Sunday, February 28, 2010

Government database content and getting to work

The Victorian State Government has made available data on a range of topics. It is hoped that this will encourage entries in the App My State competition.

There are two data sets that directly concern public transport. Transnet is a database containing stop data and timetable information for most public train, tram and bus services in Victoria. It is maintained by Metlink. In addition another file contains stop data only. It is important to note that both files are frequently updated and there's a chance that what you download today may not necessarily be the latest version. Nevertheless it will still be useful for competition purposes.

The format of the above information may be unfamilier to those whose computer experience is limited to basic Word and Excel use. You will need to be into databases to get most benefit from this data.

Other information provided also has relevance for transport. An example is Suburbs in time for which data is presented in Excel format.

Suburbs in Time data is a bit old, based on the 2006 census, but is still interesting as there is a census question related to travel to work. For example, in 2006 20% (574/2661) of employed Glenhuntly residents took the train, versus barely 3% for Frankston (450/14898). And for Frankston just 5 more people took the train in 2006 than 1996, compared to a 221-person increase for Glenhuntly.

While there are differences such as Frankston Station's large catchment and the possibility that some Glenhuntly residents would board the more frequent service available at either Elsternwick or Carnegie, it does indicate that even mature suburbs can generate high patronage growth. Brunswick, another suburb attractive to city workers, has shown even more impressive growth, but for tram rather than train. As an outer growth area, Hoppers Crossing would have expected high patronage growth, and it did get some, but at least up to 2006 its growth has been modest relative to the high population growth. This is likely due to the distance of most local employment areas from railway stations, limited bus services (soon to be increased) and high car and parking availability.

The proportion who walked is another interesting stat. Walkable suburbs have a high proportion of walkers whereas less walkable suburbs do not. For example in 2006 202 Werribee residents walked to work (compared to just 74 who got the bus). Still a small proportion of the 16251 employed. Inner Werribee (both north and south) is a long-established and walkable area.

Hoppers Crossing is less walkable but has a similar workforce and proportion of walkers to Werribee. Possibly this could be attributed to Werribee Plaza (actually in Hoppers Crossing), which is a major employer and is walkable from some residential areas.

In 2006 Rowville had the same number employed as either Werribee or Hoppers but only 148 walked, about 25% less than either Werribee or Hoppers. It's a pedestrian-hostile neighbourhood and the largest employer would be Stud Park Shopping Centre. This is not walkable from most of Rowville and is smaller than Werribee Plaza. Rowville also has higher incomes and more multi-car households than Werribee.

Reservoir is another large suburb with similar numbers of employed persons as Werribee, Hoppers Crossing and Rowville (about 17000). Though there are some pedestrian-hostile intersections, much of the suburb is a walkable grid. The number of walkers to work here was 246, ie about 25% higher than Werribee and about 70% higher than Rowville.

While detailed knowledge of workplace locations would be desirable for further research, the above numbers indicate that suburb walkability does have some bearing on the numbers of people who walk to work. However given that work trips are often longer distance than other trips such as shopping and recreation, and this would work against walking, the extent to which non-work trips are made by walking is also important for a true idea of walking's share, and one not answered by the census data.


Riccardo said...

As someone who knows (and I suspect Mr Bowen does too!) - there's no chance a Glenhuntly person would board at Carnegie or Elsternwick within the bona fide catchment!

The Dandenong rail service is much too unreliable and full.

Peter Parker said...

I lived at Carnegie but would still take Carnegie peak over Glenhuntly peak.

True average loading is likely higher at Carnegie than Glenhuntly. But Carnegie's frequency is superior and more consistent (Glenhuntly did have two 16 min gaps in am peak, though one was recently plugged).

Glenhuntly lacks consistency due to the mix of loop and non-loop trains. Especially in the pm peak where they get a particularly raw deal (26 min gap between identical stopping pattern trains around 5pm assuming everything's on time).

Carnegie at least has most Dandenong trains stopping there and if there's none there's a 900 bus from Caulfield if Frankston trains are running.

In Glenhuntly's favour though is the 67 tram and the possibility of getting it across from Elsternwick; in contrast Carnegie doesn't have strong links from stations on other lines.

What about reliability? shows the Frankston line with a 1.5% cancellation rate versus 1.2% for Cranbourne and 1.4% for Pakenham.

For punctuality the above reports 74.8% for Frankston versus 82.4% and 74.9%.

The verdict is: Crowding - Carnegie loses, Frequency - Carnegie wins, Reliability - Carnegie wins (marginally).

The difference is probably not worth walking or driving too much out of your way to, but if you had other errands Carnegie offers more chance to do them. Ditto if you lived in between and cycled.

If you were choosing somewhere to live on the basis of trains, Carnegie is the clear winner, though you do miss the 20 min tram frequency Glenhuntly enjoys in the evenings.

glowingforest said...

Hi, I'm moving to Melbourne in a few weeks. Are there a handful of your posts that would be especially helpful for someone looking to get familiar with the transit system? Any other links (besides Metlink, of course) that are not to be missed?

Thank you!

Unknown said...

I'd have to agree with what Riccardo says.