like Monash and Dandenong, which do not feature housing constructed in the nineteenth or early twentieth centuries, make little use of heritage overlays.

In Boroondara and Stonnington, even though the respective councils have made an effort with heritage overlays, the majority of the existing housing is still vulnerable to redevelopment. Location in an area affected by a heritage overlay does not ensure preservation. Rather the bar is just raised a little higher as regards the requirement for the would-be developer to construct new housing in harmony with the existing housing stock.

In Monash, a location that prides itself on its leafy ‘garden city’ streetscapes, the City has instituted a vegetative overlay which purports to preserve established canopy trees in its residential rules. It has also instituted other measures to do with site layout (discussed below), partly because of its concern to maintain the city’s leafy ambience.

ResCode has added some protective measures in terms of minimum setbacks. New dwellings adjacent to established dwellings with nine metre setbacks can no longer be built six metres from the boundary as they could be under The Good Design Guide (although porches, pergolas and verandahs less than 3.6 metre high may encroach up to 2.5 metres into setbacks). The standard for front setbacks under ResCode is the average of the setbacks of the two abutting dwellings, or nine metres, whichever is the lesser. Whether this tightening is sufficient to facilitate the planting of canopy trees and shrubs is an important issue which is explored later in this chapter.

Figure 5.2 Heritage overlays incorporated in the planning schemes of the Cities of Boroondara, Stonnington, Whitehorse, Monash and Greater
Dandenong, 2003
Victorian Planning Scheme Data

Municipal planning schemes can, however, add their own requirements. To date, most have not. The City of Monash is one of the few exceptions. The city has set a minimum front setback of 7.6 metres. This minimum acts to stop dwellings being built with a six metre frontage (as they could still be under the ResCode standard based on the average distance of the abutting dwellings if these dwellings were built between 1995 and 2001 when the six metre minimum was in force). In addition, the city has increased the private open space requirement from the standard 40 square metres of ResCode to the Monash standard of 75 square metres. Monash also requires that at least 35 square metres of the 75 square metres be at the rear or side of the building and that this 35 square metres space should be at least five metres in width. This means that the minimum area required is five metres by seven metres – the size of a generous sized double garage. This compares with the ResCode standard of 25 square metres with a width of at least three metres which translates to a minimum area of three metres by a bit more than eight metres – enough room to park a car and a trailer.

A few other municipalities have also added local amendments. In Bayside the minimum front setback is the greater of the setbacks of the existing buildings on the abutting allotments or nine metres, whichever is the lesser. Side setback requirements are double those specified by ResCode and rear setbacks are double those specified in ResCode plus an extra metre.1 The allowable site coverage in Bayside is 50 per cent rather than the 60 per cent specified in ResCode. In Glen Eira, the minimum site coverage for much of the city has also been designated at 50 per cent and the minimum area of private open space required has been put at 60 square metres.

These policy shifts have not prevented further residential infill, as Table 5.1 confirms. Apart from the local examples given, ResCode lacks teeth and its rhetoric about preserving neighbourhoods is hollow. As Buxton and Tieman (2004 p. 7) conclude,

... [ResCode] is a powerful statutory tool favouring dispersed instead of concentrated medium density development and does not promote better integration between mixed use development around activity centres and public transport.


The precise scale of infill in Melbourne’s housing market remains something of a mystery, partly because of the definitional problem discussed above. As indicated, the Urban Development Program (UDP) 2003 report estimates that infill constitutes around 35 per cent of the total new dwelling stock (Department of Sustainability and Environment 2003). The authority does not indicate how this percentage was calculated, or how it overcame the statistical problems cited in the technical note (at the end of this chapter) on the estimation of infill. The 2004 UDP report acknowledges that it still does not have the capacity to identify infill activity (Department of Sustainability and Environment 2004 p. xi). Nor do the 2003 or 2004 reports make any further comment on the prospects of such infill for the future. As for the Melbourne 2030 planning documents, they assert that the proportion of new dwellings constructed as infill in established suburbia will decline to around 30 per cent in the decade 2006 to 2015 and subsequently drop to 24 to 25 per cent in the decade 2021 to 2030 (see Table 2.1).

The prospects for infill have to be considered in the context of Melbourne’s demographic outlook. It was indicated in Chapter Four that most of the growth in Melbourne’s households will be amongst those where the household head will be aged 55 plus. Most will also be couple-without-children or lone-person households. As such they are considered by the Melbourne 2030 planners to be prime prospects for movement into activity centre apartments.

As argued in Chapter Four, most of the moves of older households that are likely to occur in Melbourne will be to dwellings built as infill. The same point applies to the new generation of prospective home owners currently aged in their twenties and early thirties. As noted, there is currently a large cohort of persons in their early thirties. These are the baby boomers’ children who were so important in providing an impetus to the apartment market in the 1990s. They are now in the partnering and family building phase of their lives where, if past experience is a guide, they will be looking to move from apartment living to a larger suburban home where they can raise a family. If they cannot afford a house, a good quality infill dwelling will be an attractive alternative. The same transition will occur with the smaller five-year age groups currently aged 25–29 and 20–24, though at a later point in the ensuing thirty years. Their numbers, as with the young cohorts to follow, will all be much smaller than the current 30–34 year old age group.

These broad demographic trends outlined above and the supply situation will favour infill though they will play out in different ways across Melbourne.


In the future, younger households looking to emulate their parents’ housing choices face a daunting prospect. They will find an inner and middle suburban housing market marked by relatively few sellers – given the baby boomers alleged propensity to sit. This characteristic will help keep the detached-housing market in these locations tight and expensive. Few of the younger generation, even if professionally qualified, will be able to afford the current $700,000 plus price (or the future prices when adjusted for inflation) required to purchase a detached house in many locations within this area. The effect will be to create a strong market for an alternative dwelling type, either in infill sites or activity centres (which in inner areas will be mainly high density buildings). It is hard to see how an apartment in such a centre can compete with infill, given that the price of the former could be greater than the latter. In addition, an infill dwelling will have at least some backyard space and some opportunity to create a garden, even if on a much smaller scale than that typical of their parents’ homes.

Under current planning regulations, there is likely to be a continued supply of infill dwellings. Even with detached properties selling to $700,000 and $800,000, an infill investment is likely to be profitable. This pattern of ownership is likely to become more common over the next few decades as those seeking to reside in a suburb like Camberwell recognise that a townhouse or an apartment is the best affordable long-term option.

Hence demand and supply of infill residential construction in the inner city will remain significant. It will be in addition to the specialist niche markets for high-density dwellings pitched at younger professionals and international students, as at Docklands and near the big educational institutions.


The reference here is to suburban areas bordering the affluent suburbs discussed above. Examples include parts of Monash, Whitehorse and Kingston in the East, and Moreland in the North. The housing stock dates primarily to the 1960s; thus there is no heritage component. The housing is generally modest. It is almost all detached housing on lots of 600 to 700 square metres. The dwellings are typically small double or triple fronted cream brick veneers (with a minority of wooden houses interspersed). The style and fittings are outmoded by contemporary standards. However, there is a premium associated with their location which is particularly evident near to commercial, employment and transport nodes such as Box Hill and Glen Waverley. Houses in these areas sell for around $400,000.

There is strong demand for properties in well located sites in middle suburbia. Persons moving into these locations gain the benefit of past public and private investment. For example, within these areas, some local public schools have amongst the best academic results in the government school system. Families, including skilled Asian migrant families, appear to be strongly attracted to these locations because of their concern about the educational future of their children. Much of middle suburbia is served by the rail system, which also provides private school access for the children of families in these areas.

Despite the rhetoric about the maintenance of neighbourhood character, the current record is that investors and developers can readily meet municipal design demands applied to infill proposals. As noted, there is a well established industry of builders and investors who are looking for sites to redevelop. When the demand for medium density housing is there, they are well set up to provide the product. The outcome of these supply and demand pressures is that the builder/investor can contemplate the prospect of a reasonable profit from purchasing a house at around $400,000 and investing in at least two dwellings which will ultimately bring a price of well over $400,000 each.

For people with modest resources who are considering locating in these areas, the choice is between an established, perhaps dowdy, detached house and a new dwelling on a site half the size of a conventional block. The price difference is small, reflecting the positive and negatives of the two alternatives. The purchase of an older detached house usually implies considerable further investment downstream in order to upgrade it to contemporary standards. The medium-density home provides less open space, but modern fittings and design. Apartments located in activity centres in these municipalities, at least for the foreseeable future, are unlikely to be competitive with infill, because of infill’s added attraction of some private outdoor space. Infill buildings usually do not share common walls with neighbours, nor do they have people living above or below, which is also an important consideration for people buying what is in effect a substitute for a detached house.

The conclusion is that there is a strong potential demand for infill in middle suburban areas and relatively few constraints to its growth. It may provide a greater share of new housing supply than the Melbourne 2030 estimates currently suggest.


At present, there are not many medium-density dwellings in the rest of Melbourne. However, it is likely that this demand will grow in some outer-suburban areas during the next few decades.One reason is the escalation of new housing prices on the Melbourne’s suburban frontier. As noted in Chapter Three, the minimum price of house and land right across the fringe of Melbourne is now in the $250,000 to $300,000 range. As a consequence, battlers at the lower end of the first home market now have to look for some alternative.

On the supply side there is a large expanse of detached housing in established outer suburban Melbourne which could accommodate infill developments. The housing tends to be a later vintage than that described above but again is typically modest in size and outmoded in style relative to contemporary standards. Detached houses in the established outer suburbs had reached around a median price of $230,000 by 2003.

Table 5.2 Selected Melbourne localities, tenure by dwelling structure, occupied dwellings, 2001

^Other rental is overwhelmingly private rental but it does include some rented dwellings where the landlord was not stated.

*Other tenure includes dwellings being occupied rent free or under life tenure schemes and those where the tenure was not stated.

Percentages may not add to 100 because of rounding.

Australian Bureau of Statistics, Census 2001, Census Basics, Table 19

The median price of flats and townhouses sold in these areas was generally much less than $200,000, for example the median price in Dandenong in 2003 was $162,000 (see Table 4.7 in Chapter Four). Thus for the would-be first home purchaser a medium-density dwelling offered a cheaper entry point into the housing market. In addition, for those forced by necessity into the rental market, such dwellings represented the most feasible option. An analysis of the 2001 Census data shows that the medium density market in these areas is predominately a rental one (See Table 5.2). However, if the argument pursued above is correct, in future a higher proportion of the residents will purchase these properties if available. Given the likely high cost of high density flats in activity centres – should any investor take up the challenge – it is unlikely that these would compete with infill.

In all areas shown in Table 5.2, at least half of the flats are rented. Flats and, to a lesser extent, semi-detached dwellings tend to provide housing for those unable or not ready to purchase a home. These include young people and other people new to the Melbourne market such as migrants from overseas, inter or intrastate, as well as households preferring to rent for lifestyle reasons.

This points feeds into a larger issue about the housing market which, though not tackled in this book, is of great concern to many housing analysts. This is the increased difficulties aspiring first-home buyers face in purchasing a dwelling because of the escalation in home prices since the housing boom began in the mid-1990s. More such persons are being forced into the rental market. The demand for low cost medium and high density housing which is directed at the rental market is likely to increase, particularly in areas of relatively low housing amenity outer-suburban areas like Dandenong, Sunshine and Whittlesea.


The analysis shows that the potential demand for infill is high – indeed much greater than is likely for high-density dwellings constructed in activity cities. Furthermore, the current ResCode regulations governing infill facilitate its proliferation. The only cloud on the horizon might be possible limits to the availability of older style detached housing suitable for infill. However, as the following comments indicate, there is a huge stock of detached housing vulnerable to infill.

Figure 5.3 Development of Melbourne residential subdivisions to 1976

Source: VicUrban, Property data and market research

Melbourne’s physical hallmark has long been the predominance of detached houses. Most of the sites where detached houses were built in the twentieth century can accommodate dual occupancy. The map in Figure 5.3 shows the phasing of suburban development around the original ring which was built in the period up to 1922. There is an extensive zone built in the period since then to World War 2, but, as can be seen on the map, the scale of development between 1961 and 1976 was huge. This means that there were very large tracts of modest-sized detached houses, typically on 600 to 700 square metre lots, built at least thirty years ago. Much of this area is ripe for refurbishment or redevelopment as infill.

Because new housing after World War 2 was usually located near railway lines, it now constitutes the oldest and generally the most outmoded of post-war suburbia. It is therefore particularly vulnerable to infill. Some of these areas are near to activity centres and public transport hubs. However, it cannot be concluded that infill and proximity to public transport are causally associated. As indicated, the age and appearance of the dwelling are better indicators of propensity for infill. Unlike modern housing estates, few sites are protected by covenants preventing their redevelopment into higher-density housing.

As the following case studies show, the process is well underway in middle and outer suburban areas which are distant from the zones which prompted the formation of the Save Our Suburbs movement.


Figure 5.4, showing infill patterns in parts of Glen Waverley in the City of Monash, indicates the extent of infill in one such area. This area at the end of the Glen Waverley railway line is on the eastern fringe of the 1947 to 1961 development zone shown in Figure 5.3. The map in Figure 5.4 is based on the cadastral (property title) map, as of 2003, which shows the boundaries of properties held in private ownership. Medium density housing that can be identified visually on this map has been shaded grey. However, a recent field inspection of the area2 enclosed by the heavy black line identified many more dwellings built on subdivisions of conventional lots than were shown by the cadastral map. These are marked in black. Some of these are dual occupancy dwellings built in the 1980s which are not shown on the cadastral map, but others are dwellings not recorded on the 2003 map because they were built after this date.

Locations where the existing dwellings are of pre-1970 stock are the most common location for infill, particularly if the block size is large and the dwelling style is modest and lacking modern amenities such as en suites. This description fits much of the area close by the Glen Waverley activity centre. The map illustrates the way infill dwellings are dispersed in a random fashion across target areas – a dispersal which reflects the opportunistic nature of infill. As the older style houses come onto the market they are purchased, demolished and replaced by two or more dwellings. Once started, the infill process moved quickly across this part of Glen Waverley. It was fed by the willingness of the owners to sell (including owner occupiers interested in developing a dual occupancy house in their own backyard), and helped by many small builders and developers looking for such sites, and a growing pool of potential buyers (either landlords or owners). A calculation of the number of sites covered by infill relative to the number of sites with conventional detached housing in the area inspected showed that more than one in five (or 20 per cent) of the original lots now contained infill dwellings.3 This means that for the area in question there are now at least 40 infill dwellings to every 80 conventional detached houses. Many of these detached houses will be replaced by infill dwellings in the not too distant future.

Figure 5.4 Recent infill, observed and as indicated by changes on the cadastral maps for 2000 and 2003, area around Glen Waverley Activity Centre

Victorian Planning Scheme Data, Department of Sustainability and Environment and field work

Figure 5.5 Recent infill as indicated by changes on the cadastral maps for 2000 and 2004, Noble Park area in City of Greater Dandenong

Victorian Planning Scheme Data, Department of Sustainability and Environment and City of Greater Dandenong

The map of Noble Park, a neighbourhood located to the south east of Springvale in the municipality of Greater Dandenong, provides another example of the extent of infill in an older suburban area. Figure 5.5 provides a comparison of medium density sites between 2000 and 2004 as shown on the cadastral maps only. A site-by-site field inspection was not carried out as in Glen Waverley. The area in question was attractive for infill for the same reasons as Glen Waverley: block sizes tend to be quite large (thus allowing two to four new dwellings to be constructed on the one lot), road pavements are ample and street tree plantings have matured to the point that the area offers a pleasant suburban appearance. Though distant from the CBD, the location near Princes Highway and the rich job availability in Greater Dandenong and the adjoining municipalities of Monash and Kingston add to its attractiveness as a location.


The foregoing analysis indicates that the proportion of dwellings constructed as infill in established suburban areas will almost certainly increase if there are no changes to the present Melbourne 2030 strategies.

Melbourne 2030 has nothing to say about what infill might mean for Melbourne’s built heritage and its suburban lifestyle. The following discussion and accompanying photographs illustrate the likely outcomes. Readers will be aware of the positive attributes associated with infill, which have been well canvassed by advocates. The material below looks at the other side of the coin.


The most obvious threat has to do with congestion. As residents living in areas with high levels of infill can hardly fail to notice, a trip out of their neighbourhood usually involves an exercise in street decorum and advanced steering skills. The reason is that infill multiplies the number of cars relative to the number of allotments. Double garages are a ubiquitous accompaniment to contemporary infill. Another result is the greater incidence of street parking. The combination of more resident-induced car traffic and more obstacles along the street multiplies the congestion effect. The streets in question were never designed for the scale of the traffic resulting from infill.

Figures 5.6a and 5.6b show the reductio ad absurdum of this with a contemporary high-density subdivision in Osaka in Japan. This an extreme example of the point made in Chapter Two that higher density living in the contemporary world does not necessarily lead to a reduction in the possession of private automobiles. The coupling of high residential densities with continued car use is now characteristic of some new infill subdivisions in Melbourne. Figures 5.7a and 5.7b show a Melbourne outcome not dissimilar to the Osaka subdivision shown.

Figure 5.6a New housing estate in suburban Osaka; compact separate detached housing; some housing lots are down to less than 80 square metres in area; continued car dependency.

Figure 5.6b Same housing estate as Figure 5.6a

Figure 5.7a Osaka? – No. High-density semi-detached housing estate on reclaimed industrial land, in the inner Melbourne suburb of Brunswick; heavy car congestion due to continued car dependence

Figure 5.7b Dwellings in this housing estate have carports at rear; on weekends, carport areas are often used for recreation in lieu of back yard, leaving cars crammed into narrow access roads; no footpaths


As stated earlier, for most of Melbourne’s residents, it is the distinctive suburban style of the city which gives it its special character: its liveability. The cosmopolitan centre – to the extent that it has not been destroyed by an earlier attack of high-density dwelling construction – is also important, but not at the heart of daily life for the great majority of the city’s residents.

Melbourne’s suburbia is a living thing to which Melbourne’s residents are continually contributing. Householders may be highly individualistic in their attachment to their allotment and jealously guard their right to do with it as they please. Nonetheless, in the great majority of cases, each householder contributes his/her share of the garden plantings, which makes up the overall ambience of their streetscape. Municipalities do their bit by providing street trees. The dominance of the gardens over the built structures is illustrated in Figures 5.8 to 5.10.

Figure 5.8 A suburban streetscape in the middle suburb of Glen Waverley (City of Monash)

Figure 5.9 Generous street setbacks have facilitated an aesthetically pleasing streetscape for the residents of these early post-war homes in this modest northern suburb in the City of Moreland

Figure 5.10 Modest post-war separate detached dwelling, visually integrated into its garden surrounds (Preston, City of Darebin)

The garden style shown in the photos will not survive an infill invasion. The high site coverage, inevitable with the addition of extra dwellings on a conventional house site, combined with the limited setbacks required under ResCode, is destructive of the traditional tree and shrub canopy. As noted, under The Good Design Guide, developers only had to provide six metre front setbacks – now extended to nine metres (or less if the adjoining houses have smaller setbacks) under ResCode.

In a current dual occupancy development, the usual nine metre setback may be encroached upon by an entrance portico and nearly always by a relatively wide paved area to provide entry for the household cars. The potential space for canopy trees and shrubs is thus much reduced. Householders also have to worry about whether any trees planted will compete for root space with house foundations, water, sewerage and drainage lines, or for air space with solar access, house gutters and any overhead lines. There are also close-by neighbours to consider. The result is that plantings are generally small scale – pencil pines, iceberg roses, box hedges, cottage style shrubbery – not canopy trees or shrubs.

This situation is even more evident with corner infill. Under ResCode, the ‘front’ setback of the dual occupancy dwelling facing the side-street need only be three metres so that there is no possibility of canopy trees. Figure 5.11 provides an illustration. Where this occurs in a neighbourhood with short streets, and thus many corner sites, the overall impact is to strip the location of any prominent trees.

The absence of canopy trees is even more marked at the side and rear of infill. Canopy trees have long been a feature of the suburban backyard. Figure 5.12 shows an areal view of part of Glenroy (in Moreland North). It illustrates the contribution of backyard canopy trees to the landscape in the area. Backyard trees tend to be taken for granted in Melbourne’s suburban landscape. It is only when an overview like that provided in the figure is shown that an appreciation of their significance becomes evident.

Figure 5.11 Corner block infill; setback at one end of block is 7.6 metres, but does not apply to rear residence in side street, where residence entrance is only 3 metres from footpath (City of Monash)

Figure 5.12 Abutting back yards provide opportunity for spines of large canopy trees, which run parallel to local streets (Glenroy, City of Moreland)

This vegetative cover will not be replicated in infill projects. Depending on the interpretation of ResCode for a particular project, it is highly unlikely that a canopy tree would be planted in the small available space at the rear or that, if trees are present, they will escape the developer’s chainsaw.

A typical case of the rear aspect of an infill allotment is shown in Figures 5.13a and 5.13b. In another example shown in Figure 5.14, the backyard garden area is sacrificed in order to accommodate a new dwelling behind the existing house. Perhaps more prevalent are developments where the original house is demolished and two or more medium density dwellings replace it. These vary in quality (and price) according to location. More up-market versions are shown in Figures 5.15a and 5.15b and more downmarket types in Figures 5.16a and 5.16b. The developments have in common the absence of canopy trees or shrubs.

Figures 5.13a and 5.13b Infill dwellings constructed under ResCode rules, with minimum 9 metre setback and rear space dominated by additional dwelling and excessive impermeable paving (City of Monash)

Figure 5.14 Basic dual occupancy; original dwelling in front and additional dwelling occupying back yard (Keilor/Sunshine, city of Brimbank)

Note the scale of the dwellings in 5.15a and 5.15b. Especially in higher income areas, developers seek to maximise the size of the building (notwithstanding the limited space surrounding it) because it maximises their financial return from investment in the property. Inevitably the building dominates the allotment. This is the reverse of the situation with most established detached housing on conventional blocks where the garden plays a more important visual role.

Figure 5.15a Several large two-storey dwellings to replace former single detached dwelling; large paved central access area; little space remaining for garden or large trees

Figure 5.15b More expensive infill development; several large dwellings in place of original modest dwelling; nearly completely paved

Figures 5.16a and 5.16b Cheap infill development; single detached house replaced with a number of small unit dwellings; common paved access along one side; small shrubbery virtually the only option for garden development

The lower priced areas of Melbourne have long been subject to the worst kind of infill: the six pack. Some examples are shown in Figures 5.17a and 5.17b. While ResCode does not allow the replication of such dwellings, once the neighbourhood character has been changed on account of these dwellings, it is easier to gain approval for buildings at similar densities and setbacks.

Figure 5.17a (Chelsea) and 5.17b (Keilor) Older style row units characteristic of much 1960s and 1970s infill development; many areas with a history of this style of development often appear prone to subsequent low-cost medium-density infill

Once infill occupies most of the sites in a street, the denuding of the previous green landscape is almost total (see Figure 5.18). When examined closely, the aerial photo provided (Figure 5.19) reveals the extent of the loss. It shows a suburb in Melbourne’s north in two parts. One, to the left, is of an area composed largely of medium density infill. Canopy trees are notable by their absence. The adjoining area to the right shows detached housing with a relatively thick canopy covering.

Figure 5.18 Virtual denuding of the streetscape through cumulated infill (city of Monash)

Figure 5.19 Aerial view showing negative impact of multi-unit infill development upon gardens and canopy cover (Preston, City of Darebin)


The dominant form of dwelling in traditional Melbourne suburbia was the low-slung bungalow which was largely obscured by the surrounding garden. It can be read in retrospect as an intelligent adaptation to the harshness of Melbourne’s climate in that it provided shade in the summer and shelter from cold winds in the winter. As indicated, this protective canopy disappears with infill. But that is not all that changes. So, too, do the houses. The contemporary trend in infill design is towards large, blocky, two-storey buildings. As Figures 5.15a and 5.15b illustrate dramatically, these buildings often dominate the streetscape. They represent a visual intrusion which many residents see as a violation of their suburban style of life.

Melbourne 2030 continually trumpets that it is about creating new dwelling opportunities without sacrificing the best of old Melbourne. However, there would be few residents of Melbourne who would consider that this is the case where infill is rampant.


The Australian Bureau of Statistics (ABS) information detailed in Figure 5.1 and Table 5.1 understates the extent of infill in many parts of established suburbia during the 1990s. In the discussion in this chapter, infill refers to new dwellings constructed on sites formerly occupied by detached houses or on vacant lots in established areas where the lots were originally intended for detached houses. As noted above, the ABS data only include flats, apartments and town houses in the ‘other residential’ category where the dwellings are attached to each other.4 The data do not include the product of a dual occupancy development where a new detached dwelling is constructed adjacent to an existing dwelling on the same block, or where two new dwellings replace an original dwelling which has been demolished and the two dwellings are separate from each other. Separate dwellings are defined by the ABS as houses; thus the official ABS statistics undercount the extent of infill.

One indication of the potential scale of the undercount is provided by the Western Australian branch of the ABS which investigated how many Perth building approvals for houses over the 1998-99 to 2000-01 period were actually ‘grouped dwellings’ — that is houses which were separate or detached, yet built in multiples of two or more on one block. It concluded that some 14 per cent of all dwelling approvals for separate houses could be defined in this way (Australian Bureau of Statistics 2002). This means that an estimate of the share of total approvals which were for medium or high density dwellings should include both the official ‘other residential’ dwellings (20 per cent in Perth over this period) and the separate houses which were actually grouped dwellings. This calculation suggests that between 31 and 32 per cent of the Perth dwelling approvals over this three year period were for medium and high density dwellings rather than 20 per cent.

There is one study on this issue which covers some municipalities of inner Melbourne. Its findings were that in 2002-03, if detached houses completed as part of a dual occupancy are included in the medium density count, the proportion of total dwelling approvals which are medium density increases markedly relative to the ABS figures. In the case of Boroondara, the proportion increases from around 50 per cent to nearly 70 per cent (Buxton and Tieman 2004 p.15). The impact of detached houses built as part of dual built-up municipalities of Port Phillip and Yarra.

On the other hand, it is also possible that some dwellings included in the other residential category should not be regarded as infill as defined above, particularly during the Kennett era. During this period, sites that once accommodated schools or other public building sites were often sold to developers and, in some cases, these were developed in the form of medium or high density housing. Such dwellings did not replace existing detached houses. They did, however, have similar effects to infill. They added to congestion in the surrounding areas, as well as causing the loss of open space (for example where school sites were built upon).

Table 5.3 indicates the number of building approvals for detached houses in Melbourne over the period 1996-97 to 2003-04 as defined by the ABS. It is obvious from the substantial number of ‘detached houses’ in the inner and middle areas of the city that these figures greatly overstate the number that would have been built on conventional lots. Most of the building approvals for detached houses in places like Preston, Banyule, Maribyrnong and Box Hill (as Buxton and Tieman have shown for Boroondara) would be the product of dual occupancies.

Table 5.3 Building approvals for detached houses by region, number and per cent of Melbourne, 1996–97 to 2003–04

Australian Bureau of Statistics Building Approvals


1     Side and rear setbacks under ResCode vary with the height of the building. The minimum is one metre plus 0.3 metres for every metre of height over 3.6 metres up to 6.9 metres, plus 1 metre for every metre of height over 6.9 metres. In the case of Bayside, buildings should be set back 2 metres from the side boundary and 3 metres from the rear boundary plus 0.6 for each metre in height to 6.9, plus 2 metres for each metre in height over 6.9.

2     The area contains four Census Collection Districts.

3     Note that many of the dwellings fronting the East side of Springvale Road and facing the commercial area have been converted to medical rooms and are not included in this calculation.

4     ABS ( describes an ‘other residential’ building as a building other than a house primarily used for long-term residential purposes. An ‘other residential’ building contains more than one dwelling unit. Other residential buildings include semi-detached, row or terrace houses or townhouses of one or two storeys; flats, units or apartments whether in a building of one or more storeys or attached to a house; and other or unknown building types.


Australian Bureau of Statistics, ‘Housing, special article – a view of housing density in Perth’, Western Australia Statistical Indicators 2002.

Buxton, M; Tieman, G. Urban Consolidation on Melbourne 1988–2003. Melbourne: RMIT; 2004.

Department of Infrastructure (Victoria). The Good Design Guide. Revision 1. Aug1997.

Department of Infrastructure (Victoria). ‘Implementing ResCode – new provisions for residential development in Victoria’. [Publication date unknown. Citation date Feb 2005]. Brochure available from:$FILE/Implementing%20ResCode%20brochure.pdf.

Department of Sustainability and Environment (Victoria). Urban Development Program Report 2003. 2003. 204 p.

Department of Sustainability and Environment (Victoria). Urban Development Program Report 2004. 2004. 208 p.

Eccles, D; Bryant, T. Statutory Planning in Victoria. 2nd ed. Leichhardt, New South Wales: Federation Press; 1999. 245 p.

Lewis, Miles. Suburban Backlash: The Battle for the World’s Most Liveable City. Melbourne: Bloomings Books; 1999. 296 p.

Productivity Commission (Australian Government). First Home Ownership. Productivity Commission Inquiry Report no. 28. 2004. Available from: index.html.

Cite this chapter as: Birrell, B; O’Connor, K; Rapson, V; Healy, E. ‘Residential infill and its threat to Melbourne’s liveability’. In: Melbourne 2030: Planning Rhetoric Versus Urban Reality, Monash University ePress: Melbourne; 2005.

Melbourne 2030: Planning Rhetoric Versus Urban Reality

   by Bob Birrell, Kevin O’Connor, Virginia Rapson and Ernest Healy