A look at Melbourne CBD transport

Sun 23 January, 2011

My last post looked at suburban employment areas, but what about the CBD? With the review of the City of Melbourne’s Transport Strategy, I’ve taken on a detailed analysis of transport to and from the CBD.

In this post I’ll look at questions like:

  • Do CBD commuters come from the inner or outer suburbs?
  • Do wealthy executive types snub public transport?
  • How does mode share vary between the sexes and young and old?
  • What impact are employer parking and driving subsidies having on mode choice?

I’m mostly focussing on the inner Melbourne CBD – using the ABS definition of “Melbourne – inner” SLA, which is essentially the Hoddle grid. However I’ve included Southbank or Docklands a couple of times, and there are also some comparisons with Sydney, Brisbane and Perth CBDs.

This is a long post, so grab a cuppa and get comfortable.

Where do the commuters come from?

According to the 2006 census, there were 137,853 commuter journeys into the CBD.

The first map shows the number of commuters from each SLA in Melbourne. The shading represents simple density of CBD commuters by area, which is not ideal because outer metro SLAs can be impacted by low average population density. At the same time, not all SLAs have the same population so some will always have large numbers (eg Manningham west). As always, click to zoom in.

The CBD attracted workers from all over Melbourne, but certainly with a high concentration from the inner suburbs.

To get around the density issue, I’ve drawn a map showing the percentage of workers from each SLA who work in the CBD, Southbank or Docklands:

You can see the percentage drops off fairly uniformly by distance. The CBD is not a major destination for most middle and outer suburban areas.

What modes of transport do commuters use? (by area)

Firstly a map showing the public transport mode share from each SLA (green = higher):

Public transport mode share was largely above 70% for much of Melbourne and indeed most surrounding areas.

A few low spots stick out:

  • Manningham west and east, serviced only by buses (that have recently been signficantly upgraded)
  • Northern parts of Boroondara and 52% and 55%. These wealthy areas are serviced by frequent trams and buses, although with a relatively slow trip in.
  • Rowville (Knox south) is at 57%, but bear in mind there were only 800 commuters from Rowville to the CBD (and I expect most of these would be park and ride train commuters). In fact, the catchment of the proposed Rowville rail line passes through three SLAs, with a total CBD commuter population of 4138. Allowing for catchments of other radial public transport lines in the SLAs, the CBD commuter catchment of the proposed Rowville line might be 2000-3000, or about 3 full trains. But of course a line would also be used for trips to other destinations (particularly Monash), and it would probably cause changes in travel patterns over time once built. I might look at this more in a future post. In the meantime you might want to read Alan Davies take, and a 2004 pre-feasibility study (here is a summary presentation).
  • Wealthy Brighton is well serviced by the Sandringham line, but only half used public transport to get to the CBD. There is no easy freeway connecting Brighton and the CBD, so why are they driving? I’ll come back to that.
  • The inner SLAs in Melbourne, Yarra and Port Phillip are slightly lower, probably due to a high rate of walking and cycling. More on that later too.

You can see a high PT mode share for the relatively small numbers of commuters from Geelong (around 800 in total). $4.3b is being spent on a regional rail link, that will separate regional trains from suburban trains. Regional trains from outside Melbourne seat less than 500 people, but because they run express through much of Melbourne they each consume probably around two all-stopping suburban train paths (which have a capacity of around 1000 each). I haven’t seen any debate about whether encouraging regional commuting by train into central Melbourne is worthwhile, though I’m sure people living in those areas appreciate the trains.

Next a map showing private transport mode share (red = higher):

Private transport mode share was highest for Manningham, northern Boroondara, Wyndham South (including Point Cook), Bayside, Rowville, and the outer northern fringes.

But a high car mode share may not be a huge issue if the number of car commuters is low. The next map shows the number of private transport commuter trips from each SLA, shaded by relative density:

Observations:

  • Like we saw in my last post for South Melbourne, there were large numbers of car commuters coming from the inner suburbs, particularly to the south-east. These areas are well connected to the CBD by public transport, and also quite wealthy. Is wealth a driver of higher car mode share? Read on.
  • Manningham west had a large number of car commuters (with a reasonable density). This area is entirely reliant on bus services, which have been upgraded considerably since 2006, with strong patronage growth resulting. In 2006, the last bus from the CBD on the Eastern Freeway – Doncaster Road route (307) was around 6:45pm. It’s now around midnight (on route 907 that replaced 307).
  • There were also a large number from Wyndham north-east (Werribee – Hoppers Crossing area) which is not shaded dark on the map due to low average population density. In 2006, peak train services on the Werribee line were often 20 minutes apart, and bus services only ran every 40 minutes. The train frequency has since increased to 6/hour but the (feeder) bus frequencies are still 40 minutes in peak periods.
  • Moonee Valley (Moonee Ponds-Essendon area) was a large contributor of cars, despite frequent trains and trams to the CBD. Not sure why that is, although Essendon is a relatively wealthy area.

Here is a another map of private transport commuters, except it is shaded by numbers rather than density. Manningham west stands out, but bear in mind it is one of the largest SLAs in Melbourne by population. You can see the outer western SLAs show up on this map also.

And for a flip side, here is where the public transport passengers were coming from (shaded by density):

There are large concentrations coming from the inner suburbs, but also the middle eastern suburbs which are well connected by trains. The Manningham west area had over 2000 public transport commuters to the CBD, many of which would have been on buses only.

Again, to get around the low population density problem, I’ve also drawn a similar map shaded by total numbers:

We saw low PT and car mode shares for the inner city. I haven’t drawn a map of walking mode share for the CBD but you can see public and private transport mode shares are low in the inner city, with walking likely to fill the gap. A map of walking mode share to any work destination is in another post.

The cycling figures are quite interesting. Next map shows the bicycle mode share to the CBD (any trip involving bicycle) (green=higher):

The figures are for Yarra north, Brunswick and Northcote are surprisingly high at 8-10%. Remember that the census is taken in winter (August). As I recall it wasn’t a rainy day. Bicycle mode share is also lower for commuters from the City of Melbourne itself. SLAs in grey lacked sufficient data.

Here are the total number of CBD bicycle commuters per SLA (shading by numbers, not density):

According to the data, people also rode from as far out as Frankston, Croydon, Ringwood and Sunbury! Census data is like that (as I recall, someone in Banyule claimed to have gone to work by ferry).

What modes did people use overall?

Here is a chart showing the overall mode split for all CBD workers:

Trains accounted for almost half of all CBD arrivals.

While buses accounted for only 2% of all CBD commuters, they were the only mode used by 32% in Manningham west, 11% in Kew, 9% in Camberwell north, 7% in Maribyrnong, and 5% in Altona.

Next chart shows mode split in a more simplified form:

Public Transport dominates, but still over a quarter came by car – including over 32,000 car drivers.

Public transport took 67% of motorised commuter trips into the CBD.

Active transport is at 8%, which probably represents those who live within walking distance of the CBD.

So how does Melbourne compare to other large Australian cities? The following chart compares Sydney, Brisbane, Perth and Melbourne CBDs. I’ve used the SLA that represents the inner core of business activity in each city to try to make in a reasonably fair comparison. Unfortunately Adelaide does not have a true inner CBD SLA to compare against (the central SLA includes all of North Adelaide, including lower density residential areas).

Sydney has the highest public transport mode share, with Melbourne and Brisbane very close (to my surprise). Perth is very much a car CBD, although mode shares are likely to have changed following the opening of the Mandurah rail line since 2006. The 2011 figures will be very interesting.

Perth walking more share was 3.0%, lower than 5.3-5.8% in the other cities – probably because of a lack of inner city residents.

And for the record, cycling was highest in Melbourne at 2.3%, followed by Perth at 2.0%, Brisbane at 1.5%, and Sydney at 0.8%.

The number of car driver journeys to work in the Melbourne CBD actually decreased from 34,289 in 2001 to 30,570 in 2006, a mode share drop from 27% to 23% (ref). This happened despite a 20% increase in the number of parking spaces in the CBD between 2000 and 2006 (ref):

I’ve included Southbank and Docklands in this chart for interest – Southbank parking supply actually went down between 2006 and 2008.

[parking stats updated June 2012 with 2010 CLUE data:]

Looking at commercial parking spaces only:

The number of commercial parking spaces has actually declined in the CBD and there has been very little growth in Docklands (despite an increase in employment).

Here is the ratio of employees to commercial parking spaces:

While the ratios are flat in three of the areas, Docklands has seen strong growth in employment without equivalent growth in commercial car parking.

Colliers International have recently begun surveying CBD parking costs. Here are the results for Australia (adjusted to AUD using 1 July exchange rates):

I don’t pretend to be an expert in CBD parking markets, but the differences between daily and monthly rates suggest some complexity. In Melbourne at least, it is quite common to find “early bird” parking for $13-17 (and “early bird” usually means parking your car before 10am).

I’m perhaps more inclined to go on the monthly rates, as they are probably more competitive. Melbourne prices collapsed in 2010, at the same time that public transport patronage growth stalled. Prices also went down in all other cities except Perth (which had the strongest public transport growth of the major cities in 2009-10).

So is CBD parking price a driver of public transport patronage? Probably too early to tell because of a lack of much time series on parking cost data (including 2006 data), but worth looking at in future.

What modes are different commuters using?

Firstly, mode share of motorised journeys by age and gender:

As you might expect, public transport mode share is higher amongst younger people and females. But for females it is also high for older women, with a curious dip at 35-44 years (typical kids at primary school years?). For men, private transport mode share was higher for older men. I’ve not shown 65-74 because the total number of such commuters was very small.

I’ve put non-motorised modes on a separate chart as they are much lower shares:

Walking was much higher for younger people. Is this because of lower car ownership, less willingness/ability to pay for transport, higher residential proximity to the CBD, and/or higher health and fitness focus? Unfortunately I don’t have the datasets to answer those questions.

Cycling mode share peaked with men aged 35-44, with men much more likely to cycle than women.

For reference, here is a demographic breakdown of CBD workers – it peaks at 25-34, with women slightly younger on average:

And here are the same charts for Brisbane:

Sydney:

and Perth:

You can see:

  • cycling mode share peaked for men aged 35-44 in all cities
  • walking tended to peak for people aged 25-34
  • public transport mode share dipped for women aged 35-44 in all cities
  • In Perth, men aged over 35 had a higher private transport mode share than public transport, the only city where this occurred.

So, do executives (presumably many from wealthy inner city suburbs) shy away from using public transport?

Indeed they do. They represented 16% of Melbourne CBD workers, but 24% of car commuters (9538 car trips in total). Maybe because many of them get company cars/parking as parts of their packages? More on that coming up.

Lower paid clerical and administrative workers were most likely to use public transport (and probably least able to afford driving and parking costs).

Note that Machinery operators & drivers also had a higher private transport mode share – I expect many are professional drivers coming in their work car (there were only around 1000 in this occupational category).

Back to managers – the next chart shows they are also more likely to snub public transport in Sydney, Brisbane and Perth:

What about other trip purposes?

The following charts show data from the VISTA 2007 household travel survey, that includes all trip types and all of Melbourne.VISTA is a survey, not a census, so there is a margin or error involved, and unfortunately the sample sizes are not large (provided in charts as “n=”). The total VISTA 2007 dataset has 2955 surveyed trips into the Melbourne CBD (across all days of the week), of which 1973 were motorised.

First chart shows mode split for trip legs into and out of the CBD, by time of day on weekdays:

Weekday AM peak is 7-9am, and PM peak is 3-6pm, anything else is classed as off-peak. Unfortunately there are only 190 trips in/out of the CBD on weekends in the sample, which has too large a margin of error to be too meaningful (7%).

Active transport (walking and cycling) and public transport were clearly dominant. When looking only at motorised trips, Public transport took 74% of inbound AM peak and outbound PM peak trips, and 67%/62% of off-peak in/out bound trips.

Recall above that motorised journeys to work in 2006 were 67% by public transport, suggesting people travelling for reasons other than work in peak periods were slightly more likely to use public transport.

What about wealth? I’ve used average household income per occupant, to remove the impacts of household size, and grouped this by $500 amounts. Note: the sample sizes are quite small for larger income groups.

Sure enough, there appears to be a trend that people from higher income households were more likely to use private transport for travel into the CBD.

What about age?

While the sample sizes are relatively small, there certainly appears to have been a higher propensity to use private transport for travel to the CBD by middle-aged people.

There may be a trend back to public transport for older people, but the margin of error is around 10% for the last two age groups so this is not certain. However it would fit with Seniors being able to access cheaper public transport fares.

In terms of gender, 73% of females who used motorised transport came by public transport, compared to 67% of males – a similar difference to commuters.

Who’s paying for the private transport?

While for many people driving to the CBD for work everyday is something of a non-option, there are still tens of thousands who do. Is employer sponsored driving and parking costs influencing their mode choice?

VISTA lets us take a look at that also, although there is only a sample of 183 AM peak private transport trips (margin of error around 7%).

According to the data, around 29% of cars driven into the CBD in the AM peak had their running costs paid by a company, and 36% had parking paid for by employers. Remarkably, 34% reported no parking costs for off-street parking (these trips mostly for work purposes) – which doesn’t sound right for the CBD in the AM peak! I’m not aware of any publicly available free off-street parking spaces. Perhaps the respondents overlooked the fact that someone else was at least paying for the land on which they parked? If that is the case, then it would appear that less than a third of cars driven into the CBD in the AM peak were not employer subsidised in parking or running costs.

Employer subsidies appear to be an incentive to drive to the CBD. By contrast, only around 2% of general Melbourne AM peak car drivers had employee paid parking, and only around 13% had car running costs paid by an employer (VISTA 2007).

One of the most effective ways to reduce car mode share for journeys to work in the Melbourne CBD would appear to be reducing employer subsidies for parking and driving costs. Schemes such as parking cash out help employees see how much their parking and driving costs are being subsidised. If they have the option of receiving that money directly as salary they might make different choices (depending on tax treatment of course!).

That said, with current capacity issues on Melbourne’s trains and trams, trying to shift more CBD commuter trips from car to public transport in the short-term might not be a government priority just at the moment.

And lastly, for the record, 6 and 8 cylinder cars parked in the CBD did not appear to be over-represented. Cars of well-known luxury brands were over-represented (15% v 6% metro average).

I think that’s enough now! ūüôā

Active transport is at 8%, which probably represents those who live within walking distance of the CBD. In order to take out the walking component, I’ve also taken a sample that excludes an “inner ring” around the CBD, as shown in the following map:

If you take out the inner ring, the mode split is 69% PT, 28% car, 1.9% cycling and 1.4% walking longer distances.

Advertisements

Where do the employees come from? (Melbourne 2006)

Sat 15 January, 2011

If we want to improve transport options into employment areas, it helps to know where the employees are coming from.The answers are in the gold mine that is census journey to work data.

This post maps where employees come from for major employment destinations around Melbourne, and looks at whether public transport is servicing these areas well. The CBD will be the subject of a separate post.

About the maps

Skip this section at your own peril – there’s some important things to keep in mind:

  • For each employment centre I have generated two or three maps:
    • the number of commuter trips originating in each SLA – ie where do the workers live?
    • the private transport mode share for commuter trips from each SLA. While I often look at public transport mode share, that doesn’t account for walking/cycling for shorter trips, and car trips are really what we need to reduce. Car mode share is only shown for an SLA where there are more than 100 originating trips, Note that given census data never reports values of 1 or 2 (for privacy reasons), we need to be careful about reported car mode shares that are 97%+.
    • for inner city areas only: the number of private transport based trips originating from each SLA (a high car mode share might not be such an issue if there aren’t many car trips being made).
  • I have used an SLA-to-destination zone journey to work dataset from the 2006 census (with thanks to the Victorian Department of Transport).
  • I have defined each employment area as a collection of destination zones, and then looked at the SLA origins for people working in those areas. These employment areas are shown in black shading on the maps.
  • On the maps showing quantities, shading represents relative density while the numbers are absolutes for each SLA. Some SLAs have larger areas and/or populations than others. For outer metro SLAs (which are often only partially urbanised) the density figures will always be low, but the concentration within the urban are might be higher. In an ideal world I would use density of residential areas only, but that takes a fair bit of work and I do this blogging in my own time. So you need to interpret those carefully.
  • The private transport mode share maps use the same colour scale for mode share values (hence some are very green and some very red).
  • As usual, you will need to click to enlarge maps.

Inner City Destinations

South Melbourne employees by SLA:

You can see large sources from the inner southern suburbs, particularly to the east and south.

South Melbourne private transport commuter mode share:

In general only about half commuters come by car, but this includes the nearby inner southern suburbs, despite being directly connected by tram routes. Perhaps the high car mode share reflects relative ease of parking and some awkwardness in getting to all parts of South Melbourne via public transport (for many it would require changing trams at Domain Interchange). A previous post showed that car mode share varied from 35% near Flinders Street Station up to 58% along St Kilda Road, and 67% in the south-west.

Car mode share is 87% from Point Cook/Werribee South (227 commuters) and Rowville (188 commuters), both of which lacked good radial public transport connections (Rowville is now served by a SmartBus direct to Huntingdale Station). Another high car share area is the affluent suburbs in Bayside, particularly south of the Sandringham train line (refer previous post).

South Melbourne employees commuting by private transport, by SLA:

This third picture is interesting:

  • The highest car origin density is still from the inner southern and south-eastern suburbs, on relatively short trips where public transport is generally frequent and direct (although not always fast).
  • There are many cars coming from the inner north and Williamstown, which requires crossing the city or Westgate Bridge.

Perhaps the number of cars coming from the inner south-eastern suburbs could be reduced if it were possible to travel by public transport directly to more parts of South Melbourne? Tram 8 does provide a link from South Yarra station to St Kilda Road (you can then change again to tram 55), but perhaps better access is required to other parts of South Melbourne?

Or perhaps it is more to do with ease of parking arrangements?

I’ve had a look at VISTA 2007 data for the Docklands/Southbank SLA – although the sample is very small so need to treat this with caution!

Only 41% involved self-paid parking of a personal car. Another 21% was free parking of personal cars, while 38% involved employer parking and/or company cars. While this is a very small sample, it suggests ease of parking is quite probably a strong determinant of car mode share. I’ll do a similar analysis for the CBD in an upcoming post where there is a larger sample (n=250).

Parkville employees by SLA:

The major sources for Parkville are the inner northern suburbs, which is not surprising. People tend to work and live on the same side of the city because transport is usually easier. Parkville is not directly served by the rail city loop for those coming from the south.

The proposed metro rail tunnel would connect the Sunbury/Sydenham line (north-western suburbs) to Parkville direct. However, the maps shows that the north-western suburbs are not currently a major source of Parkville employees (there will be students as well of course). This may of course change once the line is open, but it does present a patronage challenge to the project.

The relatively high catchment along the Epping and Hurstbridge lines suggests a more direct link from there to Parkville might have some potential for public transport mode shift. Already there is a high frequency bus service along Johnston Street which connects to Victoria Park Station on these lines. However, only half of peak period trains stop at Victoria Park station at present.

Parkville private transport commuter mode share:

(17/1 – this map has now been corrected since original posting)

A standout is Moonee Valley West at 87%. Not quite sure what is going on there – although most residents would have had to catch a bus, train and tram to get to Parkville by public transport in 2006. Such a trip would still involve two transfers now, but to a more direct and high frequency route 401 bus at North Melbourne station.

Car mode share is also high to the east in Boroondara and Manningham and in the outer western and northern suburbs. Being effectively on the same side of the city would make car commuting relatively easier.

Those coming from the south-eastern suburbs would appear to be largely content with public transport, although wealthier Brighton had a higher car mode share.

Parkville employees commuting by private transport, by SLA:

Car mode share and volume is also relatively high from Moonee Valley. While the 59 tram connects these two areas, it is relatively slow. The recently introduced high frequency 401 shuttle bus from North Melbourne to Parkville may have since increase public transport mode share from Moonee Valley workers in the catchment of the Craigieburn rail line.

There are also quite a few cars making the awkward trip from the St Kilda and South Yarra areas. But otherwise they tend to come from nearby suburbs – many of which are directly connected to Parkville (at least from the north).

Fishermans Bend is perhaps one of the most interesting areas in this analysis. It is on the eastern side of the Yarra River and relatively close to the CBD, but take a look at where the employees come from:

Fishermans Bend employees by SLA:

While the densities are highest from the inner southern suburbs, there are actually large numbers from the western suburbs (this is where low average density SLAs impact the results). There are around 3000 Fishermans Bend employees who live on the western side of the Maribyrnong River.

In fact, here is a similar map, except shaded by total number of employees (rather than density):

Most of them come from the western suburbs. And they come by car…

Fishermans Bend private transport commuter mode share:

No wonder the Westgate Bridge is heavily congested. As shown in the figure below (from a report to the City of Melbourne by Paul Mees citing a 2005 VicRoads survey), 34% of cars on the Westgate Bridge exit at Todd Road (the main Fishermans Bend exit) (for trucks the figure was 31%). That’s a third of the traffic on the Westgate Bridge . We seem to have a signficant public transport gap here!

Current public transport access to Fishermans Bend is primarily by bus from the CBD (routes 235/237/238), although one bus (route 232) runs from a small catchment in North Altona over the Westgate Bridge and along the southern edge of the industrial area (Williamstown Road). Driving over the Westgate Bridge would appear to be a more attractive option than the current train-bus option (here is a map of what such a trip can look like). Perhaps a stronger public transport link from the western suburbs is needed, maybe one that connects with the train network in the west. The current route 232 commences from the site of the now closed Paisley railway station, through which about half of Werribee peak period trains pass.

That said, the major bus service from the CBD have been realigned recently to operate from directly outside Southern Cross Station, and Werribee and Williamstown line trains run direct to Southern Cross in the AM peak. However there are still headway gaps of 20-30 minutes during some parts of the AM peak on these routes. It might be possible to slightly increase the frequency of these routes with existing resources if the routes terminated at Southern Cross station instead of Flinders & Market Streets (although I am sure a minority of existing users would not like this). Travelling via Southern Cross might indeed be the fastest way to reach Fishermans Bend by public transport, unless bus services could get priority over the Westgate Bridge (which is currently congested by single occupant cars).

Fishermans Bend employees commuting by private transport, by SLA:

The cars also come from the inner suburbs, including some density from Port Phillip west and St Kilda (which has a direct although infrequent bus service to Fishermans Bend).

Docklands is a fast growing area directly to the west of the CBD. Because of this growth, patterns may well have changed significantly since 2006.

Docklands employees by SLA:

For an employment area on the western side of the CBD, employees tended to live on the eastern side of the city – which is not what you would expect. But if they are coming via public transport, then trains from most parts of Melbourne provide relatively good access to Southern Cross Station, on the eastern edge of Docklands.

Docklands private transport commuter mode share:

Car mode share is significantly higher from the western suburbs, from which it is easier to drive to Docklands, and using public transport requires an indirect journey via Southern Cross.

However North Melbourne Station is only a short distance from Docklands. When the “E-Gate” site is redeveloped, it seems like a perfect opportunity to link the two with quality public transport (maybe a short tram extension, although we’d need a larger tram fleet). This might significantly increase public transport mode share and take pressure off the city loop railway.

Docklands employees commuting by private transport, by SLA:

The car numbers in 2006 were quite low, so we cannot read too much into them.

Suburban Employment Centres

(For subsequent centres, I will only show two maps, as car mode share is very high in most places)

While not a household name, Notting Hill is a relatively job dense area just north of the Clayton campus of Monash University.

Notting Hill employees by SLA:

You can see many workers live to the west, east and nearby north. But note that the two SLAs to the south (Kingston and Greater Dandenong) have a low average population density, so it is likely that the actual density of employees is higher to the south.

However there are also many employees from the Hallam and Berwick parts of Casey (which show as low density given the SLAs have a low urban density). There are fewer local jobs in Casey, meaning workers need to travel longer distances to work.

I would guess that nearby Clayton and Mulgrave industrial areas would show similar patterns (alas, I haven’t had time to analyse these areas).

Notting Hill private transport commuter mode share:

Notting Hill is full of cars (see the car parks on NearMap for yourself).

In 2006 there was a SmartBus route along Blackburn road (the eastern edge) and the 733 bus ran every 15 minutes in the peak on the western edge. East-west bus routes (693 and 742) run more like every 30 minutes in the peak. (All of the timetables hare hardly changed since 2006).

You can see on the map slightly lower car mode shares from the local area, but also up towards Blackburn (89%) – perhaps reflecting the attraction of the SmartBus service.

The south-eastern suburbs present a challenge: how does public transport service people who live in the suburbs of Casey and work in Notting Hill and surrounds? The current public transport system requires transferring from a low-frequency bus to a train, and then another (slightly more frequent) bus. Two transfers will struggle to compete with the car. Indeed, looking at VISTA 2007 data, it seems only 4.8% of public transport journeys from home to work in Melbourne involve two transfers.

There is a school of through that says bus routes should not be designed to connect everywhere to everywhere with a direct route – because you end up with a complex network of infrequent routes (than you would otherwise have with the same resources). However, if there are clear concentration of origins and destinations, might it make sense to run a direct service between the two?

Would a bus route that performs a collection function in Casey, runs express along the Monash freeway, and then performs a distribution function in Clayton/Notting Hill be viable? It may not need to run at a very high frequency, because transfers are not required. Perhaps the only way to find out is to do a trial (at a not insignificant cost). Such a service might not need operate extended hours, particularly if a guaranteed ride home was provided by employers who can otherwise save on parking costs. It would also still be possible to travel home using other public transport routes in off-peak times.

And now for the other major destination of Casey workers…

Dandenong South employees by SLA:

A majority of Dandenong South employees come from Casey to the east, although there are also quite a few from the Frankston area.

I think the lack of an east-west bus route connecting Casey to South Dandenong is probably one of the largely missing links in the Melbourne’s public transport network. Recent focus in the area has been on upgrading north-south bus routes (901 and 857), which will certainly help the a couple of thousand commuters from those corridors. But the 6000 odd employees from the east gained no better access following the December 2010 local bus upgrade.

Dandenong South private transport commuter mode share:

You can see almost universal car dependence for South Dandenong employees. It only drops to 91% in central Dandenong (sorry that number has been obscured by the shaded areas on the map).

Moonee Ponds employees by SLA:

While not recognised as a central activities district (CAD), Moonee Ponds actually has quite a bit of activity, includes some multi-storey office blocks (something other centres could only hope to achieve). Most workers come from the local area, or the nearby north-west.

Moonee Ponds private transport commuter mode share:

The reason I have included Moonee Ponds is the low car mode share from the Brunswick area (to the east). Perhaps this reflects the high frequency east-west bus routes that feed into Moonee Ponds from Brunswick? That said, the number of employees coming from Brunswick is low. Maribyrnong is connected to Moonee Ponds by one tram and two bus routes, but still 79% of people drive.

Central Activities Districts

The Melbourne @ 5 Million urban plan for Melbourne places greater emphasis on six suburban regional centres, to act like CBDs in the suburbs. I’ve taken a look at a couple of these that are perhaps better served by public transport. My earlier post showed many of these centres already have very low public transport mode shares.

Box Hill employees by SLA:

People working in Box Hill largely come from the local area, but from the north (Manningham west) and towards the east along the Lilydale and Belgrave rail lines. This is good – in theory – for the rail system in that it frees up some capacity on citybound trains in the morning.

But do they use public transport to get there?

Box Hill private transport commuter mode share:

Unfortunately not – car mode share is 85%+ for most of the eastern catchment. It drops to 59% in the Box Hill SLA itself – probably a combination of walking and bus access.

In 2006, 90% of commuters from Manningham west drove. This might have been reduced slightly by the introduction of the 903 SmartBus which runs every 7.5 minutes in the peak (although the previous 291 service was around every 10 mins in the peak). Some local routes connecting Manningham to Box Hill do run frequently in peak periods (eg 279, 286).

Dandenong CAD employees by SLA:

Central Dandenong attracts workers from the local area, particularly to the south and east. No surprises there.

Dandenong CAD private transport commuter mode share:

Car mode share is very high, with slight dips in the central Dandenong SLA (probably a fair amount of walking), the awkwardly U-shaped “Dandenong – remainder” SLA, and slightly lower in Pakenham. Overall Public transport mode share to central Dandenong is most around 4-7% (refer earlier post). It will be interesting to look at 2011 mode share from Frankston and Knox, which are now connected to Dandenong via the 901 SmartBus service.

I have looked at many other employment areas, but for reasons of space and time, I have not included them in this post. Many of those have quite predictable patterns (eg most Footscray employees come from the western suburbs). I’ve attempted to pick out centres with more interesting patterns.

Concluding remarks

I’ve not seen this sort of analysis done elsewhere, and I think it is important evidence to support planning transport systems – particularly public transport.

In the above analysis I’ve identified some “missing links” in Melbourne’s public transport network, including:

  • Casey to South Dandenong
  • Western suburbs to Fishermans Bend (via Westgate Bridge)
  • Casey to Monash industrial area

It might also be worth investigating new links to short-circuit trips to near-CBD locations:

  • North Melbourne station to Docklands
  • South Yarra station to South Melbourne (other than where route 8 runs)
  • More peak trains stopping at Victoria Park station to allow for convenient bus connections to Parkville

But it also looks like ease of car parking is having an impact on public transport mode share. South Melbourne sees many car commuters from nearby suburbs that are well connected by tram. Many areas outside the inner city would offer free employee parking, and driving is likely to be faster than public transport in most cases, particularly where on-road public transport is not insulated from traffic congestion. Unfortunately the census does not include data on who pays for parking and vehicles, so analysis of this issue is limited by the data available.

My next post will focus on the Melbourne CBD, and following that I hope to look at employment destinations of various SLAs (where do the workers go, rather than where they come from).


Trends in Melbourne Traffic

Sun 31 October, 2010

[first posted January 2010, last updated July 2013]

Each year, VicRoads publishes Traffic Systems Performance Monitoring bulletins that contain a wealth of data. However the bulletins don’t always show the longer term trends, so here are some charts to illuminate some of what is happening, particularly around congestion and total traffic volumes in Melbourne, and also a comparison of road traffic and public transport use growth.

Since the 2010-11 bulletin, VicRoads have changed the way they report various statistics. This post includes some charts of measures no longer reported, and in many cases the figures for 2010-11 and 2011-12 have been read off charts (as numbers were not published), so there is a little less precision.

Travel Speeds

The first chart shows average travel speed around Melbourne by time of day.

Melbourne average travel speed 5

The figures show a gradual slowing of speeds, although with some more erratic results in the AM peak in recent time (VicRoads haven’t offered an explanation, and I don’t have one either).

What about for different types of roads?

Melbourne average speed AM peak 3

Speeds have been declining on some road types, particularly inner freeways and inner undivided roads. Upgrades to the M1 completed in late 2009 may have increased speeds in 2010-11, but this appears to have been largely eroded in 2011-12.

Curiously, freeways slowed down after CityLink opened in 2001. Inner roads have been trending downwards slightly while outer roads seem to be steady (though with a slight increase in 2009-10). Outer freeways showed an increase in speed in 2008-09, which is related to the opening of Eastlink, but this was more than eroded in 2009-10, possibly related to the Ring Road widening works.

How does Melbourne compare to other Australian cities? This data is from AustRoads (except the 2011-12 figure for Melbourne):

Australian cities average speed AM peak 2

A gradual slowing trend is evident in most cities, except for speeding up in Melbourne and Brisbane up to 2010-11 (although Melbourne slowed down again in 2011-12). The Perth and Brisbane data seems to move around a lot more than the other cities. See the AustRoads website for qualifications on the data.

All-day average speeds by road type have not been reported for years after 2009-10, but here are the previous results:

Melbourne all day speed by road type

Is speed a good thing? If you want to get to your destination sooner, then obviously it is. If you want to improve safety and the liveability of streets, it possibly isn’t. A lot of road safety initiatives (eg right turn signal controls) actually slow down traffic. The other consideration is whether you experience the same travel speed each day – which brings us to…

Travel time variability

Another measure of congestion is travel time variability. If it sometimes takes 40 minutes to get to work, and you need to arrive at work on time, you have to leave yourself 40 minutes, even if on average it takes 30 minutes.

Melbourne variability of travel time 3

The recent trend has been towards increased variability after some improvements in 2009-10.

Traffic volumes

Since the 2010-11 Traffic Monitor report, VicRoads have completely revised their estimates for total arterial road traffic volumes and have divided Melbourne into three zones (inner, middle and outer) whereas previous results were for two zones (inner and outer). The new results show some quite different trends. I’ve read the data off the charts (not an exact process), to produce the following chart showing growth index figures (old figures in blue, revised figures in orange):

Melbourne traffic volumes by zone old and new

The previous estimates suggested traffic volumes stopped growing around 2004-05, whereas the revised estimates suggest an outer suburban growth spurt from around 2006, and a strong middle suburban growth spurt from around 2010. The 2011-12 bulletin did not include estimates for 2011-12.

This is what VicRoads says about the new method:

VicRoads has recently developed a new method for estimating road use in Melbourne. The new method uses all available vehicle counts to calculate VKT for all arterial roads and freeways in Melbourne. VKT values used in previous editions of the Traffic Monitor were derived from the monitored network only. The new method captures travel for the whole of Melbourne and allows estimates to be retrospectively made for previous years.

BITRE also estimate total Melbourne vehicle kilometres (using fuel sales data, so including all roads, not just arterials and freeways). The most recent data at the time of updating this post was Information Sheet 44. Here’s a comparison (again using index values):

Melbourne total vkms growth estimates

The BITRE data suggests a slightly different trend, but a similar overall growth figure and recent stronger growth. Putting the two VKT estimates together suggests that around 75% of Melbourne vehicle kilometres are travelled on arterials and freeways.

To add a little more context, here is a chart comparing road traffic and public transport patronage growth:

Melbourne total vkms and PT growth estimates

This chart really tells a story. Over eleven years to 2012, Melbourne public transport patronage has grown around three times more than road traffic.

Traffic growth on different road types

Here is a chart with 2010-11 data using the new road type classifications:

Melbourne traffic growth by road type

And here is the longer term trend using the older road types (up to 2009-10 only):

Melbourne index of traffic volume by road type 4

Traffic on freeways has risen substantially ‚Äď not surprising given massive investment in freeway infrastructure over the last couple of decades. There has been a substantial growth in the last few years, reflecting the opening of Eastlink in 2008, the widening of the M1 in 2009, and diversion of some traffic off other outer roads (and you can see a decline in “outer – divided”). Otherwise, freeway traffic volumes seem to be growing while arterial roads are showing stagnant or declining traffic volumes.

Take the freeways off the chart and you can see trends on other roads a bit more clearly:

Melbourne index of traffic volume by road type ex freeways 3

Traffic volumes have trended downwards on undivided inner roads (including those with trams), and have largely stagnated on all other non-freeway roads in recent times. You can also see a significant drop on outer-divided roads in 2008-09 – possible some of it from Springvale and Stud Roads that parallel EastLink.

For more detail (including other charts), I suggest reading the Traffic Systems Performance Monitoring bulletins on the VicRoads website.


Public transport mode shift and road congestion

Sun 4 July, 2010

Is a modal shift to public transport an effective way to reduce road traffic congestion pressures?

I’ve discussed this issue in a few posts, but I think the following simple chart pretty much sums it all up:

You can see a significant change in both trends from around 2004 onwards.

Arguably public transport mode shift has been the most effective method for relieving congestion pressures in Melbourne in the last five years.

That is not to say traffic congestion has been solved or significantly eased, but it would likely have become much worse if road traffic volumes had continued to grow after 2004-05.

More posts on road traffic and mode shift.


Car and transit use per capita in Australian cities (Peak Car)

Fri 8 January, 2010

[post updated in January 2016 with data from the 2015 BITRE Yearbook. For some more recent data see this post published in December 2018. First published January 2010]

Thanks to the Bureau of Infrastructure, Transport and Regional Economics’ Australian Infrastructure Statistics Yearbook 2015, a great set of time series data is available on transport behaviour. This post looks at trends in private and public transport use in Australian cities up until 2013-14.

The first chart shows car passenger kms per capita peaked in 2004 for all cities and has been mostly in decline since then (with the possible exception of Adelaide in recent years).

car pass kms per capita 4

This is¬†clear evidence of “peak car” use per capita in Australian cities.

Canberra has the highest car passenger kms per capita – perhaps due to the low density city, relatively sparse bus services, and general ease of using private transport.

The underlying figures show relatively little growth in total car passenger kms in most cities in the ten years since 2003-04:

index of car passenger kms 2

Perth has shown the greatest increase in total¬†car passenger kms despite the reduced car use per capita, presumably mostly reflecting strong population growth. Canberra, Sydney and Melbourne show a flat-lining in car passenger kms between 2004 and 2009 (a period of time that is becoming very¬†familiar on this blog). I’m not sure why Adelaide has seen¬†declining car passenger kms (it’s population did grow over this period by 10%).

Does this mean we are travelling less? The following chart shows estimated motorised passenger kms per capita, and yes in all cities there was a peak in 2003-04 followed mostly by declines. This might be people taking shorter trips, people taker fewer trips, and/or people substituting motorised transport with non-motorised transport.

motorised pass kms per capita 4

There is another story in the data. Mass transit passenger kms per capita rose significantly in Melbourne and Perth between 2005 and 2011:

mass transit share of pass kms 4

However, more recently there have been slight declines in Brisbane, Perth, Adelaide, and Canberra.

Sydney shows up with many more mass transit kms per capita than any other city (68% of which is on rail). While I am not sure exactly how BITRE sourced their train patronage data for “Sydney”, my experience is that patronage data is only readily available for a very wide definition of Sydney’s railways including lines beyond the Sydney Statistical Division including Newcastle, Nowra, and the Hunter Valley.

You can also see a spike in mass transit use in Sydney in 2000/01 – the year containing the Olympic Games.

Previous strong growth in Brisbane might reflect considerable investment in bus services (BITRE have published three years of bus data showing Brisbane increasing from 53.0 million kms in 2005-06 to 61.2 million kms in 2007-08). More recent stagnation in mode share might reflect above-CPI fare rises and changes to patronage measurement methodology.

The data also allows a calculation of mass transit’s share of motorised passenger kms:

Melbourne and Perth are the stand outs for mass transit mode shift based on these figures, particularly in the years leading up to 2009. Adelaide, Sydney and Brisbane also saw mode shift between around 2004 and 2009, but Adelaide and Brisbane have gone backwards since then.

Note there are other ways to more directly measure transport mode share – primarily household travel surveys, covered in another post.

Finally, here is a stark comparison of total car and mass transit passenger kilometre growth since 2003-04 (with population growth included for reference):

car v pt growth aus large cities 2

Car use has grown more than four times slower than population growth, and almost seven times slower than mass transit use.

Business as usual no longer

In the frequently cited¬†BITRE 2007 report on congestion costs, they made a “business as usual” assumption that mode shares will not change (ref page 8) and that car use per capita would increase (ref page 47), and then forecast that avoidable congestion costs in Australia will more than double between 2005 and 2020.¬†The latest BITRE evidence is that mode shares are not business as usual in most cities and that car use per capita is decreasing. See another post reviewing the forecasts of the costs of congestion.

An updated 2015 BITRE report on the costs of congestion looks at the issues of per capita vehicle user into the future in a bit more detail, including acknowledging researchers highlighting peak car.

Their new high case scenario assumes the recent downturn in per capita vehicle use is essentially all GFC related (and has persisted for¬†10 years), and that things will now¬†bounce back to old patterns (and grow even faster than the 2007 forecast). Their medium case also assumes a reversal of trend, but that we’ll only return to the previous peak level by 2030,¬†and a low case continues the current trend.

The following chart shows forecasts compared to estimated actual vehicle kilometres per capita.

BITRE vkms per capita estimates

The 2007 forecast was not only trending in the wrong direction, their starting assumption about vehicle km per capita for 2007 was well above the (revised) estimated actual.

The new estimates for the “avoidable social costs of congestion” in 2030 are $48.2b for the high case, $37.3b for the median case, and only $25.1b for the low case. BITRE suggest this median estimate is the upper range of what is plausible, but if current trends continue,¬†the avoidable social costs of congestion would be a third lower.

See also a recent piece on this subject by Professor Peter Newman in The Conversation.

Some disclaimers:

  • I’ve called it “mass transit” as the data does not seem to differentiate public, school and private bus passengers. Passengers on chartered buses aren’t usually considered “public transport”, but they still are a very space efficient way to move people on roads.
  • km figures are reported on financial years, while population figures are for June at the end of that financial year. So the “real” per km figures are slightly less, but I’m really just looking at trends and relative numbers here.