How is density changing in Australian cities? (2nd edition)

Sun 21 April, 2019

While Australian cities are growing outwards, densities are also increasing in established areas, and newer outer growth areas are some times at higher than traditional suburban densities.

So what’s the net effect – are Australian cities getting more or less dense? How has this changed over time? Has density bottomed out? And how many people have been living at different densities?

This post maps and measures population density over time in Australian cities.

I’ve taken the calculations back as far as I can with available data (1981), used the highest resolution population data available. I’ll discuss some of the challenges of measuring density using different statistical geographies along the way, but I don’t expect everyone will want to read through to the end of this post!

[This is a fully rewritten and updated version of a post first published November 2013]

Measuring density

Under traditional measures of density, you’d simply divide the population of a city by the area of the metropolitan region.

At the time of writing Wikipedia said Greater Sydney’s density was just 4.23 people per hectare (based on its Greater Capital City Statistical Area). To help visualise that, a soccer pitch is about 0.7 hectares. So Wikipedia is saying the average density of Sydney is roughly about 3 people per soccer field. You don’t need to have visited Sydney to know that is complete nonsense (don’t get me wrong, I love Wikipedia, but it really need to use a better measure for city density!).

The major problem with metropolitan boundaries – in Australia we use now Greater Capital City Statistical Areas – is that they include vast amounts of rural land and national parks. In fact, in 2016, at least 53% of Greater Sydney’s land area had zero population. That statistic is 24% in Melbourne and 14% in Adelaide – so there is also no consistency between cities.

Below is a map of Greater Sydney (sourced from ABS), with the blue boundary representing Greater Sydney:

One solution to this issue is to try to draw a tighter boundary around the urban area, and in this post I’ll also use Significant Urban Areas (SUAs) that do a slightly better job (they are made up of SA2s). The red boundaries on the above map show SUAs in the Sydney region.

However SUAs they still include large parks, reserves, industrial areas, airports, and large-area partially-populated SA2s on the urban fringe. Urban centres are slightly better (they are made of SA1s) but population data for these is only available in census years, the boundaries change with each census, the drawing of boundaries hasn’t been consistent over time, they include non-residential land, and they split off most satellite urban areas that are arguably still part of cities, even if not part of the main contiguous urban area.

Enter population-weighted density (PWD) which I’ve looked at previously (see Comparing the densities of Australian, European, Canadian, and New Zealand cities). Population-weighted density takes a weighted average of the density of all parcels of land that make up a city, with each parcel weighted by its population. One way to think about it is the average density of the population, rather than the average density of the land.

So parcels of land with no population don’t count at all, and large rural parcels of land that might be inside the “metropolitan area” count very little in the weighted average because of their relatively small population.

This means population-weighted density goes a long way to overcoming having to worry about the boundaries of the “urban area” of a city. Indeed, previously I have found that removing low density parcels of land had very little impact on calculations of PWD for Australian cities (see: Comparing the residential densities of Australian cities (2011)). More on this towards the end of this post.

Calculations of population-weighted density can also answer the question about whether the “average density” of a city has been increasing or decreasing.

But… measurement geography matters

One of the pitfalls of measuring population weighted density is that it very much depends on the statistical geography you are using.

If you use larger geographic zones you’ll get a lower value as most zones will include both populated and unpopulated areas.

If you use very small statistical geography (eg mesh blocks) you’ll end up with a lot fewer zones that are partially populated – most will be well populated or completely unpopulated, and that means your populated weighted density value will be much higher, and your measure is more looking at the density of housing areas.

To illustrate this, here’s an animated map of the Australian Capital Territory’s 2016 population density at all of the census geographies from mesh block (MB) to SA3:

Only at the mesh block and SA1 geographies can you clearly see that several newer outer suburbs of Canberra have much higher residential densities. The density calculation otherwise gets washed out quickly with lower resolution statistical geography, to the point where SA3 geography is pretty much useless as so much non-urban land is included (also, there are only 7 SA3s in total). I’ll come back to this issue at the end of the post.

Even if you have a preferred statistical geography for calculations, making international comparisons is very difficult because few countries will following the same guidelines for creating statistical geography. Near enough is not good enough. Worse still, statistical geography guidelines do not always result in consistently sized areas within a country (more on that later).

We need an unbiased universal statistical geography

Thankfully Europe and Australia have adopted a square kilometre grid geography for population estimates, which makes international PWD comparisons readily possible. Indeed I did one a few years ago looking at ~2011 data (see Comparing the densities of Australian, European, Canadian, and New Zealand cities).

This ABS is now providing population estimates on a square km grid for every year from 2006.

Here is what Melbourne’s estimated population density looks like on a km square grid, animated from 2006 to 2017:

The changes over time are relatively subtle, but if you watch the animation several times you’ll see growth – including relatively high density areas emerging on the urban fringe.

It’s a bit chunky, and it’s a bit of a matter of luck as to whether dense urban pockets fall entirely within a single grid square or on a boundary, but there is no intrinsic bias.

There’s also an issue that many grid squares will contain a mix of populated and non-populated land, particularly on the urban fringe (and a similar issue on coastlines). In a large city these will be in the minority, but in smaller cities these squares could make up a larger share of the total, so I think we need to be careful about this measure in smaller cities. I’m going to arbitrarily draw the line at 200,000 residents.

How are Australian cities trending for density using square km grid data 2006 to 2018?

So now that we have an unbiased geography, we can measure PWD for cities over time.

The following chart is based on 2016 Significant Urban Area boundaries (slightly smaller than Greater Capital City Statistical Areas but also they go across state borders as appropriate for Canberra – Queanbeyan and Gold Coast – Tweed).

Technical notes: You cannot perfectly map km squares to Significant Urban Areas. I’ve included all kilometre grid squares which have a centroid within the 2016 Significant Urban Area boundaries (with a 0.01 degree tolerance added – which is roughly 1 km). Hobart appears only in 2018 because the Hobart SUA crosses the 200,000 population threshold in 2018.

The above trend chart was a little congested for the smaller cities, so here is a zoomed in version without Sydney and Melbourne:

You can see most cities getting denser at various speeds, although Perth, Geelong, and Newcastle have each flat-lined for a few years.

Perth’s population growth slowed at the end of the mining boom around 2013, and infill development all but dried up, so the overall PWD increased only 0.2 persons/ha between 2013 and 2018.

Canberra has seen a surge in recent years, probably due to high density greenfield developments we saw above.

How is the mix of density changing? (2006 to 2018)

Here’s a look at the changing proportion of the population living at different densities for 2006-2018 for the five largest Australian cities, using square km grid geography:

It looks very much like the Melbourne breakdown bleeds into the Sydney breakdown. This roughly implies that Melbourne’s density distribution is on trend to look like Sydney’s 2006 distribution in around 2022 (accounting the for white space). That is, Melbourne’s density distribution is around 16 years behind Sydney’s on recent trends. Similarly, Brisbane looks a bit more than 15 years behind Melbourne on higher densities.

In Perth up until 2013 there was a big jump in the proportion of the population living at 35 persons / ha or higher, but then things peaked and the population living at higher densities declined, particularly as there was a net migration away from the inner and middle suburbs towards the outer suburbs.

Here’s the same for the next seven largest cities:

Of the smaller cities, densities higher than 35 persons/ha are only seen in Gold Coast, Newcastle, Wollongong and more recently in Canberra.

The large number of people living at low densities in the Sunshine Coast might reflect suburbs that contain a large number of holiday homes with no usual residents (I suspect the dwelling density would be relatively higher). This might also apply in the Gold Coast, Central Coast, Geelong (which actually includes much of the Bellarine Peninsula) and possibly other cities.

Also, the Central Coast and Sunshine Coast urban patterns are highly fragmented which means lots of part-urban grid squares, which will dilute the PWD of these “cities”.

Because I am sure many of you will be interested, here are animated maps for these cities:

Sydney

Brisbane

Perth

Adelaide

Canberra – Queanbeyan

Sunshine Coast

Gold Coast

Newcastle – Maitland and Central Coast

Wollongong

Geelong

What are the density trends further back in time using census data?

The census provides the highest resolution and therefore the closest measure of “residential” population weighted density. However, we’ve got some challenges around the statistical geography.

Prior to 2006, the smallest geography at which census population data is available is the collector district (CD), which average around 500 to 600 residents. A smaller geography – the mesh block (MB) – was introduced in 2006 and averages around 90 residents.

Unfortunately, both collector districts and mesh blocks are not consistently sized across cities or years (note: y axis on these charts does not start at zero):

Technical note: I have mapped all CDs and MBs to Greater Capital City Statistical Area (GCCSA) boundaries, based on the entire CD fitting within the GCCSA boundaries (which have not yet changed since they were created in 2011).

There is certainly some variance between cities and years, so we need to proceed with caution, particularly in comparing cities. Hobart and Adelaide have the smallest CDs and MBs on average, while Sydney generally has larger CDs and MBs. This might be a product of whether mesh blocks were made too small or large, or it might be that density is just higher and it is more difficult to draw smaller mesh blocks. The difference in median population may or may not be explained by the creation of part-residential mesh blocks.

Also, we don’t have a long time series of data at the one geography level. Rather than provide two charts which break at 2006, I’ve calculated PWD for both CD and mesh block geography for 2006, and then estimated equivalent mesh block level PWD for earlier years by scaling them up by the ratio of 2006 PWD calculations.

In Adelaide, the mesh block PWD for 2006 is 50% larger than the CD PWD, while in the Australian Capital Territory it is 110% larger, with other cities falling somewhere in between.

Would these ratios hold for previous years? We cannot be sure. Collector Districts were effectively replaced with SA1s (with an average population of 500, only slightly smaller) and we can calculate the ratio of mesh block PWD to SA1 PWD for 2011 and 2016. For most cities the ratio in 2016 is within 10% of the ratio in 2011. So hopefully the ratio of CD PWD to mesh block PWD would remain fairly similar over time.

So, with those assumptions, here’s what the time series then looks like for PWD at mesh block geography:

As per the square km grid values, Sydney and Melbourne are well clear of the pack.

Most cities had a PWD low point in 1996. That is, until around 1996 they were sprawling at low densities more than they were densifying in established areas, and then the balance changed post 1996. Exceptions are Darwin which bottomed out in 2001 and Hobart which bottomed in 2006.

The data shows rapid densification in Melbourne and Sydney between 2011 and 2016, much more so than the square km grid data time series. But we also saw a significant jump in the median size of mesh blocks in those cities between 2011 and 2016 (and if you dig deeper, the distribution of mesh block population sizes also shifted significantly), so the inflection in the curves in 2011 are at least partly a product of how new mesh block boundaries were cut in 2016, compared to 2011. Clearly statistical geography isn’t always good for time series and inter-city analysis!

How has the distribution of densities changed in cities since 1986?

The next chart shows the distribution of population density for Greater Capital City Statistical Areas based on collector districts for the 1986 to 2006 censuses:

You can more clearly see the decline in population density in most cities from 1986 to 1996, and it wasn’t just because most of the population growth was a lower densities. In Hobart, Canberra, Adelaide, Brisbane and Melbourne, the total number of people living at densities of 30 or higher actually reduced between 1986 and 1996.

Here is the equivalent chart for change in density distribution by mesh block geography for the capital cities for 2006, 2011, and 2016:

I’ve used the same colour scale, but note that the much smaller geography size means you see a lot more of the population at the higher density ranges.

The patterns are very similar to the distribution for square km grid data. You can see the how Brisbane seems to bleed into Melbourne and then into Sydney, suggesting a roughly 15 year lag in density distributions. This chart also more clearly shows the recent rapid rise of high density living in the smaller cities of Canberra and Darwin.

The next chart shows the 2016 distribution of population by mesh block density using Statistical Urban Area 2016 boundaries, including the smaller cities:

Gold Coast and Wollongong stand out as smaller cities with a significant portion of their population at relatively high densities, but a fair way off Sydney and Melbourne.

(Sorry I don’t have a mesh block times series of density distribution for the smaller cities – it would take a lot of GIS processing to map 2006 and 2011 mesh blocks to 2016 SUAs, and the trends would probably be similar to the km grid results).

Can we measure density changes further back in history and for smaller cities?

Yes, but we need to use different statistical geography. Annual population estimates are available at SA2 geography back to 1991, and at SA3 geography back to 1981.

However, there are again problems with consistency in statistical geography between cities and over time.

Previously on this blog I had assumed that guidelines for creation of statistical geography boundaries have been consistently applied by the ABS across Australia, resulting in reasonably consistent population sizes, and allowing comparisons of population-weighted density between cities using particular levels of statistical geography.

Unfortunately that wasn’t a good assumption.

Here are the median population sizes of all populated zones for the different statistical geographies in the 2016 census:

Note: I’ve used a log scale on the Y-axis.

While there isn’t a huge amount of variation between medians at mesh block and SA1 geographies, there are massive variations at SA2 and larger geographies.

SA2s are intended to have 3,000 to 25,000 residents (a fairly large range), with an average population of 10,000 (although often smaller in rural areas). You can see from the chart above that there are large variances between medians of the cities, with the median size in Canberra and Darwin below the bottom of the desired range.

I have asked the ABS about this issue. They say it is related to the size of gazetted localities, state government involvement, some dense functional areas with no obvious internal divisions (such as the Melbourne CBD), and the importance of capturing indigenous regions in some places (eg the Northern Territory). SA2 geography will be up for review when they update statistical geography for 2021.

While smaller SA2s mean you get higher resolution inter-censal statistics (which is nice), it also means you cannot compare raw population weighted density calculations between cities at SA2 geography.

However, all is not lost. We’ve got calculations of PWD on the unbiased square kilometre grid geography, and we can compare these with calculations on SA2 geography. It turns out they are very strongly linearly correlated (r-squared of over 0.99 for all cities except Geelong).

So it is possible to estimate square km grid PWD prior to 2006 using a simple linear regression on the calculations for 2006 to 2018.

But there is another complication – ABS changed the SA2 boundaries in 2016 (as is appropriate as cities grow and change). Data is available at the new 2016 boundaries back to 2001, but for 1991 to 2000 data is only available on the older 2011 boundaries. For most cities this only creates a small perturbation in PWD calculations around 2001 (as you’ll see on the next chart), but it’s larger for Geelong, Gold Coast – Tweed Heads and Newcastle Maitland so I’m not willing to provide pre-2001 estimates for those cities.

The bottom of this chart is quite congested so here’s an enlargement:

Even if the scaling isn’t perfect for all history, the chart still shows the shape of the curve of the values.

Consistent with the CD data, several cities appear to have bottomed out in the mid 1990s. On SA2 data, that includes Adelaide in 1995, Perth and Brisbane in 1994, Canberra in 1998 and Wollongong in 2006.

Can we go back further?

If we want to go back another ten years, we need to use SA3 geography, which also means we need to switch to Greater Capital City Statistical Areas as SA3s don’t map perfectly to Significant Urban Areas (which are constructed of SA2s). Because they are quite large, I’m only going to estimate PWD for larger cities which have reasonable numbers of SA3s that would likely have been fully populated in 1981.

I’ve applied the same linear regression approach to calculate estimated square kilometre grid population weighted density based on PWD calculated at SA3 geography (the correlations are strong with r-squared above 0.98 for all cities).

The following chart shows the best available estimates for PWD back to 1981, using SA3 data for 1991 to 2000, SA2 data for 2001 to 2005, and square km grid data from 2006 onwards:

Technical notes: SA3 boundaries have yet to change within capital cities, so there isn’t the issue we had with SA2s. The estimates based on SA2 and SA3 data don’t quite line up between 1990 and 1991 which demonstrates the limitations of this approach.

The four large cities shown appear to have been getting less dense in the 1980s (Melbourne quite dramatically). These trends could be related to changes in housing/planning policy over time but they might also be artefacts of using such a coarse statistical geography. It tends to support the theory that PWD bottomed out in the mid 1990s in Australia’s largest cities.

Could we do better than this for long term history? Well, you could probably do a reasonable job of apportioning census collector district data from 1986 to 2001 censuses onto the km grid, but that would be a lot of work! It also wouldn’t be perfectly consistent because ABS use dwelling address data to apportion SA1 population estimates into kilometre grid cells. Besides we have reasonable estimates using collector district geography back to 1986 anyway.

Melbourne’s population-weighted density over time

So many calculations of PWD are possible – but do they have similar trends?

I’ve taken a more detailed look at my home city Melbourne, using all available ABS population figures for the geographic units ranging from mesh blocks to SA3s inside “Greater Melbourne” and/or the Melbourne Significant Urban Area (based on the 2016 boundary), to produce the following chart:

Most of the datasets show an acceleration in PWD post 2011, except the SA3 calculations which are perhaps a little more washed out. The kink in the mesh block PWD is much starker than the other measures.

The Melbourne SUA includes only 62% of the land of the Greater Melbourne GCCSA, yet there isn’t much difference in the PWD calculated at SA2 geography – which is the great thing about population-weighted density.

All of the time series data suggests 1994 was the year in which Melbourne’s population weighted density bottomed out.

Appendix 1: How much do PWD calculations vary by statistical geography?

Census data allows us to calculate PWD at all levels of statistical geography to see if and how it distorts with larger statistical geography. I’ve also added km grid PWD calculations, and here are all the calculations for 2016:

Technical note: square km grid population data is estimated for 30 June 2016 while the census data is for 9 August 2016. Probably not a significant issue!

You can see cities rank differently when km grid results are compared to other statistical geography – reflecting the biases in population sizes at SA2 and larger geographies. Wollongong and Geelong also show a lot of variation in rank between geographies – probably owing to their small size.

The cities with small pockets of high density – in particular Gold Coast – drop rank with large geography as these small dense areas quickly get washed out.

I’ve taken the statistical geography all the way to Significant Urban Area – a single zone for each city which is the same as unweighted population density. These are absurdly low figures and in no way representative of urban density. They also suggest Canberra is more dense than Melbourne.

Appendix 2: Issues with over-sized SA1s

As I’ve mentioned recently, there’s an issue that the ABS did not create enough reasonably sized SA1s in some city’s urban growth areas in 2011 and 2016. Thankfully, it looks like they did however create a sensible number of mesh blocks in these areas, as the following map (created with ABS Maps) of the Altona Meadows / Point Cook east area of Melbourne shows:

In the north parts of this map you can see there are roughly 4-8 mesh blocks per SA1, but there is an oversized SA1 in the south of the map with around 50 mesh blocks. This will impact PWD calculated at SA1 geography, although these anomalies are relatively small when you are looking at a city as large as Melbourne.


Are Australian cities sprawling with low-density car-dependent suburbs?

Wed 30 January, 2019

Many people talk about urban growth in Australian cities being car-dependent low-density suburban sprawl. But how true is that in more recent times? Are new greenfield density targets making a difference? Are cities growing around their rapid public transport networks? And how do growth areas compare to established areas at a similar distance out from city centres?

This post takes a look at what census data can tell us about outer urban growth areas in terms of population density, motor vehicle ownership, distance from train/busway stations, and journey to work mode shares.

How much of city population growth is in outer areas?

Firstly a recap, here is the percentage of annual population growth in each city that has occurred in “outer” areas (defined by groupings of SA3s around the edges of cities – refer my previous post for maps showing outer areas) for Greater Capital City Statistical Areas.

Sydney has had less than a third of its population growth in outer areas since around 2003, while Perth has mostly had the highest outer growth percentage (since 1996), and more recently pretty much all population growth in Perth has been on the fringe. You can see how the other cities sit in between.

However, not all of this “outer” population growth was in urban growth on the fringe. For that we need to distinguish between urban growth and infill development, even in “outer” areas. So we really need a better definition of outer growth areas.

How to define outer urban growth areas

I have built groupings of SA1s (Statistical Area Level 1) that try to represent outer urban greenfield residential development. SA1s are the smallest census geographic areas (average population 400) for which all census data variables are available.

I’ve selected 2016 SA1s that meet all of the following criteria:

  • Brand new SA1 or significant population growth: The 2016 SA1 is new and cannot be matched to a 2011 SA1 (by location/size and/or ABS correspondences), or if it can be matched, the population at least doubled between 2011 and 2016. Brand new SA1s are very common in urban growth areas as new SA1s are created to avoid oversized SA1s on last census boundaries (except this doesn’t always happen – more on that shortly).
  • In an SA2 with significant population growth: The SA2 (Statistical Area Level 2 – roughly suburb sized with typically 3,000 to 25,000 residents) that contains the SA1 had population growth of at least 1000 people between 2011 and 2016 (based on 2016 boundaries). That is, the general area is seeing population growth, not just one or two SA1s.
  • Are on – or close to – the urban fringe. I’ve filtered out particular SA2s that I’ve judged to be contain all or mostly in-fill development rather than greenfield development, or that are largely surrounded by existing urban areas and are not close to the urban fringe. I’ll be the first to admit that some of the inclusions/exclusions are a little arbitrary.

The criteria aren’t perfect, but it seems to work pretty well when I inspect the data. I’m calling these “Growth SA1s” or outer urban growth in this post.

For urban centres, I’m using Significant Urban Area 2016 boundaries (rather than Greater Capital City boundaries), and I’ve bundled Yanchep with Perth, Melton with Melbourne, and the Sunshine Coast and Gold Coast with Brisbane to form South East Queensland (SEQ).

Where are these outer urban growth areas?

What follows are maps for each city with the density of these growth SA1s shown by colour.

Melbourne’s northern and western growth areas:

Technical note: The maps do not show non-growth SA1s with fewer than 5 people per hectare, or “growth SA1s” with fewer than 1/hectare, although these SA1s are including in later analysis.

And the south of Melbourne:

Note: not shown on these Melbourne maps are isolated tiny growth SA1s in Rosebud and Mooroolbark.

Here are Sydney’s growth SA1s – all in the western suburbs:

Next up South East Queensland, starting in the north with the Sunshine Coast:

Northern Brisbane:

Outer urban growth is scattered in southern Brisbane and northern Gold Coast:

Gold Coast – Tweed Heads:

Perth’s northern and eastern growth areas:

Perth’s southern growth areas:

Note: Canning Vale East is an inclusion you could debate – the previous land use of the growth SA1s appear to have been rural based on satellite imagery.

Northern Adelaide:

Southern Adelaide:

And finally Canberra:

So how much of each cities’ population growth has been in outer growth areas?

Here’s a breakdown of the population growth for my six urban areas:

Over the five-year period, outer urban growth areas accounted for 19% of Sydney’s population growth, 43% of Melbourne’s, 37% of SEQ’s, 60% of Perth’s, 27% of Adelaide’s and 69% of Canberra’s.

Technical note: These “outer urban growth” figures are different to the chart at the top of this post which had a coarser definition of “outer” and used Greater Capital City boundaries. Some of my “outer urban growth” areas actually don’t quality as “outer” in the coarser definition, and I’ve also excluded several “outer” SA2s from “outer urban growth” where I’ve deemed the growth to be mostly infill. Hence the differences.

In case you are wondering, it’s not easy to create a longer-term time-series analysis about the proportion of population growth in “outer urban growth” areas because the classification of SA2s would have to change on a year-by-year basis which would be messy and somewhat arbitrary.

A challenge for density analysis: some SA1s are over-sized

You might have noticed some SA1s in the maps above are very large and show a low average density of 1-5 persons per hectare (I’ve coloured them in a light cyan). Many of these SA1s had thousands of residents in 2016, which is way more than the ABS guideline of 200 to 800 residents. Unfortunately what seems to have happened for 2011 and 2016 in some cities is that the ABS did not create enough SA1s to account for new urban areas. Some Melbourne SA1s had a population over 4000 in 2016. Many of these SA1s contain a combination or urban and rural land use, so their calculated density is rather misleading.

I’m designating any SA1s with more than 1000 residents and larger than 100 hectares as “oversized”, and I’ve exclude these from some density analysis below. Here’s a chart showing the proportion of outer growth area populations that are in oversized SA1s:

You can see it is a substantial problem in Sydney, Melbourne, Perth and South East Queensland, but miraculously not a problem at all in Adelaide or Canberra (I’m sure someone in ABS could explain why this is so!).

If you are interested, in 2011 it was a bigger problem in Melbourne, and only Canberra was fully clean.

So how dense are outer urban growth areas?

Firstly, I am excluding over-sized SA1s from this analysis for the reasons just mentioned.

Secondly, all cities will also have growth areas that were partially developed at the time of the census (ie some lots with occupied houses and other lots empty) so the densities measured here may be understated of the likely fully built-out density of these SA1s. That said, those areas perhaps are more likely to be in over-sized SA1s, but it’s hard to be sure. So keep this in mind when looking at growth area densities.

You can see dramatic differences, with Sydney, Canberra, and Melbourne showing higher densities, and South East Queensland with much lower densities. As we saw on the maps above, South East Queensland’s outer growth areas are very dispersed, so perhaps more of them are growing slowly and more of them are partially built-out? It’s hard to be sure.

But perhaps what is most remarkable is that Canberra had the highest densities in outer urban growth areas of any city – nothing like what you might consider suburban sprawl. Here’s what was 144.5 people per hectare in 2016 in Wright on Canberra’s new western growth front looks like:

(pic from Google Streetview, dated December 2016)

The densest SA1 in Sydney’s growth areas was 101 persons/ha. Nothing like this was seen in other cities.

Canberra’s outer growth areas are actually, on average, denser than the rest of Canberra (on a population weighted density measure):

The same was also true by a slim margin in both Perth and Adelaide, but they have relatively “suburban” densities for both growth and established areas. The growth areas of Sydney and Melbourne are more dense than Perth and Adelaide, but not compared to the rest of these cities as a whole. That’s probably got to do a lot with the large cities having dense inner suburbs.

So perhaps it is better to compare the urban growth areas with established areas a similar distance from city centres, which the following chart does (I’ve filtered out 5 km distance intervals without growth areas of at least 2000 population, and apologies for rather squashed Canberra label):

Technical note: for South East Queensland I’ve measured distances from the Brisbane CBD.

Outer growth areas were much more dense than the rest of each city at most distances from the city centre, except in Sydney.

One issue with the above chart is that different distance intervals have different populations – for example only 2,815 people were in growth SA1s at a distance of 45-50 km from the Perth CBD (just above my threshold of 2000), so the low population density of that interval is not hugely significant.

To get around that issue, I’ve calculated the overall population weighted density of non-growth SA1s that are within these 5 km distance intervals from the CBD (including all of SEQ beyond 15 km from the CBD). The following chart compares those calculations with the population weighted density of the growth areas overall:

This shows that urban growth areas are on average more dense than other parts of the city at similar distance from the CBD, except in South East Queensland. And remember, many of the growth SA1s will be partially built out, so their expected density is understated.

Are outer urban growth areas near rapid public transport?

The next chart shows the proportion of growth SA1 population by distance from the nearest train or busway station:

Technical notes: Distances are measured from the centroid of each SA1 to a point location defined for each station (sourced from August 2016 GTFS feeds). For oversized SA1s these distances might be a little longer than reality for the average resident. I haven’t excluded oversized SA1s because I want to see the population alignment of growth areas overall. Canberra excluded due to lack of separated rapid transit.

What sticks out clearly is that just over half the of the population in Perth’s outer growth areas was more than 5 km from a station in 2016. That is to say Perth has had the least alignment of outer urban growth areas and rapid public transport networks of all five cities. I’m not sure many urban planners would recommend such a strategy.

However, Perth’s MetroNet program appears to be trying to rectify this with new lines and stations proposed near urban growth areas such as Yanchep, Canning Vale East, Ellenbrook, Byford, and Karnup (Golden Bay). It will however take some time to get to them all built and open.

South East Queensland was second to Perth in terms of urban growth remote from stations, with a lot of the growth scattered rather than concentrated around rail corridors. I haven’t included the Gold Coast light rail in my proximity calculation – it runs at an average speed of 27 km/h (which is slower than most train networks) and doesn’t serve outer urban growth areas.

Sydney and Adelaide had the highest proximity of growth areas to stations.

Around half of Melbourne’s growth SA1s that were more than 5km from a train station were in Mernda and Doreen, a corridor in which a rail extension opened in 2018. Many of the rest are not in the current designated growth corridors, or are where future train stations are planned. Melbourne’s current designated urban growth corridors are fairly well aligned to its train network. From a transport perspective this is arguably a better kind of sprawl than what Perth has been experiencing.

Adelaide’s outer growth areas more than 5 km from a station were in Mount Barker (satellite town to the east) and Aldinga (on the far south coast of Adelaide).

Are the outer urban growth areas better aligned to rapid public transport stations than non-growth areas at the same distance from city centres? Here’s the chart as above but with an extra column for non-growth areas within the same distance intervals from the CBD (as before).

The populations of urban growth areas are less likely to be within a couple of kilometres of a station (most of that land probably has long-established urban development), but curiously in Adelaide and South East Queensland the urban growth areas are more likely to be within 5 kilometres of a station than the non-growth areas, suggesting better rapid public transport alignment than older urban growth areas. Older urban areas in other cities are more closely aligned to stations, particularly in Perth.

As an interesting aside, here’s a breakdown over the last three censuses of population by distance from train/busway stations (operational in 2016 – so it overstates 2006 and 2011 slightly):

You can really see how Perth has had much population growth remote from its rapid public transport network, which probably goes some way to explaining the overall 1.2% journey to work mode shift towards private transport between 2011 and 2016.

So how did people in these outer growth areas get to work?

Technical note: The figures here for “private transport” are for journeys involving only private transport modes – i.e. they exclude journeys involving both private and public transport (eg car+train).

While private transport (mostly car driver only journeys) dominated journeys to work from almost all growth areas, Melbourne and Sydney were the only cities to see significant numbers of residents in outer growth areas with private transport mode shares below 80%.

South East Queensland’s outer urban growth areas were the most reliant on private transport to get to work, with an overall private transport mode share of 93%, followed by Adelaide on 92%, Canberra on 91%, Perth on 90%, Melbourne on 86%, and Sydney on 81%.

Here’s how the growth area mode shares compare to other areas a similar distance from city centres (note: the Y-axis is not zero-based):

Significantly, the growth areas of Sydney and Melbourne had lower private transport mode shares of journeys to work than other parts of the city a similar distance out – even though they are generally further away from train or busway stations (as we saw above)! That’s not to say they didn’t drive themselves to a train station to get to work.

Similar to population density, here is a summary of growth areas compared to other areas in the same distance interval from the CBD:

There’s really not a huge amount of difference within cities. Sydney’s growth areas had a mode share 1.5% lower than non-growth areas, while Canberra’s growth areas had a mode share 2.5% higher.

What are motor vehicle ownership rates like in the outer growth areas?

My preferred measure is household motor vehicles per persons aged 18-84 (roughly people of driving age).

Motor vehicle ownership rates are generally very high across the growth areas – with the notable exceptions of Melbourne and Canberra where around a quarter of the growth area population had a motor vehicle ownership rate of less than 80 (although that is still pretty high!). (I explored this in more detail in an earlier post on Melbourne)

South East Queensland, Perth, and Adelaide outer urban growth areas had the highest motor vehicle ownership rates. Perth’s urban growth areas overall averaged 96.7 motor vehicles per persons aged 18-84 – pretty close to saturation.

How does motor vehicle ownership compare to established areas a similar distance from the city centre? The following chart compares motor vehicle ownership between urban growth and other areas at the same distance from the CBD (note: the Y-axis is not zero-based):

Motor vehicle ownership tends to increase with distance from the CBD, and in Sydney and South East Queensland the growth areas have higher ownership compared to non-growth areas. But the opposite is true in Melbourne, Perth and Canberra.

The population at each distance interval varies considerably, so here is a summary of the data across all distance intervals that have growth SA1s for each city:

The growth areas of Melbourne, Perth and Canberra had slightly lower motor vehicle ownership than other areas a similar distance from the city, while the opposite was true in other cities. That said, motor vehicle ownership rates are very high across all cities.

 

Finally, I’ll look at the relationships between these measures for growth areas (see another post for analysis for whole cities).

How does motor vehicle ownership relate to distance from stations?

Technical note: for scatter plots I’ve filtered out SA1s with less than 50 population as they are more likely to have outlier results (one person can change a measure by 2% or more).

Lower rates of motor vehicle ownership are generally only found close to train/busway stations (and are dominated by Melbourne examples), but close proximity to a station does not guarantee lower rates of motor vehicle ownership. Quite a few Adelaide SA1s are found the top middle part of the chart – these are all in Mount Barker which has frequent peak period express buses to the Adelaide CBD operating along the South East Freeway – which is similar to rapid transit although without a dedicated right of way.

How do journey to work mode shares relate to distance from stations?

Here’s a scatter plot of private transport mode shares of journeys to work and distance from train/busway station:

This shows that lower private transport mode shares are only generally seen within proximity of train or busway stations, and areas remote from stations are very likely to have high private transport mode shares. But also that proximity to a station does not guarantee lower private transport mode shares of journeys to work (particularly in SEQ).

Technical aside: You might have noticed that almost no SA1s report 99% private mode share. How can that be? The ABS make random adjustments to small figures to avoid identification of individuals which means you never see counts of 1 and 2 in their data. To get a mode share of 99% you’d need at least 300 journeys to work with “3” being non-private (or a similar but larger ratio). Very few SA1s have 300+ journeys to work, and even for over-sized SA1s, they are very unlikely to have only 3 or 4 non-private journeys to work. A mode share of 100% is much easier because you can get that no matter the total number of journeys.

How does population density relate to distances from train/busway stations?

Densities above 45 persons/ha were mostly only found within 5 km of stations, and almost entirely in Sydney and Melbourne. The highest densities were very close to train stations in Sydney. In the middle area of the chart you can see quite a few Perth SA1s that are around 30-40 persons/ha but remote from stations. These are all in the Ellenbrook area of Perth’s north-east, generating a lot of car traffic.

How does motor vehicle ownership relate to private transport mode shares of journeys work to work?

For interest, here is the relationship as a scatter plot:

There is certainly a relationship, but it’s not strong (r-squared = 0.22). Other factors are at play.

Conclusions

  • Perth and Canberra are seeing most of their population growth on the fringe, with Sydney, Adelaide, Melbourne, and South East Queensland seeing most of their population growth in established areas.
  • Growth areas in Sydney, Melbourne, and Canberra have higher than traditional urban densities, indeed Sydney and Canberra have a few very high density greenfield developments. Perth, Adelaide, and particularly South East Queensland have urban growth at relatively low densities. In fact, SEQ is the only major urban centre where growth areas are measured as less dense than non-growth areas at similar distances from the CBD.
  • Perth’s urban growth areas are largely remote from rapid transit stations, and this is likely contributing directly to very high and increasing rates of motor vehicle ownership and private transport mode shares. Melbourne’s current urban growth corridors are closely aligned to train stations (thanks to the opening of the Mernda line), and this is also largely true of Sydney and Adelaide.
  • Almost all outer urban growth areas had high rates of motor vehicle ownership. Overall, Melbourne, Perth, and Canberra’s outer urban growth areas had slightly lower rates of motor vehicle ownership compared to other areas at the same distance from the CBD. Only Sydney, Melbourne and Canberra have some growth areas with lower motor vehicle ownership and/or lower private transport mode shares of journeys to work – and these were all close to train or busway stations.

I hope you’ve found this at least half as interesting as I have.

For a similar and more detailed analysis around these topics, see this excellent 2013 BITRE research report on changes between 2001 and 2006.


What explains variations in journey to work mode shares between and within Australian cities?

Thu 6 December, 2018

Private and public transport journey to work mode shares vary considerably both between Australian cities and within them. Are these differences related to factors such as population density, motor vehicle ownership, employment density, proximity to train stations, proximity to busway stations, jobs within walking distance of homes, and distance from the city centre?

This posts sheds some light on those relationships for Australia’s six largest cities. I’m afraid it isn’t a short post (so get comfortable) but it’s fairly comprehensive (over 30 charts).

I should stress up front that a strong relationship between a certain factor and high or low mode shares does not imply causation. There are complex relationships between many of these factors, for example motor vehicle ownership rates are generally lower in areas of higher residential density (which I will also explore), and more factors beyond what I will explore here.

If you are interested in seeing spatial mode share patterns, see previous posts for Melbourne, Brisbane, and Sydney. You might also be interested in my analysis explaining the mode shifts between 2011 and 2016.

Population density

Higher population densities are commonly associated with higher public transport use. This stands to reason, as high density areas have more potential users per unit of area, but also higher density is likely to mean high land prices, which in turn increases the cost of residential parking. But higher public transport mode share can only happen if government’s invest in higher service levels, and this isn’t guaranteed to happen (although it often does, through pressures of overcrowding).

My preferred measure is population weighted density, which is the weighted average density of land parcels in a city, weighted by their population (this gets around problems of including sparsely populated urban land). I’ve measured it at census district (CD) geography for 2006 and Statistical Area Level 1 (SA1) geography for 2011 and 2016, using 2011 Significant Urban Area boundaries to define cities. The 2006 density figures are not perfectly comparable with 2011 and 2016 because CDs are slightly larger than SA1s, so the density values will be calculated as slightly smaller.

Here is the relationships at city level (the thin end of each worm is 2006 and the thick end 2016, with 2011 in the middle):

The relationship is very strong for Melbourne and Sydney over time. Between 2011 and 2016, Perth and Brisbane saw increased population density but reduced public transport mode share (mostly because of changes in the distribution of jobs between the centre and the suburbs).

Brisbane was a bit of an outlier in 2006 and 2011 with high public transport mode share relative to its lower population density.

Canberra is also perhaps a bit of an outlier, with much lower public transport mode share compared to similarly low density cities. This might be explained by the smaller total population, lower jobs density, and lack of rapid public transport services segregated from traffic.

But Canberra does have higher active transport mode share, so it’s worth doing the same analysis with private transport mode shares:

Brisbane was still an outlier in the relationship in 2006 and 2011, but Canberra is more in line with other data points.

Another interesting note is that Canberra went from being the least dense city in 2006 to the third most dense in 2016.

Drilling down to SA2 geography (SA2s are roughly the size of a suburb), here’s a chart showing all SA2s in all cities across the three census years (filtered for CDs and SA1s with at least 5 persons per hectare). I’ve animated it to highlight one city at a time so you can compare the cities, and I’ve used a log scale on the X-axis to spread out the data points (only the Sydney and Melbourne CBDs go off the chart to the right).

(if these animated GIF charts are not clear on your screen, you can click to enlarge the image, then use “back” to come back to this page).

You can see a fairly strong relationship, although it is very much a “cloud” rather than a tight relationship – there are other factors at play.

What I find interesting is that Sydney has had a lot of SA2s with population weighted densities around 50-100 but private mode shares over 55% (toward the upper-right part of the cloud of data points) – which are rare in all other cities. That’s a lot of traffic generation density, which cannot be great for road congestion. In a future post I might focus in on the outlier SA2s that are in the top right of these charts (can public transport do better in those places?).

In case you are wondering about the Brisbane SA2 with low density and low private transport mode share (middle left of chart) it is the Redland Islands where car-carrying ferries are essential to get off an island, and are counted as public transport in my methodology. The Canberra outlier in the bottom left is Acton (which is dominated by the Australian National University).

Employment density

I’ve calculated a weighted job density in the same way I’ve calculated population weighted density, but using Destination Zones (which can actually be quite large so it certainly isn’t perfect). Weighted job density is a weighted average of job densities of all destination zones, weighted by the number of jobs in each zone. In a sense it is the density at which the average person works

(technical notes: I’ve actually only counted jobs as people who travelled on census day and reported their mode(s) of travel. Unfortunately I only have 2006 data for Sydney and Melbourne)

This chart suggests a very strong relationship at the city level, with all cities either moving up and left (Adelaide, Perth and Brisbane) or down and right (Sydney, Melbourne, Canberra).

So is the relationship as strong when you break it down to the Destination Zone level? The next chart shows jobs density and private mode share for all destination zones for 2016. Note that there is a log scale on the x-axis, and Adelaide dots are drawn on top of other cities in the top left which explains why that dense cloud of dots appears mostly green.

There’s clearly a strong relationship, although again the data points form a large cloud rather than tightly bunch around a line, so other factors will be at play.

It’s also interesting to see that the blue Sydney dots are generally lower than other cities at all job densities. That is, Sydney generally has lower private transport mode shares than other cities, regardless of employment density.

Which leads us to the next view: the private transport mode shares for jobs in different density ranges in each city for 2011 and 2016.

(click to enlarge if the chart appears blurry)

You can see a fairly consistent relationship between weighted job density and mode shares across all cities in both 2011 and 2016.

At almost all job density ranges, Sydney had the lowest average private transport mode share, while Adelaide and Perth were generally the highest (data points are not shown when there are fewer than 5 destination zones at a density range for a city). This shows that something other than jobs density is impacting private transport mode shares in Sydney. Is it walking catchment, public transport quality & quantity, or something else?

For more on the relationship between job density and mode share, see this previous post.

Proximity to public transport

Trains generally provide the fastest and most punctual public transport services (being largely separated from road traffic and having longer stop spacing), and are the most common form of rapid transit in Australian cities. So you would expect higher public transport mode shares around train stations.

Here is a chart showing average journey to work public transport mode shares by home distance from train stations. It’s animated over the three census years, with a longer pause on 2016.

Technical note: A limitation here is that I’ve measured all census years against train stations that were operational in 2016 – so the 2006 and 2011 mode shares will be under-stated for the operational stations of those years. For example, in Melbourne the following stations opened between 2011 and 2016: Williams Landing, South Morang, Lynbrook, and Cardinia Road.

You can see that public transport shares went up between 2006 and 2011 in most cities at all distances from train stations. In both Perth and Brisbane there were new train lines opened between 2006 and 2011, which will explain some of that growth.

But if you watch carefully you will see public transport mode shares near train stations fell in both Brisbane and Perth between 2011 and 2016. That is, there was a mode shift away from public transport, even for people living close to train stations. As discussed previously, this is most likely related to there being only small jobs growth in the CBDs of those cities between 2011 and 2016, compared to suburban locations.

You can also see that public transport mode shares aren’t that much higher for areas near train stations in Adelaide (I’ll come back to that).

We can do the same for train mode shares (any journey involving train):

Again, Sydney’s train stations seem to have the biggest pulling power, while Adelaide’s have the least.

Busways are the other major form of rapid transit in Australian cities, with major lines in Brisbane, Sydney and Adelaide. Here is a chart of public transport mode share by distance from busway stations, excluding areas also within 1.5 km of a train station:

Note for Adelaide this data only considers suburban stations on the O-bahn, and not bus stops in the CBD. For Sydney all “T-Way” station are included, plus the four busway stations on the M2 motorway for which buses run into the CBD (but not the relatively short busway along Anzac Parade in Moore Park). Sydney’s north west T-Ways opened in 2007

Proximity to a busway station appears to influence public transport mode share strongly in Brisbane and Adelaide, where busways are mostly located in the inner and middle suburbs and cater for trips to the CBD. Sydney’s busway stations are in the “outer” western suburbs, feeding Blacktown, Parramatta, but also relatively long distance services to the Sydney CBD via the M2.

Curiously, public transport mode shares were higher in places between 3 and 5 km from busway stations in Sydney, compared to immediately adjacent areas. I’m not sure that I can explain that easily, but it suggests equally attractive public transport options exist away from busway and train stations.

The station proximity influence appears to extend around 1 km, which possibly reflects the fact that few busway stations have park and ride facilities, and are therefore more dependent on walking as an access mode (although cycling may be another station access mode).

Over time Sydney public transport mode share lifted at all distances from busway stations, while in Brisbane it rose in 2011 and then fell again in 2016, in line with other city mode shares.

So are busway stations similar to train stations in their impact on public transport mode share? To answer this I’ve segmented cities into areas near train stations, near busway stations, near both, and near neither. I’ve used 1.5 km as a proximity threshold that might represent an extended walking catchment.

In Sydney, train stations appear to have a much stronger influence on public transport mode shares than busway stations, but the opposite is true in Brisbane and Adelaide. This possibly reflects the much higher service frequencies on Adelaide and Brisbane busways compared to their trains, and the fact Sydney’s busway stations are so far from the CBD (and thus have fewer workers travelling to the CBD where public transport dominates mode share).

Also of note in this chart is that for areas more than 1.5 km from a train or busway station, Sydney had a much higher public transport mode share compared to the other cities. These areas will be served mostly by on-road buses, but also some ferries and one light rail line. Adelaide has the least difference between mode share for areas near and not-near train or busway stations.

We can do the a similar analysis for workplaces:

The most curious pattern here is Adelaide – where public transport mode share was highest for jobs between 1.5-2.5 kms from train stations. This distance band is dominated by the centre of the Adelaide CBD (the station being on the edge, arguably a “corner”), for which bus was the dominant public transport access mode. Also, there was no destination zone small enough near Adelaide central train station to register as 0 – 0.5 km away, and only one that is 0.5 – 1 km away (I use distances between station data points and destination zone centroids). So the results might look slightly different if smaller destination zones were drawn in the Adelaide CBD.

In all other cities there was a very strong relationship between train station proximity and public transport mode share, as you would expect. And Sydney again stands out with high public transport mode shares for workplaces more distant from train stations.

If you are wondering, the bump in Sydney at 2.5 to 3 km includes the Kensington / Randwick area which has high employment density and a strong bus connection to the central city (partly assisted by the Anzac Parade busway). And the relatively high figure for Melbourne at 1 – 1.5 km includes parts of Docklands, Parkville, Southbank, and St Kilda Road, which all have high tram service levels.

Unfortunately destination zones around busway stations are generally too large to provide meaningful insights so I’m not presenting such data.

Motor vehicle ownership

It will come as little surprise that there is a relationship between household motor vehicle ownership and journey to work mode shares.

Here’s a summary chart for each city for the 2006, 2011 and 2016 censuses:

There appears to be a fairly strong relationship between the two factors at city level.

Sydney and Melbourne have seen the largest mode shift away from private transport, but only Melbourne has also seen declining motor vehicle ownership rates.

Canberra saw only weak growth in motor vehicle ownership between 2011 and 2016, and at the same time there was a shift away from private transport (and a large increase in population weighted density).

Perth and Brisbane saw increasing private transport mode share and increasing motor vehicle ownership between 2011 and 2016.

Here’s a more detailed look at the relationship over time for Melbourne at SA2 geography:

The outliers on the upper left are generally less-wealthy middle-outer suburban areas (lower motor vehicle ownership but high private mode share), while the outliers to the lower-right are wealthy inner suburbs where people can afford to own plenty of motor vehicles, but they didn’t use them all to get to work.

In the bottom left of the chart are inner city SA2s with declining private mode share and declining motor vehicle ownership. For motor vehicle ownership rates around 70-80 (motor vehicles per persons aged 18-84), there are many SA2s with private mode shares that declined 2006 to 2016, but not significantly lowering motor vehicle ownership rates. That suggests that just because people own many motor vehicles, they don’t necessarily use them to drive them to work.

Here is the same data for Sydney:

There are many SA2s with motor vehicle ownership rates around 50 to 70 where the private mode shares are dropping faster than motor vehicle ownership. But there are also many areas with high private mode shares and increasing rates of motor vehicle ownership.

How do the other cities compare? Here are all the SA2s for all cities on the same chart, with alternating highlighted cities:

You can see big differences between the cities, but also that Brisbane and Perth have many SA2s with very high private mode share and rapidly increasing motor vehicle ownership (ie moving up and right, although it’s a little difficult to see with so many lines overlapping). Melbourne and Sydney have plenty of SA2s moving down and left – reducing motor vehicle ownership and declining private transport mode share (which may make some planners proud).

Of course there will be a relationship between motor vehicle ownership and where people choose to live and work. People working in the central city may prefer to live near train stations so they can avoid driving in congested traffic to expensive car parks. People who prefer not to drive might choose to live close to work and/or a frequent public transport line. People who are happy to drive to work in the suburbs might avoid higher priced real estate near train stations or the inner city.

As an aside, we can compare total household motor vehicles to the number of people driving to work, to estimate the proportion of household motor vehicles actually used in the journey to work. Here is Melbourne:

SA2s with a lower estimate are generally nearer the CBD, are wealthier areas, have reasonable public transport accessibility, and/or might be areas with a higher proportion of people not in the workforce (for whatever reason). The areas where the highest proportion of motor vehicles are required for the journey to work are relatively new outer suburbs on the fringe (perhaps suggesting forced car ownership), where adult workforce participation is probably high and public transport accessibility is lower.

The proportion of cars used in the journey to work declined on average in many parts of Melbourne. Given that motor vehicle ownership rates in Melbourne barely changed between 2011 and 2016, this probably represents people mode shifting, rather than people acquiring more motor vehicles even though they don’t need them to drive to work.

Jobs within walking distance of home

It stands to reason that people would be more likely to walk to work if there were more work opportunities within walking distance of their home.

For every SA1 I’ve measured how many jobs are approximately within 1 km as a notional walking catchment (measured as the sum of jobs in destination zones whose centroid are within 1 km of the centroid of each home SA1, so it is not perfect). Here’s the relationship with walking mode share:

(there are a lot of dots overlapping in the bottom left-corner and Adelaide dots have been drawn on top so try not to get thrown by that).

You don’t have to have a lot of nearby jobs to get a higher walking mode share, but if you do, you are very likely to get a high walking more share. The exceptions (many jobs, but low walking share) include many parts of Parramatta (Sydney), and areas separated from nearby jobs by water bodies or other topographical barriers (eg Kangaroo Point in Brisbane).

Workplace distance from the city centre

As was seen in a previous post, workplaces closer to city centres had much lower private transport mode shares, which is unsurprising as these are locations with generally the best public transport accessibility, high land values that can lead to higher car parking prices (which impact commuters who pay them), and often higher traffic congestion.

Here is a chart showing private transport mode share by workplace distance from the city centre. I’ve used faded lines to show 2011 and 2006 results (2006 only available for Sydney and Melbourne).

Here’s a chart that shows the mode shifts between 2011 and 2016:

Inner Melbourne had the biggest mode shifts away from private transport (particularly in Docklands that falls into the 1-2 km range, which saw significant employment and tram service growth), but Sydney had more consistent mode shifts across most distances from the city centre. Adelaide and Canberra saw mode shifts away from private transport in the inner city but towards private transport further out.

Brisbane and Perth saw – on average – a mode shift to private transport across almost all distances from the city centre, with the highest mode shift to private transport in Brisbane actually for the CBD itself(!).

Home distance from the city centre

There’s unquestionably a relationship here too, and it’s probably mostly driven by public transport service levels being roughly proportional to distance from the CBD, but also the proportion of the population who work in the CBD being much higher for homes nearer the CBD.

Sydney had the lowest average private transport mode share at all distances up to 20 km from the CBD, followed by Melbourne and Brisbane, in line with overall mode shares.

The trends over time are also interesting. Brisbane saw mode shifts towards private transport at all distances more than 2 km from the city centre between 2011 and 2016. However there were not significant shifts for Perth outside the city centre – that is: modes shares by geography didn’t change very much. The mode shift away from public transport in Perth is best explained by the shift in jobs balance away from the city centre.

Here are public transport mode shares by home distance from city centres:

In most cities, public transport mode share peaked at a few kilometres from the city (as active transport has a higher mode in the central city).

Here are public transport mode shifts by distance from the city centre between 2011 and 2016:

The significant shift in central Melbourne is likely to be largely explained by the Free Tram Zone introduced in 2015. Outside of the city centre the mode shifts are surprisingly uniform across each city.

Here’s the same chart for 2006-2011, and you can clearly see the impact of the opening of the Mandurah railway line in Perth with significant mode shift beyond 30 km:

Curiously there was a massive shift to public transport for CBD residents in Melbourne (and this is before the free tram zone was introduced).

So which factors best explain the patterns in mode shares across cities?

What we’ve clearly seen is that higher public transport mode shares are seen for journeys to work…

  • to higher density workplaces
  • from areas of lower motor vehicle ownership
  • to workplaces closer to train stations
  • from higher density residential areas
  • from areas around train and busway stations
  • to and from areas closer to city centres (except from the central city where walking takes over)
  • from less wealthy areas (while I haven’t tested this directly, wealth seems to explain a lot of the outliers in the scatter plots)

I’ve listed these roughly in order of the strength of the relationships seen in the data, but I haven’t put them all in a regression model (yet, sorry).

Of course most of these factors are inter-related, so we cannot isolate causation factors. I’m going to run through many of them, because they are often interesting: (note I have sometimes used log scales)

Population density is roughly related to distance from the city centre:

Motor vehicle ownership has a strong relationship with population density (see this post for more analysis):

Motor vehicle ownership has a weaker relationships with distance from the city centre:

Motor vehicle ownership is related to home distance from train stations, except in Adelaide:

Technical note: For this chart (and some below) I’ve calculated average quantities for the variable on the Y axis, as there would otherwise there are too many data points on the chart and it becomes very hard to see the relationship (I would need to show all SA1s because SA2s are too large in terms of distance from stations). The downside is that these style of charts don’t indicate the strength of relationships.

Population weighted density is related to distance from train stations, especially in Melbourne and Sydney, but not at all in Adelaide:

There is a relationship – although not strong – between weighted job density and distance from city centres:

There’s some relationship between average weighted jobs density and distance from train stations, except in Adelaide:

Here’s the same data, but as a scatter plot with a point for each destination zone, scaled by the number of journeys to each destination zone, and a linear Y-axis:

Technical note: the X-axis appears green mostly because Adelaide data points are drawn on top of other cities, but those data points aren’t of much interest.

In most cities, destination zones with high jobs density (over 700 jobs/ha) were only found within 1 km of a train station – with the notable major exception of Adelaide (again!).

(If you are curious, the large Melbourne zone at 1.4 km from a train station and 659 jobs/ha is the Parkville hospital precinct – where incidentally a train station is currently under construction).

There is a relationship between motor vehicle ownership and proximity to busway stations, but it varies between cities:

But there’s not much relationship between population density and proximity to busway stations (except in the immediate vicinity of busway stations in Brisbane):

Final remarks: there’s something about Adelaide’s train network

A few key observations come through clearly about the catchments around Adelaide’s train stations:

  • In aggregate they do not have higher population density, unlike other cities.
  • In aggregate they do not have particularly high public transport mode shares, unlike other cities.
  • In aggregate they do not have lower rates of motor vehicle ownership, unlike other cities.
  • They do not include the area of highest job density in the CBD (a longer walk or transfer to tram or bus is required), unlike other cities.

Few cities have spare land corridors available for new at-grade rapid public transport lines, and so transport planners generally want to make maximum use of the ones they’ve got, before opting for expensive and/or disruptive tunnelling or viaducts solutions. It looks like Adelaide’s rail corridors are not reaching their people-moving potential.

By contrast, Adelaide’s “O-Bahn” busway does go into the job dense heart of the CBD and the busway station catchments do have higher public transport mode share and lower motor vehicle ownership. However they do not have higher population density, possibly because the stations are surrounded by car parks, green space, and one large shopping centre (Tea Tree Plaza).

Mode shares, population densities, and motor vehicle ownership rates would quite probably change significantly if Adelaide could address the fourth issue by building a train station near the centre of the CBD.

In fact, Auckland had a very similar problem with its previous main city station being located away from the centre of the CBD. They fixed that with Britomart station opening in 2003 and train patronage soon rose quite dramatically (off a very low base, and also helped by service upgrades, subsequent electrification, and many other investments).

Should Adelaide do the same? It would certainly not be cheap and you would have to weigh up the costs and benefits.


What might explain journey to work mode shifts in Australia’s largest cities?

Mon 28 May, 2018

[Updated 29 June 2018 with further analysis of parking levies and their impact]

Between 2011 and 2016, journey to work public transport mode shares went up significantly in Melbourne and Sydney but dropped significantly in Perth and Brisbane. Private transport mode shifts did the opposite. Can this be explained by the changing distribution of jobs within cities, or other factors such as changes in transport costs?

In a recent post focused on Brisbane I found that stronger growth in suburban jobs relative to central city jobs could explain around half of the city’s mode shift towards private transport, with other factors (mostly the changes in relative attractiveness of modes) explaining the rest.

So how is job distribution changing in other Australian cities? How much of the mode shifts can be attributed to changing job distribution and how much could be attributed to other factors like changes in transport costs, or increasing employment density?

(for details about how I define public, private and active transport, see the appendix in this post)

How is job distribution changing in Australian cities?

Here’s a view of the changing distribution of all jobs within each city by workplaces distance from the city centre.

(Unfortunately I only have 2006 data for Sydney and Melbourne)

The changes are relatively subtle, but if look at how the bands shift between years, you’ll see increasing centralisation in Sydney but a decentralisation in all other cities between 2011 and 2016.

The strongest decentralisation was in Brisbane and Perth, which also showed the biggest increases in private transport mode share.

However Melbourne saw both a slight decentralisation of jobs and a mode shift away from private transport between 2011 and 2016.

So we need to dig deeper to find out what’s going on here.

How does private mode share vary by distance from the city centre?

The following chart shows private transport mode shares by distance from the city centre for the last two or three censuses for each city. The darkest line for each city is for 2016, with lighter lines being previous years (I only have 2006 data for Melbourne and Sydney).

There’s a clear pattern in all cities that private mode shares are lower in areas closer to the city centre, with Sydney the lowest, followed by Melbourne, Brisbane, Perth, Adelaide, and Canberra (which is also the order of their population size).

Notably Sydney private mode share averaged lower than 90% out as far as 24km from the city centre, whereas Adelaide sees 90% mode shares as close as 2km from the city centre.

If you look carefully you can see that Brisbane increased private transport mode shares in the central city between 2011 and 2016, while private mode shares dropped or were stable in all other cities at most distances.

You can also see that the central city mode shifts away from private transport were largest in Melbourne, something I’ll come back to.

Here’s the same again but for public transport:

Sydney and Melbourne saw mode shifts to public transport at most distances from the city centre, unlike all other cities.

What mode shift can we attribute to changing job distributions?

A city’s mode share (measured by place of work) will be fundamentally impacted by two types of changes between censuses:

  • Changes in the volume of jobs in each SA2 – because different SA2s generally have different mode shares due to factors like proximity to the city centre and public transport access. If there is stronger jobs growth in areas that already had lower private mode shares, you would get a mode shift away from private transport, all other things being equal.
  • Changes in the mode share in each SA2 – because different modes became more or less attractive for commuters between census years. This might be due to changes in public transport service quality, transport infrastructure provision, and relative changes in the cost of public transport, private motoring, and commuter parking. It could also be influenced by broader demographic changes.

For each city I have calculated what the city-level private transport mode share would have been in 2016, had mode shares in each workplace SA2 remained exactly the same as 2011, but the job volumes in each SA2s had still changed. The city level mode shift due to SA2 volume changes is then the difference between this hypothetical 2016 mode share and the 2011 mode share. The remainder of the city-level mode shift between 2011 and 2016 results can then be attributed to mode shifts at the SA2 level.

Here’s a chart showing the mode shift impact of both volume changes at the SA2 level, and mode shifts at the SA2 level:

As we noted above, Sydney saw a slight trend to centralisation of jobs between 2011 and 2016, and it had the largest volume change attributed reduction in private mode share (-0.4%). However other factors were responsible for a further 2.5% of the mode shift away from private transport.

The story is similar in Melbourne but to a smaller magnitude in both aspects. Both of these cities also saw increasing inner city job density – which matters – and I’ll back come to that in a moment.

In Brisbane you can see that the total mode shift towards private transport was roughly equally attributable to SA2 volume changes and SA2 mode shifts (as I discussed in my earlier post).

Perth had an overall 1.3% mode shift to private transport, and the majority of this was due to significant jobs growth in the suburbs compared to the CBD (in fact, the SA2 with the largest jobs growth was Murdoch in the southern suburbs). But there were also other factors that led to a mode shift to private transport.

In Canberra – Queanbeyan, volume changes by themselves would have seen a mode shift to private transport, but other factors were larger and led to an overall mode shift away from private transport (although it is actually complicated because the 2011 census day was in a federal parliamentary sitting week, while 2016 was not).

Nothing much changed in Adelaide.

Next I’m going to explore what could be behind the mode shifts at SA2 level, in terms of job density and real transport costs.

Can increases in workplace density impact mode shares?

As discussed in my Brisbane analysis, if the relative attractiveness of modes hadn’t changed, you might still expect a mode shift to public transport in high density employment areas with increasing jobs numbers because you would expect the cost of parking provision to increase with increasing land use density (i.e. more competition for space).

Indeed, in Sydney and Melbourne a number of inner city SA2s became significantly more job dense between 2011 and 2016, and also saw mode shifts away from private transport:

(inspect this data in Tableau)

A similar thing happened in Civic (the main centre of Canberra).

But Adelaide and Perth saw both declining job density and declining private transport mode share, which suggests something else is at play.

Job density didn’t really go down in Brisbane – see my Brisbane post for an explanation (basically, ABS redrew the SA2 boundary along the Brisbane River).

Could changes in the real cost of transport be causing mode shifts?

The following chart shows the real change in urban transport fares in Australian cities since 2000, as measured by the ABS as part of the Consumer Price Index series (which unfortunately includes public transport, taxis, and “ride share” but is for a representative sample of journeys so hopefully mostly dominated by public transport fares):

The lines are somewhat saw-toothed because public transport fares generally only rise once a year, and become better value in real terms over the course of the following 12 months.

Many cities have seen above-CPI public transport fare increases at various times, most notably Brisbane in 2010-2014. Melbourne has had above CPI fare increases, but also reduced zone 1+2 fares in 2015 which lead to a reduction on the ABS measure (the fare reduction only really applied to people travelling across zones 1 and 2 – which roughly summarised means travel between the outer and inner suburbs). Brisbane fares peaked in 2014, which was followed by a freeze and then a large reduction in 2017.

By contrast, here is the (negative) growth in the cost of “private motoring” (which includes vehicles, fuel and maintenance):

Private motoring costs have declined in real terms since 2000, although they increased a little during the second half of 2017.

The next chart shows the change in ratio between the two costs. Urban transport fares have become less competitive than private motoring over time in all cities:

But if we are looking at changes between census figures, we should probably also look at cost changes between the times of each census. Here’s how prices changed in real terms between the September quarters of 2011 and 2016 (which cover the August census dates):

The real cost of private motoring dropped in all cities, but so did the real “average” cost of urban transport fares in Sydney and Melbourne (the Melbourne drop being mostly around large fare reductions for travel across zones 1 and 2).

The biggest differences in cost changes were in Brisbane and Perth (around 18%), which I think will go a fair way to explaining why these cities had the biggest shifts to private transport attributable to SA2 mode shifts.

Brisbane saw a rapid increase in public transport fares between 2011 and 2014 which is likely to have changed many commuting habits, but those habits may or may not have changed back when fares were subsequently reduced (e.g. if someone bought a car due to fare increases, they may not have subsequently sold their car when fares reduced). Perth certainly had less mode shift at the SA2 level compared to Brisbane, which might support this hypothesis.

What about changes in car parking costs?

The ABS CPI’s private motoring cost index does not include car parking costs – which would be difficult as they vary considerably with geography.

However we do know about central city car parking levies that governments charge in a bid to reduce road congestion and fund inner city transport initiatives. Sydney, Melbourne, and Perth apply levies to central city non-residential car parking spaces, and ultimately these levies will need to be recovered through parking prices.

I’ve calculated these levies in 2017 dollars (adjusting for inflation as measured in June quarters), and here’s how they have changed since 2000:

Melbourne increased its central city parking levy by 40% per space in 2014 (category 1), and created a new lower-priced levy area in some neighbouring areas to the north and south in 2015 (category 2, see map). This is likely to have contributed to the larger mode shifts away from private transport in the central city area of Melbourne compared to most other cities (particularly considering there were similar changes in average private motoring and urban transport fares in Melbourne between 2011 and 2016).

Sydney’s category 1 fee applies in the Sydney CBD area, Milsons Points and North Sydney. It was $2390 in 2017, and has only risen with indexation since 2009 (when it was doubled). A lower category 2 levy applies in the business districts centres of Bondi Junction, Chatswood, Parramatta, and St Leonards.

Perth has an annual licence fee per bay which ranged from $1039 to $1169 in 2017.  The Perth fee was increased by around 167% in 2010, and there were also above-inflation increases from 2014. The fee increased 63% in real terms between 2011 and 2016 for “long stay” spaces, and 69% for “tenant” spaces.

I am not aware of any such fees or levies in place in Brisbane or Adelaide (a proposal for Adelaide was voted down).

So how are CBD parking prices changing?

Unfortunately good data is a little hard to find, but this Colliers Car Parking White Paper provides “average daily rates” for CBDs for 2009-2015, and early bird rates for 2015. I expect most commuters would pay early bird rates – which average between 28% and 62% of daily rates depending on the city (quite some variation!). I’ve adjusted the pre-2015 figures for inflation to be in 2015 dollars:

In real terms, “average daily” parking costs have declined in Melbourne, rocketed up in Brisbane and Canberra, and moved less in Sydney and Perth. I don’t know whether these reflect trends in early bird prices. And we don’t know how prices changed between 2015 and the census year of 2016.

So how much are parking levies contributing to parking prices?

I have to make some assumptions (guesstimates) here. Regular weekdays represent about 60% of the days of the year. If we assume say 80% of the levy is recovered from weekday commuter parking (there generally being less demand for parking on weekends), we can calculate the average weekday commuter cost of the levy to be 27% of the Sydney early bird price, 25% of the Melbourne early bird price, and 15% of the Perth early bird price. Certainly not insignificant.

Here’s a summary of the levy and “average daily” price changes and mode shifts in the central city parking levy areas:

Changes 2011 to 2016
Parking levy area or CBD SA2 Levy real increase Average daily real price change (2011 to 2015) Private mode shift New private trips Private share of new trips
Perth 63% -5% -0.8% -60 -3%
Melbourne – category 1 40% -11% -5.3% 3200 5%
Melbourne – category 2 (new) n/a -6.4% 5315 30%
Sydney CBD 0% +1% -2.6% 6204 9%
Brisbane City SA2 n/a +64% +1.7% 3135 68%
Adelaide SA2 n/a -11% -1.5% 2567 35%
Canberra Civic SA2 n/a +71% -3.2% 746 30%

Firstly, “average daily” parking prices don’t seem to be following the changes in parking levies in Perth and Melbourne (category 1 area). Other factors influencing parking prices will include supply (influenced by competition for real estate and planning rules) and demand (influenced by employment density) with the market ultimately determining prices.

Car park operators appear to be absorbing the increased cost of the levy (although we don’t know the trends in early bird prices so we cannot be entirely sure). But that’s not to say that the levy hasn’t had any impact on prices – for example, the price reductions might have been larger if the levies had not increased.

Secondly, price changes do not appear to be correlated with mode shifts as you might expect (except Canberra). Brisbane prices increased dramatically, but so did private mode share! Price reductions in Perth, Adelaide, and Melbourne did not result in increased private transport shares.

Maybe other factors are driving mode shift away from private transport in those cities. Maybe early bird prices are trending differently to “average daily” prices. Maybe increased traffic congestion persuaded people to shift modes. Maybe there were significant price changes between 2015 and 2016. Maybe most existing public transport users were not aware of reductions in parking prices.

I don’t know what happened to parking prices in the new category 2 areas of Melbourne but there was a large mode shift away from private transport (-6.4%), and they may well be linked. Indeed, Infrastructure Victoria has recently recommended the category 2 area be expanded to include the inner-eastern suburbs of Richmond, South Yarra, Windsor and Prahran. And the Grattan Institute has recommended increasing the levy to match Sydney’s rates.

Curiously, when I look at City of Melbourne Census of Land Use and Employment (CLUE) data, the category 1 area (approximated with CLUE areas) had an increase of only around 367 non-residential parking bays between 2011-12 and 2015-16 (a four year period), a lot less than the additional 3200 private trips, which might suggest increased average occupancy.

Also, it is likely that a significant portion of people who drive to city centres are not paying for their parking costs (eg employer provided car parking). Employers may simply be absorbing price increases.

For more interesting discussion and research about car parking in the City of Melbourne, see a recent discussion paper and background report prepared by Dr Elizabeth Taylor.

Did changes in population distribution impact mode shares?

While this post has been focused on changes by workplace location, it is possible to separate the overall mode shifts into the two components by home location. Here are the results:

In Sydney, Melbourne, and Canberra, stronger population growth in areas that already had low private mode shares in 2011 made a small contribution to overall mode shifts away from private transport. These cities have all seen densifying population in inner city areas better served by public transport.

The distribution of population growth in Perth and Brisbane had a small effect in the opposite direction.

And again, nothing much changed in Adelaide.

What about active transport?

Cycling-only mode share was pretty stable in most cities (except Canberra up 0.2%). Walking-only mode share declined in Sydney (-0.2%), Brisbane (-0.3%), Adelaide (-0.4%), Perth (-0.3%) but was steady in Melbourne and increased in Canberra (+0.2%). So Canberra has the biggest shift to active transport.

Can you summarise all that?

If your head is spinning with all that information, here’s a summary of what some of the major factors could be in each city between 2011 and 2016. I say “could be” because I’ve not looked at every possible factor influencing mode share.

Sydney: the 2.9% mode shift away from private transport was probably mostly to do with increasing job density in employment centres (more on that in my next post), but was also partly by a shift to more centralised jobs, and increasing population density in places well served by public transport.

Melbourne: The 1.8% mode shift away from private transport probably had a fair bit to do with increasing central city job density, the significant spatial expansion of the central city parking levy area and rates (although we don’t know if early bird prices also rose), a reduction in some public transport fares, and strong population growth in areas well served by public transport.

Brisbane: The 1.9% mode shift towards private transport appears roughly half about the decentralisation of jobs, and half the reduced attractiveness of public transport – particularly following significant fare rises between 2010 and 2014, and possibly/arguably declines in service quality.

Perth: The 1.2% mode shift towards private transport was probably mostly due to a decentralisation of jobs, and partly due to public transport becoming less cost competitive with private transport (despite an increase in the central city parking levy). Urban sprawl is probably also a factor.

Adelaide: The 0.2% mode shift to private transport is probably mostly due to public transport becoming less cost competitive with private transport. Changes in job and population distribution, and employment density do not appear to have had a significant impact.

Canberra:  The 1.0% mode shift away from private transport was probably the result of competing forces of higher jobs growth in car-dominated workplace areas with increasing job density in dense employment centres, increasing central city parking prices, higher population growth in areas better served by public transport (and possibly cycling facilities), and also the fact census 2016 was not a parliamentary sitting week while 2011 was (so really, it’s hard to be too sure!).

You might want to add your own views about changes in the service quality of public transport and cycling infrastructure in each city. I also haven’t looked at the impact of major new public transport infrastructure and service initiatives (such as the opening of new train stations), which we know does impact mode shares at a local level (maybe that’s for a future post).

I hope you found this interesting. My next post will look at suburban employment centres, and their role in changing mode shares in cities.


Can Airbnb explain falls in dwelling occupancy in Melbourne and Sydney?

Thu 10 May, 2018

According to census data, private dwelling occupancy has been declining in most Australian cities (refer my earlier post on the topic). Could an increase in private dwellings dedicated to Airbnb rental – but vacant on census night (a Tuesday in winter) – explain much of this decline? Let’s look at the data.

Firstly here’s a reminder of private dwelling occupancy trends in Australia’s 16 largest cities.

Dwelling occupancy rates declined in all large cities (but rose in some smaller urban areas, particularly on Central (NSW), Sunshine, and Gold Coasts).

The ABS have advised me that their field officers would have no way of telling whether a dwelling is on Airbnb, and would therefore count them as private dwellings. So vacant Airbnb dwellings could account for unoccupied private dwellings.

How many of the additional unoccupied dwellings might be dedicated to Airbnb?

The fantastic site Inside Airbnb provides data scraped from the Airbnb website about listings in various cities. I’ve used available data extracted on 4 September 2016 for Melbourne and 4 December 2016 for Sydney (the closest data sets available to the August 2016 census). I’ve then filtered for entire home/apartment listings that had more than 90 days availability in the 12 months ahead and had been reviewed at least once in the last six months, to estimate the number of “active and dedicated” Airbnb dwellings. Definitely just an estimate.

For the area for which Melbourne Airbnb data is available (I’ve approximated Inside Airbnb’s unpublished boundary as SA3s with any listings in 2016) these Airbnb dwellings account for 0.19% of total private dwellings. For that same area, dwelling occupancy dropped 0.71% from 92.61% to 91.90% between 2011 and 2016.

According to a Melbourne University study using data from commercial Airbnb data site AIRDNA, around 62% of entire home/apartment Melbourne listings (that were not blocked out by owners) were unoccupied on Saturday 27 August 2016.

I’m guessing the Airbnb vacancy rate might have been higher on census night (a Tuesday). If say 70% of the Airbnb dwellings were empty on census night (just a guess), then they would account for 0.09% out of the 0.71% decrease in dwelling occupancy in Melbourne between 2011 and 2016, which is about 19%.  Note: Airbnb barely existed in Melbourne in 2011 – there were only 161 Airbnb listings.

If somewhere between 60% to 80% of active/dedicated Airbnb properties were vacant, then they might explain between 16% and 21% the decline in dwelling occupancy in Melbourne.

For the equivalent area of Sydney, these Airbnb dwellings account for 0.22% of private dwellings, and there was a drop in private dwelling occupancy of 0.58% between 2011 and 2016. If somewhere between 60% to 80% of active/dedicated Airbnb properties were vacant, then they would explain between 23% and 30% of the decline in overall dwelling occupancy.

However I must stress these are rough estimates and might be over or under the actuals for several reasons:

  • It’s possible that some of these “entire home/apartment” listings are not counted by the ABS as dwellings (eg granny flats or segmented buildings that don’t have separate addresses) which would lead to over-estimates.
  • Some Airbnb listings that have less than 90 days availability in the 12 months ahead might just be very popular – leading to underestimates (my guess is that is unlikely).
  • The Sydney figure might be an overestimate because the Airbnb data was extracted three months after the census, and the total number of Airbnb listings almost doubled in 2016.
  • The actual Airbnb vacancy rate on census night might not have been in the 60-80% range.
  • I don’t know exactly where the city boundary was drawn for the Inside Airbnb data, but my approximation is more likely to be larger – which would lead to slight underestimates (probably very slight as the differences would be in peri-urban areas with few dwellings).
  • There may be other reasons – please comment.

That said, it looks like Airbnb might explain somewhere in the order of a fifth of the drop in private dwelling occupancy in Melbourne and Sydney between 2011 and 2016. Certainly not all of it, but probably not none of it either.

What proportion of dwellings are dedicated to Airbnb in different parts of Melbourne and Sydney?

Here’s a map of active/dedicated Airbnb dwellings (as per filter above) as a percentage of total dwellings in Melbourne at SA2 level:

It maxes out at 2.5% in the Melbourne CBD (that’s 1 in 40 dwellings), followed by Southbank and Fitzroy at around 1.9%. Mount Dandenong – Olinda is the orange patch to the east which measures 1.3%. View the data in Tableau.

As you would expect, Airbnb properties appear to be more prevalent around the inner city and tourist areas (eg St Kilda and the Dandenongs). These are also the areas with generally lower dwelling occupancy, and certainly some of the unoccupied dwellings in the census will be Airbnb dwellings.

Here is Sydney:

The highest rates are 2.6% around Bondi Beach, and 2.5% at Avalon – Palm Beach. Manly comes in at 1.5%, Surry Hills is 1.5%, Potts Point – Woolloomooloo is 1.4%, while the CBD area is 1.3%. Again, you can explore this Airbnb data in Tableau.

Could Airbnb properties explain the spatial differences in dwelling occupancy?

Here’s a plot of dwelling occupancy and Airbnb percentages for SA2s in Melbourne and Sydney.

I’ve done a linear regression on each city, and while the relationships are significant, they are not strong, and the correlation coefficients are -4.9 in Sydney and -1.5 in Melbourne. The signs are as expected (ie more Airbnb, lower occupancy), but the magnitudes are much higher than would be the case if Airbnb was the main explanation for lower dwelling occupancy (otherwise they would be around -1 or smaller). Which essentially means Airbnb presence is correlating with other drivers that would explain lower dwelling occupancy.

Indeed inner city and tourist areas had lower dwelling occupancy in both 2006 (when Airbnb didn’t exist) and 2011 (when Airbnb only had a tiny presence in Australia).

Therefore I think we can conclude Airbnb properties are more prevalent in areas where dwelling occupancy is lower for other reasons – one of which is likely to be popular places for visitors. Unoccupied Airbnb properties are almost certainly part of the pattern, but cannot explain the majority of the decrease in dwelling occupancy.

Can you do those Airbnb maps at higher resolution?

It’s getting a bit beyond the topic of transport, but yes I can go down to SA1 level. It’s not particularly important, but certainly interesting.

A disclaimer: Airbnb introduce randomised errors on property locations of up to 150 metres, so there will be some mis-attribution of properties to SA1s, but hopefully not too much. Also, I’m still only counting properties that match the above criteria.

Here’s inner Melbourne (note: different colour scale to last map):

Airbnb maxes out at 11% for three city blocks around Swanston Street and Collins Street, plus a large SA1 in East Melbourne that actually only contains a small residential area close to the CBD (including an apartment tower at 279 Wellington Parade). There’s also an SA1 in Southbank behind Crown Casino that is 10% Airbnb.

Curiously, there were only about 6 Airbnb listings in the New Quay apartments in Docklands that had 65-70% occupancy (refer earlier post), so Airbnb is definitely not to blame of the low occupancy of those towers.

Outside the central city, other hot spots are St Kilda around Acland Street (6%), just east of South Yarra Station (6%), and a patch of Olinda (5%). Explore in Tableau.

Here’s Sydney:

It tops out at 11% in the southern part of the CBD, with 10% in part of Pyrmont and 7% in pockets near Manly and Bondi Beach. Other hotspots include Coogee (6%) and Whale Beach further north (5%). Explore in Tableau.


How did the journey to work change in Brisbane between 2011 and 2016?

Wed 25 April, 2018

Between 2011 and 2016, Greater Brisbane saw a 2% mode shift towards private motorised transport for journeys to work, the largest such shift of all large Australian cities. Was it to do with where jobs growth happened, or because public transport became less attractive over that time?

This post takes a more detailed look at the spatial changes in private transport mode shares, and then examines the relative impact on spatial variations in jobs growth compared to other factors.

Greater Brisbane main mode shares

Firstly for reference, here are the Brisbane Greater Capital City Statistical Area main mode shares and shifts for 2011 and 2016, measured by place of enumeration and place of work:

2011 2016 Change
Private Place of enumeration 80.0% 81.9% +1.9%
Place of work 79.1% 81.1% +2.0%
Public Place of enumeration 15.1% 13.5% -1.6%
Place of work 15.9% 14.2% -1.7%
Active Place of enumeration 4.9% 4.6% -0.3%
Place of work 5.0% 4.7% -0.3%

More information about main mode definitions and data in general is available at the appendix at the end of this post.

Mode shares and shifts by home location

Here are private transport mode shares by home location for 2006, 2011, and 2016:

(you might need to click on these charts to see them larger and more clearly)

You can see lower private mode shares around the central city and to some extent along the rail lines. In case you are wondering, the Redcliffe Peninsula railway opened in October 2016 – after the August 2016 census.

The changes between years are a little difficult to make out on the map above, so here are the mode shifts to private transport by home location at SA2 level:

Mode shifts to private transport can be seen over most parts of Brisbane, with the biggest being Auchenflower (+6%), Lawnton (+6%), Toowong (+5%), Norman Park (+5%), Strathpine – Brendale (+5%), Keperra (+5%), and Sandgate – Shorncliffe (+5%). Many of the large mode shifts to private transport were actually seen around the train network.

The Redland Islands area had a larger shift to public transport – but keep in mind this will include use of car ferries.

Here’s a map showing the mode split of net new trips by home SA2:

There were a lot of new trips from outer growth areas in the north, west and south, and the vast majority of these trips were by private transport (although the southern growth area of Springfield Lakes, where a rail line opened in 2010, had a relatively high 15% of new trips by public transport). Private transport mode shares of new new trips were also high in middle and most inner suburbs (unlike inner Melbourne).

To sum all that up, here are the changes in trip volumes by main mode and home distance from the CBD:

Private transport dominated most new trips, and there were net declines in public transport trips beyond 2 km from the CBD.

Here’s a look at the main mode split over time, by distance from the CBD:

Brisbane achieved significant mode shift away from private transport between 2006 and 2011, but that was pretty much reversed between 2011 and 2016.

Private transport mode shares dropped in 2011 but pretty much returned to 2006 values in 2016. On average, only the city centre saw a mode shift away from private transport between 2011 and 2016, and that’s only a tiny fraction of the Brisbane’s population.

Mode shares and shifts by work location

Here are workplace private transport mode shares for 2011 and 2016:

(more areas are coloured in 2016 because they reached my minimum density threshold of 4 jobs per hectare at destination zone level for inclusion on the map)

Low private mode share is only really seen around the city centre. Some lower mode share areas further out include St Lucia (UQ campus, 52% in 2016) and Nundah (74%), but most of the suburban jobs are dominated by private transport.

Here are the mode shifts by workplace location:

The biggest mode shifts to private transport were to workplaces in Wooloowin – Lutwyche (+7%), Spring Hill (just north of the CBD, +5%) and Jindalee – Mount Ommaney (+5%). The biggest shifts away from private transport were in Newstead – Bowen Hills (-6%), St Lucia (-4%, which includes the University of Queensland main campus), and West End (-3%).

Notably, the job rich Brisbane CBD had a 2% shift to private transport (with 3,135 more private transport trips in 2016).

Here’s a map of the net new jobs and their main mode splits:

And a zoom in on the inner city to separate the overlapping pie charts:

The SA2 with the biggest jobs growth was “Brisbane City” (covering the CBD) with 4584 new jobs – with 68% of this net increase attributable to private transport. North Lanes – Mango Hill in the northern suburbs was not far behind (4472 new jobs at 96% by private transport), followed by Newstead – Bowen Hills (4266 new jobs at 49% private transport) and Brisbane Airport (4197 new jobs at 95% private transport).

The distribution of jobs growth was not heavily concentrated in central Brisbane – in stark contrast to Melbourne where the central city jobs growth was much more signficant.

Here’s a clearer view of new jobs by workplace distance from the city centre and main mode:

At all distances from the CBD, private transport new trips outnumbered active and public transport new trips (and there was a decline in public transport trips to the very city centre). The vast majority of net new trips were to workplaces more than 4 km from the city centre, and by private transport.

So why was there an overall 2% mode shift to private transport?

The relative lack of jobs growth in the public transport rich city centre is very likely to have contributed to the mode shift to private transport. The vast majority of new jobs were in the suburbs where public transport is significantly less competitive (relative to the CBD).

Others will point to factors that have made public transport less attractive relative to private transport, including problems on the train network, extensive new motorway infrastructure, and public transport fares growing around twice the rate of inflation after 2010.

There was very rapid growth in fares between 2010 and 2015, but then fares were frozen in 2016 and substantially reduced in 2017:

Looking at people working in Greater Brisbane (Greater Capital City Statistical Area), there were 94,055 new private transport commutes, just 246 new public transport commutes, and 2,506 new active transport commutes. So around 97% of net new trips in 2016 were by private transport, much higher than the 2011 baseline private transport mode share of 79% of trips (measured for workplaces in Greater Brisbane), hence the overall 2% mode shift.

Looking at people living in Greater Brisbane, there were 61,557 new private transport commutes, a net reduction of 6,069 public transport commutes, and a net reduction of 54 active transport commutes. Thus every new commute was accounted for by private transport, and further to this there was mode shift away from active and public transport.

So how much of the mode shift can be explained by spatial changes in jobs distribution? If mode shares in each workplace SA2 had not changed between 2011 and 2016 then city level mode shares would be influenced only by spatial variations in jobs growth.

I’ve done the calculations at SA2 geography: if place of work mode shares in Brisbane had not changed between 2011 and 2016 (but volumes had), then the overall private transport mode share would have increased only 1.0% in 2016 (essentially because of higher jobs growth in the suburbs compared to the centre).

Actual private mode share increased by 2.0% (measured by place of work).

So this suggests only half of the mode shift can be explained the spatial variations in jobs growth. The other half will be explained by other factors, particularly changes in the relative attractiveness of modes.

Changes in the relative attractiveness of modes will include public transport service quality, public transport fares, fuel prices, toll prices, and infrastructure provision for private and active transport. Car ownership will undoubtedly be a factor, but I suspect many ownership decisions will be influenced by workplace locations and relative modal attractiveness. Other factors might include changes in real incomes, demographic changes, changes in employment density, and the locations of population growth. I’ll explore the last two in more detail.

What about the relationship between job density and mode share?

You could argue that if general public transport “attractiveness” had not changed, you could still expect a mode shift towards public transport in areas with both high and increasing job density, as car parking might struggle to grow at the same rate as jobs growth (as the land becomes increasingly valuable/scarce). This might particularly be the case in the city centre.

I’ve calculated weighted job density for each SA2 – that is, the average density of destination zones in the SA2, weighted by the number of jobs in each zone (similar to population weighted density, so that large areas within SA2s that house few jobs make little contribution to such scores).

Here’s how weighted job density and workplace private mode share changed in Brisbane for higher density SA2s:

While there is some relationship between job density and private mode share overall, there wasn’t a consistent negative correlation between changes in those values. If there was, you would expect all lines on the chart to be on a similar diagonal orientation (upper left – lower right).

South Brisbane and Upper Mount Gravatt saw increased density but little change in private mode share. Chermside, Auchenflower, and Woolloongabba (which incidentally is at the southern end of the Clem 7 motorway) saw increased job density but also increased private transport mode share (the opposite effect of what you might expect). Spring Hill had only a small drop in job density but a large increase in private mode share.

Newstead – Bowen Hills had the largest shift away from private transport, and also one of the largest increases in job density

You might be wondering how the Brisbane City SA2 (which includes the CBD) can have had an increase in total jobs, but a slight decline in weighted jobs density. It turns out that the 2016 SA2 boundary goes further into the Brisbane River than the 2011 boundary. Here’s a map generated on the ABS website, where blue lines are the 2011 boundaries and red the 2016 boundaries:

If you discounted the increase in area, you might expect a slight increase in job density (about 4% in unweighted average density) to result in a small mode shift away from private transport, quite the opposite of what actually happened. If increasing job density by itself might have pushed a mode shift away from private transport, it appears it was overpowered by factors working in the opposite direction.

The Brisbane City SA2 accounted for 12.5% of Brisbane’s jobs so its mode split impacts more than most on overall city mode shares.

So what might be the stand-alone impact of increased job density in the city centre on private mode share? It’s very hard to quantify. I can certainly look at other city centres, but there will be so many factors at play in those cities that it would be almost impossible to isolate the impact.

But as a rough stab, had Brisbane City SA2’s private mode share increased from 29.0% to 29.5% (instead of 30.6%), and all other things were the same, then the overall Brisbane private mode share would have been 0.14% lower.

While the actual impact is uncertain, it would only increase the influence of the “other factors” that are responsible for at least half of the 2% mode share towards private transport.

And what about the spatial distribution of population growth?

All other things being equal, if population growth had disproportionately occurred in places with high private transport mode share (eg the middle and outer suburbs), you might expect a mode shift to private transport. However I don’t think this was significant in Brisbane as there has also been inner city population growth.

Indeed, if the home-based private transport mode share of each SA2 had not changed between 2011 and 2016 (but population numbers had), then the overall Brisbane private mode share (by place of enumeration) would have increased only 0.1% (rather than 1.9%). So the overall mode shift doesn’t seem to have a lot to do with where population growth happened.

So what are these effects other cities? I’ll cover that in an upcoming post.

Appendix: about the data

Here’s how I have defined “main mode”:

Private (motorised) transport any journey to work involving car, motorcycle, taxi, truck and/or “other”, but not involving any mode of public transport (train, tram, bus, or ferry)
Public transport any journey involving train, bus, tram, or ferry (journeys could also involve private or active transport modes)
Active tranport journeys by walking or cycling only

I have extracted data from the ABS census for 2006, 2011, and 2016 for areas within the 2011 boundary of the Brisbane Significant Urban Area. The detailed maps are at the smallest available geography – Census Collector Districts (CD) for 2006 and Statistical Area Level 1 (SA1) for 2011 and 2016 for home locations, and Destination Zones (DZ) for workplaces in 2011 and 2016 (detailed workplace data is not readily available for 2006 for most cities). I’ve aggregated this data for distance from city centre calculations (filtered by 2011 Significant Urban Area boundaries), which means the small randomisations will have amplified slightly.

In 2011, a significant number of jobs were not assigned to a destination zone:

  • 3.8% of jobs were assigned to an SA2 but not a DZ – I’ve imputed these proportionately to the DZs in their SA2 based on modal volumes reported for each DZ (for want of something better).
  • 18,540 Queensland jobs (0.9%) were only known to be somewhere in Greater Brisbane.
  • 115,011 jobs (5.8%) were only known to be somewhere in Queensland (hopefully mostly outside Greater Brisbane!).

These special purpose codes are not present in the 2016 data – presumably the ABS did a much better job of coding jobs to DZs. It means that the volumes in 2011 may be slightly understated, and so growth between 2011 and 2016 might be slightly overstated.

I’ve also extracted the data at SA2 (Statistical Area Level 2) based on 2016 boundaries for the purposes of calculating mode shifts and changes in trip volumes at SA2 level (to avoid aggregating small random adjustments ABS applies). However this wasn’t possible for jobs where 2011 SA2s were split into smaller SA2s in 2016 – because some 2011 jobs were assigned an SA2 but not a DZ, so we cannot map those to a specific 2016 SA2 (I aggregated imputed DZ numbers to 2016 SA2 boundaries instead).

I also extracted data at the Brisbane Greater Capital City Statistical Area level, as noted (the boundary did not change between 2011 and 2016).

I have not counted jobs that were reported to have no fixed address in my place of work analysis. I’ve also excluded people who worked at home, did not go to work on census day, or did not provide information about their mode(s) of travel. These workers are also excluded from job density calculations.


Where are the unoccupied dwellings in Australian cities?

Sat 4 November, 2017

Over one million private dwellings in Australia were unoccupied on census night in 2016 – 11.2% of all private dwellings – up from 10.2% in 2011.

This raises many questions. Where are these unoccupied dwellings and where are they now more prevalent? What type of dwellings are more likely to be unoccupied? How long have these dwellings been unoccupied? Do we know why these dwellings are unoccupied?

This post will focus on dwelling occupancy by geography, dwelling types and trends over time. In a future post I hope look into those last two questions in more detail.

I’ve prepared data for sixteen Australian cities, with various maps in Tableau (you will need to zoom and pan to your city of interest).

Why am I blogging about dwelling occupancy on a transport blog? Well partly because I’m interested in urban issues, but also because land use is very relevant to transport. If dwelling occupancy rates in the inner and middle suburbs were higher, there would be more people living closer to jobs and activities who might be less reliant on private motorised transport for their daily travel.

If you’d like to read more around the associated policy issues, Professor Hal Pawson from UNSW has a good piece in The Conversation highlighting the increasing number of empty properties and spare bedrooms, and advocates  replacing stamp duty with a broad-based land tax to improve housing mobility. Also read Eryk Bagshaw in the Fairfax press, Jonathan Jackson in Finfeed, and a piece in Business Insider where the Commonwealth Bank state that 17% of recently built dwellings are left unoccupied (not sure how that was calculated).

What are the dwelling occupancy rates in Australian cities?

Here’s a chart showing private dwelling occupancy rates for sixteen Australian cities (using 2011 Significant Urban Area boundaries) from the last three censuses:

Note the y-axis only runs from 84% to 94%, so the changes are not massive. However a small change in dwelling occupancy can still have a large impact on housing prices (rental and sales).

The Sunshine and Central Coasts have the lowest occupancy, almost certainly explained by many holiday homes in those regions, although all three have been trending upwards. Curiously, the Gold Coast – Tweed Heads had a significant increase in occupancy between 2011 and 2016 to take it above Perth, Townsville, and Darwin.

Hobart and Cairns also had increased occupancy between 2011 and 2016, but all large cities declined between 2011 and 2016. Perth, Darwin and Townsville had big slides – quite possibly related to the downturn in the mining industry and slowing population growth (all three have seen slowing population growth in recent years after a boom period). Then again, if there are more fly-in-fly-out workers in a city you might expect dwelling occupancy on census night to go down as a portion of them will be away for work on census night.

How does dwelling occupancy in capital cities compare to the rest of the country?

Private dwelling occupancy is significantly lower outside the capital city areas. While the capital city areas contain 63% of all private dwellings, they only contain 51% of unoccupied private dwellings.

How does dwelling occupancy vary by dwelling type?

Here’s a chart of 2016 dwelling occupancy by Greater Capital City Statistical Areas and the most common dwelling types:

In many cities there is a strong correlation between housing type and occupancy, with separate houses having the highest occupancy rates, and multi-storey flats/apartments having the lowest. The pattern is strongest in Perth – perhaps reflecting reduced demand for apartment living following the end of the mining boom(?).

The data suggests higher density apartments are more likely to not be occupied on census night, but it doesn’t tell us why. Of course different dwelling types have different spatial distributions, so is it the dwelling type that drives the occupancy rates? I’ll come back to that shortly.

Where are the unoccupied dwellings?

Quite simply, here is a map showing the density (at SA2 geography) of unoccupied dwellings in Melbourne over time (you might need to click to enlarge to read more clearly):

(I’ve not shaded SA2s with less than 1 unoccupied dwelling per hectare. You can look at other cities in Tableau by zooming out and then in on another city).

You can see a fairly significant increase in the number of unoccupied dwellings in the inner and middle suburbs (at least at densities above 1 per hectare).

From a transport perspective – this isn’t great. If people lived in those dwellings rather than dwellings on the fringe of Melbourne, the transport task would be easier as there would be many more people living closer to jobs and other destinations with non-car modes being more competitive.

But these areas with a relatively high density of unoccupied dwellings are also areas with a high density of dwellings in general. The density of unoccupied dwellings has risen in the same places where total dwelling density has risen:

(see in Tableau – you may need to change the geography type)

Given you would expect a small percentage of dwellings to be unoccupied for good reasons (eg resident temporarily absent, or property on the market), it makes sense that the density of unoccupied dwellings has gone up with total dwelling density.

But a decrease in the dwelling occupancy rate requires the number of unoccupied dwellings to be growing at a faster rate than the total number of dwellings. We already know that is happening at the city level through declining occupancy rates, so how does that look inside cities?

How does dwelling occupancy vary across Melbourne?

Here’s a map of dwelling occupancy in Melbourne and Geelong at CD/SA1 level geography:

(see also in Tableau)

You can see very clearly that occupancy is lowest on Mornington Peninsula beaches to the south – which almost certainly reflects empty holiday homes on census night (a Tuesday night in winter).

In fact, I’ve created a map of dwelling occupancy at SA2 level for all of Australia, and you can see many coastal holiday areas around Melbourne (and other cities) with low occupancy (with Lorne – Anglesea at 32% and Phillip Island at 40%):

The previous Melbourne map at CD/SA1 level is very detailed and so it’s not easy to see the overall trends. Also, apart from the Mornington Peninsula, occupancy rates are almost all above 80%.

So here is a zoomed-in map with a different (narrower) colour scale, with data aggregated at SA2 level (also in Tableau):

Things become much clearer.

The highest dwelling occupancy is generally on the fringe of Melbourne.

Apart from holiday home areas, the lowest occupancy in 2016 was concentrated in wealthier inner suburbs, including Toorak at 83% and South Yarra west at 84%. This was closely followed by the CBD, Docklands, East Melbourne, Southbank, and Albert Park between 84% and 86%. These areas have all had declining occupancy since 2011.

It can be a little difficult to see the changes in occupancy rates, so here is a non-animated map the change in dwelling occupancy rates between 2006 and 2016 (also in Tableau):

There are at least small declines in most parts of Melbourne. The biggest decline was 7% in Bundoora North (with lowest 2016 occupancy of 79% in these new units in University Hill ), followed by 5% in Doncaster (lowest around Doncaster Hill where there are new apartments, perhaps too new to be occupied on census night?), 4% in South Yarra East (lowest in the new apartments around South Yarra Station, again possibly because some are very new) and Prahran – Windsor.

Curiously, Docklands dwelling occupancy increased by 9% from 75% to 85% (rounding means that those numbers don’t perfectly add). Perhaps there were many new yet-to-be-occupied dwellings in 2006? For reference, Dockland’s 2011 occupancy was 84%, only slightly below the 2016 level.

The outer growth areas are a mixed bag of increases and decreases. This possibly depends again on how many brand new but not yet occupied dwellings there were in 2006 and 2016.

What are the dwelling occupancy patterns in other cities?

Sydney

You can see lower occupancy around the CBD, North Sydney, Manly, and the northern beaches, and higher occupancy in the western suburbs.

The largest declines are evident in the city centre and North Ryde – East Ryde:

Brisbane

Brisbane has some big declines to the north-east of the city centre, Rochedale – Burbank, Woodridge, Logan, and Leichhardt – One Mile. The Redland Islands in the east are presumably a popular place for holiday homes.

Perth

Low occupancy is evident around Mandurah in the south (a popular holiday home area). Lower occupancy has spread around the inner city, and beach-side suburbs of Scarborough, Cottesloe, Fremantle, and Rockingham (many of which are areas with higher concentrations of Airbnb properties).

The biggest declines were in Maylands, Victoria Park – Lathlain – Burswood, and South Lake – Cockburn Central. For the first two of these areas the decline was mostly in flats/units/apartments.

Adelaide

The lowest occupancy is on the south coast and in Glenelg. The biggest decline was in Fulham (-5%), followed by Payneham – Felixstowe (-4%):

[Canberra, Hobart and Darwin added 6 November 2017]

Canberra

Dwelling occupancy was lowest around Parliament House (the census was not during a sitting week in 2016), and highest in the outer northern and southern suburbs. The 2006 census was during a sitting week, so it’s little surprise that big dwelling occupancy reductions were seen around Capital Hill between 2006 and 2016.  There was also a 5% decline in Farrer and a 6% growth in Gungahlin between 2006 and 2016 (Gungahlin’s dwellings almost doubled between 2006 and 2011, so the 2006 result might reflect brand new dwellings awaiting occupants).

Hobart

Dwelling occupancy was lowest in central Hobart, with the biggest decline of 4% in Old Beach – Otago, but overall there was little change between 2006 and 2016 (average occupancy did drop slightly in 2011 though).

Darwin

Darwin dwelling occupancy was lowest in the city centre at 82% in 2016, while Howard Springs had 100% occupancy (in 2016). Declines are evident between 2006 and 2016 across most parts of Darwin.

Gold Coast

Here’s a map of 2016 occupancy at SA1 level, with the original broader colour scale:

You can see quite clearly that the beach-side areas have low occupancy, while the inland areas have much higher occupancy (some at 100%). Presumably many permanent residents cannot or choose not to compete with tourism for beach-side living.

Sunshine Coast

Similar patterns are evident on the Sunshine Coast, particularly around Noosa and Sunshine Beach in the north:

If you want to see other cities, move around Australia in Tableau for occupancy maps at CD/SA1 and SA2 geography (choose you year of interest), and occupancy change maps (at SA2 geography).

So are there lots of unoccupied inner city apartments in Melbourne?

Some commentators have spoken about many inner city apartments being unoccupied – perhaps through a glut or investors chasing capital gains and not interested rental incomes.

Here is dwelling occupancy in central Melbourne at SA1 geography for 2016, using the broader colour scale (also in Tableau):

There are quite a few pockets of very low occupancy, particularly areas shaded in yellows and greens. The average private dwelling occupancy for the City of Melbourne local government area was 87%, lower than the Greater Melbourne average of 91%.

The lowest occupancy is a block between Adderley, Spencer and Dudley Street in North Melbourne at 56%, which is probably related to the recent completion of an apartment tower not long before the census (from Google Street view we know it was under construction in April 2015 and completed by October 2016).

There are several patches of yellow  (65-70% occupancy) in the CBD, Docklands and Southbank.

But what about apartment towers? For that we need to drill down to mesh blocks – and thankfully 2016 census data is actually provided at this level.

Here’s a map showing dwelling occupancy of mesh blocks in the City of Melbourne (local government area) with at least 100 dwellings per hectare (as an arbitrary threshold for large apartment building – see the appendix for an example of this density):

(explore in Tableau)

Some notable low occupancy apartment towers include:

  • 48% for an apartment tower at 555 Flinders Street (Northbank Place Central Tower) between Spencer and King Street and the railway viaduct. It wasn’t brand new in 2016.
  • 47% in a block that includes the Melbourne ONE apartment tower, possibly because it was only just opened (as I write there are still apartments for sale)
  • 65% for one of the towers at New Quay, Docklands (which seems to include serviced apartments)
  • 66% for a tower at 28 Southgate Ave (corner City Road), and 67% for the Quay West tower next door (almost certainly popular places for Airbnb / serviced apartments).

Several of these towers include advertised serviced apartments, and I expect the towers would contain a mix of serviced apartments, owner-occupied apartments and rentals (regular and Airbnb). However ABS advises me that field officers do speak to building managers, and are therefore likely to not code serviced apartments as private dwellings.

That said, according to the 2016 census data there were only 11 non-private dwellings in Docklands that were classified as “Hotel, motel, bed and breakfast”, and zero non-private dwellings in the New Quay apartment towers.

I snapped this picture at 9pm on a Sunday in September 2017 of the apartments at New Quay (Docklands) that at the 2016 census had 65-70% occupancy:

Of course you wouldn’t expect lights to be on in all rooms in all occupied dwellings at 9pm on a particular Sunday, but I dare say it’s probably a time when fewer people would be out. It looks like a lot less than a quarter of rooms are lit. I know very few of these are on Airbnb (more on that in a future post!), but I don’t know how many are actually serviced apartments.

There’s huge variation in dwelling occupancy across these mesh blocks. So is the lower occupancy more concentrated in higher density areas? Here’s a scatter plot of all mesh blocks in the City of Melbourne by dwelling density and occupancy:

There’s not a strong relationship between density and occupancy. The variation in dwelling occupancy between mesh blocks will probably depend on a lot of local factors.

What about occupancy by dwelling type for the inner city?

(data points removed where dwelling counts were small, the isolated blue dot at the bottom is for Southbank).

There’s no evidence that flats / apartments have lower occupancy than other housing types in the central city. However there is evidence that inner city areas have relatively lower occupancy.

So how does the occupancy of apartment blocks of 4+ storeys vary across Melbourne?

Box Hill had the lowest apartment occupancy of 50% (perhaps some were brand new?), followed by Ringwood, Glen Waverley, and Brighton in the 70-75% range.  Croydon East, Templestowe , Seddon – Kingsville, Clayton, Carnegie, West Footscray, Braybrook and Frankston reported occupancy above 95%. The inner city areas were around 84-85% occupied, and these would make up the majority of such dwellings in Melbourne.

Apartments in blocks of 4+ storeys seem to have lower occupancy on average because most of them are located in the central city, which generally has lower dwelling occupancy.

Here’s a similar map (with a different colour scale) for dwelling occupancy of separate houses across the Melbourne region:

The lowest rates in metropolitan Melbourne are 82-83% in some inner city areas, while the urban growth shows up in pink and purple, mostly 94-96%.

Explore the 2016 occupancy rates at SA2 geography for different dwelling types for any part of Australia in Tableau. You can also view changes in occupancy rates since 2006 for separate houses, flats/units/apartments, and semi-detached/townhouses.

Why are there lower dwelling occupancy rates in the central city?

The census doesn’t answer this, and I’m not a housing expert, but I dare say there are plenty of plausible explanations:

  • Many dwellings are rented out on Airbnb (and/or other platforms) – but are not in high demand on a weeknight in mid-winter (more on that in this post).
  • Many dwellings are serviced apartments that are indistinguishable from regular private dwellings (in buildings with a mixture of dwelling use). ABS say they don’t count these as private dwellings, however they are not showing up as non-private dwellings.
  • Dwellings are more likely to occupied by executives who travel more frequently.
  • Dwellings might be second homes for people living outside the city.
  • Dwellings might be owned by employers for interstate staff visiting Melbourne.
  • Dwellings might be poorly constructed and uninhabitable (eg mould issues).
  • Investors who are not interested in rental income might deliberately leave properties vacant (something that is disputed).

But I’m just speculating.

What about dwelling occupancy in the centre of other cities?

Here’s a map of the Sydney CBD area at SA1 geography:

There are some very low occupancy rates in the north end of the CBD, but very high occupancy rates around Darling Harbour and Pyrmont.

Here’s central Brisbane:

Here are occupancy rates for different dwelling types for selected inner city SA2s in Sydney, Brisbane, Adelaide and Perth:

In all SA2s except Surrey Hills (Sydney) and South Brisbane, flats or apartments in 4+ storey blocks had the lowest dwelling occupancy in 2016. Only in Perth City SA2 (which is quite a bit larger than the CBD) is there a reasonably clear relationship between housing type and occupancy.

Summary of findings

Couldn’t be bothered reading all of the above, or forgot what you learnt? Here’s a summary of findings:

  • Dwelling occupancy, as measured by the census, has declined in most Australian cities between 2006 and 2016 (particularly larger cities).
  • Dwelling occupancy is generally very low in popular holiday home areas, but also relatively low in central city locations.
  • Dwelling occupancy is generally highest in outer suburban areas.
  • Higher density housing types generally have lower occupancy, but that is probably because they are more often found in inner city areas.
  • There are examples of low occupancy apartment towers in Melbourne, but there’s not a clear relationship between dwelling density and dwelling occupancy in central Melbourne.

In a future post I plan to look more at why properties might be unoccupied, and for how long they are unoccupied, drawing on Airbnb and water usage datasets.  I might also look at bedrooms and bedroom occupancy which is a whole other topic.

 

Appendix – About the census dwelling data

I’ve loaded census data about occupied and unoccupied private dwellings data into Tableau for 2006, 2011, and 2016 censuses for sixteen Australian cities at the CD (2006) / SA1 (2011,2016) level, which the smallest geography available for all censuses. I’ve mapped all these CDs and SA1s to boundaries of 2016 SA2s and 2011 Significant Urban Areas (as per my last post). Those mappings are unfortunately not perfect, particularly for 2006 CDs.

The ABS determine a private dwelling to be occupied if they have information to suggest someone was living in that dwelling on census night (eg a form was returned, or there was some evidence of occupation). Under this definition, unoccupied dwellings include those with usual residents temporarily absent, and those with no usual residents (vacant).

For my detailed maps I’ve only included CDs / SA1s with a density of 2 dwellings per hectare or more.

For reference, here is a Melbourne mesh block with 100 dwellings per hectare:

And here is a mesh block with 206 dwellings per hectare (note only a small part of mesh block footprint contains towers):