Key growth areas for high density housing in the City of Melbourne (source: CoM Draft Housing Strategy)

Almost every major planning strategy published in Australia over the last 20 years has stressed the importance of supporting and increasing diversity on a range of dimensions, such as industry composition, ethnicity, age, family structure and, of course, socio-economic status.

The City of Melbourne’s newly released draft Housing Strategy is no exception. It emphasises increasing the proportion of lower income residents in the City, but one aspect I’m particularly interested in is its concern with what it sees as a lack of diversity in dwelling types.

It argues that too few “family friendly” three and four bedroom apartments are being built in the municipality.

Ninety six per cent of over 20,000 apartments in over 100 developments currently in the pipeline have just two or fewer bedrooms.

The draft Strategy says that diversity of dwelling type choice is important because “the right mix of homes contributes to developing sustainable communities that work for the population at large”. It also fosters,

a community which is inclusive of different household needs and circumstances, including family size, household composition, income and health. (1)

While diversity can have problems (see Diversity in cities: does it have to be uniform?), in principle it’s a very good thing; but the particular dimension and the spatial scale matter.

Other than in the case of directly subsidised housing (e.g. social housing), I’m not sure that the aim of increasing the proportion of “family friendly” dwellings within an area as small as the City of Melbourne is a realistic, or even a sensible, idea. (2)

At present, investors and one and two bedroom households value a city centre location highest and are prepared to pay more than families who are seeking affordable housing (the rich can continue to buy penthouses, of course). Families usually want more space than smaller households and, depending on their income, they mostly find a suitable compromise between space and price outside the precincts where most apartments are being built.

Requiring developers to provide more three and four bedroom apartments won’t be costless. If they’re compelled to build what they see as a sub-optimal mix of apartment types, then it’s possible some projects won’t proceed and the supply of smaller units will decrease, with consequent price impacts. If it’s achieved by development bonuses then it’s likely it will impose amenity costs on neighbouring land uses.

That might be OK if the goal of getting more families into the city centre is worthwhile, but it’s not at all clear to me that it’s such a valuable objective, especially compared to other choices like (say) providing housing for low income residents.

There’s no compelling reason why the city centre has to exhibit diversity on all dimensions and, in particular, why it needs to have a higher representation of families. It might be nice, but the centre can’t, and needn’t, mirror the metropolitan area as a whole on every or even most variables of interest.

The City of Melbourne isn’t like other municipalities; it’s already highly specialised on many dimensions. For example, it’s got by far the largest concentration of high-skill jobs in the metropolitan area, particularly in finance and government.

It alone of all municipalities has the State’s major cultural institutions, the Parliament, the courts, key entertainment and sporting venues, and many national and state headquarters for corporates, trade unions, industry associations, not-for-profits, and much more. It differs markedly from even its neighbours on key variables like dwelling density and infrastructure services.

It shouldn’t be surprising that the population in the core where most apartment construction is occurring is also very different from that of other municipalities. As I’ve pointed out before (see Who lives in the city centre?), the age, socio-economic status, tenure, household type and dwelling type profile of the resident population in the core is markedly different to the other 30 metropolitan municipalities and, for that matter, to the rest of the City of Melbourne.

The specialisation of the core in small dwellings doesn’t come at the expense of the rest of the metropolitan area. The City of Melbourne is tiny; it covers a mere 35 sq km. The areas where new apartments will be built – the “growth areas” – occupy only part of the municipality (especially Southbank, see exhibit).

The wider metropolitan area in contrast covers more than 2,500 sq km and 73% of the housing stock is “family friendly” detached dwellings. More than 20% of the existing housing stock in the City of Melbourne is made up of detached houses and low-rise terraces, semis and town houses that offer families direct access to the ground. This type of housing also accounts for 63% of the dwelling stock in the adjoining City of Yarra. (3)

The important point here is that while diversity is vital for a city, in most cases it needs to be understood at a larger spatial scale than individual buildings, streets, precincts, suburbs, or in many instances even municipalities. The scale depends on the variable; for example, communities with a common interest like ethnicity or sexual orientation often tend to concentrate in a few locations (e.g. see here, here, and here).

Cities aren’t geographically uniform. Businesses and people tend to locate in clumps. The “diversity dividend” of big cities lies in the greater probability that people will meet others who’re just like them and seek to locate nearby rather than spread more or less evenly across the metropolitan area. A diverse metro looks like a patchwork, not a uniform blend (see Are bigger cities less diverse?).

What matters is that the whole metropolitan area should welcome, embrace and find room for diversity; but in many cases the spatial pattern will quite reasonably manifest at the regional or even metropolitan level rather than the local level.

Melbourne City Council should recognise it’s managing the city centre, not a suburban municipality. It should understand it’s a specialised location and work out what sort of diversity matters to its role and how it can reinforce it by improving opportunities and lowering costs e.g improving housing affordability.

The draft Housing Strategy also has a goal to increase the proportion of low income residents in the centre. I think that should be one of Council’s key priorities; not because it improves socio-economic diversity at the very local level but because it provides housing for people who need it. (4)


  1. I can’t be sure what the Strategy specifically proposes should done to increase the supply of three and four bedroom apartments because, curiously, the section detailing the proposed actions doesn’t mention dwelling mix.
  2. An immediate practical problem is that families looking for a larger, affordable apartment to rent in the city centre will almost certainly be outbid by young professionals willing to share an apartment. Three incomes can pay more than one or two (renting is the appropriate tenure to consider because Council’s draft Housing Strategy says 85% of new apartments are bought by investors).
  3. Note also that while the draft Housing Strategy says only 4% of apartments “currently in the pipelines” have three or more bedrooms, not all these projects will proceed. The earlier discussion paper, Future Living, says 9% of all apartments built over 2006-12 had three bedroom or more; in 2012 it was 18%.
  4. It seems clear that spatial concentrations of socio-economic disadvantage are negative, but it’s less clear that lower socio-economic groups benefit unambiguously from living physically close to high income neighbours (see What can planners do about socio-economic polarisation?)