Friday, 29 December 2017

Voices from the 'Other' Melbourne - Time for truly needs-based public transport policy

Source: Charting Transport

Cost-Benefit Analysis really is a rubbish way of delivering public transport. I say this in particular reference to Infrastructure Victoria's debatable means of telling us Doncaster Rail was a dud, but a possible future Metro Two tunnel whose longest inner north component would in practical terms do little more than grant hipsters in the inner north subway stations, may well stack up.

Close the Upfield Line

The Upfield line would not pass a cost benefit analysis if proposed today - it would be projected to exclusively cannibalise existing Craigieburn line traffic. Yet we are today actually planning for the line as a crucial conduit to new northern growth suburbs.

But none of the cost of the tens of thousands of cars that would be thrown onto the choked roads network if we closed the Upfield line today would be on that original cost-benefit study.

And none of the benefits of having two rail lines rather than one serving entire new growth regions for decades to come would be in there either. And all of the usage projections for the new well serviced commuter stations in the outer west have exceeded original projections.

The point in bold here, is the benefits of having heavy rail covering your entire urban geography have been obvious with hindsight. Society today considers ALL the prior money here well spent, no smart politician has talked about closing about closing a train line in this city for 30 years.

The point is that if you are going to talk about benefits in such an economically rigid way that you hive off MOST of what society at large infers by the term, then you're deliberately and systemically building a world that is less responsive to society's values.

Your entire methodology is actually invalid, not any scientific basis, but on a moral one. And because of this, your science is incapable of responding the problem.

Cost-benefit analysis is a discipline built on the same foundational faultline that undermines the entire discipline of economics. Statistical modelling is central to most of its foundational and core theories. This requires that an approximation model of the real world be created, and to do this, all manner of "assumptions" are made about the normal operation and behaviour of our world, and us the actors in it.

And here begins the problem, because the essential purpose of these assumptions is to allow for what is largely chaotic and random behaviour to be expressed in some form of mathematical expression.

Close the Frankston Line

To model Doncaster Rail, Infrastructure Victoria have made the assumption that fully 98% of its users would be cannibalised from the existing public transport system. You read that right. 98%. Without, apparently having anything obvious that might contextually support this figure, such as survey data from commuters in Doncaster (call me crazy). This number just appears from the air, and it's essentially this number that hobbles the entire cost-benefit study.

So, I have an unsolicited proposal for Infrastructure Victoria. Let's close the ENTIRE Frankston Line. Today. It doesn't stand up to cost benefit, because 98% of commuters would simply use other public transport modes. OK?
THIS IS INSANE, Y/N????

But the other point is the only actual benefit by which Doncaster rail was assessed by Infrastructure Victoria was as a conduit to further enable access to the CBD employment zone, which is correctly assessed as a low priority.

But we will ask the question shortly, what if the real possibilities of Doncaster rail are being deliberately missed? What if rather than connecting Doncaster radially to the CBD, we explored options to connect it ORBITALLY to the heavy rail network? The needs this is then addressing are around creating suburban employment clusters and are instantly much more pressing.

You don't need to dig terribly deeply around the media at present to find articles talking of the increasing creation of TWO Melbournes - an inner city one that has excellent access to all the best paying jobs and which has viable public transport options and a suburban one which faces extensive and increasing daily commutes which are necessarily longer, car based and mostly to suburban destinations.

I believe Infrastructure Victoria has built a number of assumptions about this second Melbourne that are actually deeply classist and that literally dial in more of the same. Working people ('bogans') just love their cars, and this is an ideology of theirs.

Whereas I would start with, "we've never provided adequate public transport to the places where these people live, and so they don't use it, and if this isn't the most obvious causal factor here then there really must be something wrong with me."

And there's an increasing body of evidence that says there's apparently nothing wrong with me at all, and the other viewpoint is in fact dangerous.

Charting Transport in a Non-Parallel Universe

The latest, which has prompted this rant, is from the fabulous people at the Charting Transport blog. Certainly the most thorough publicly available data on journey to work in Melbourne is to be found there for anyone interested. I'll take up the rest of this post by referencing what I thought relevant and really shouting out what I think the lessons here are.

The report is about journey to work, so the data divides fairly neatly into data relating to ORIGIN and that relating to DESTINATION. On the origin side of things, one thing hits us square in the face.

Over the past 15 years, Public Transport mode share in Melbourne has increased ONLY in areas of the city with direct access to rail (tram, light or heavy).


Source: Charting Transport

The data shows it is only suburban household very proximate to the stations who will use the network in volume, further underlining the apparent weakness of the existing suburban bus network as a systemic support to rail.

And all this with quite a bit invested in the system in recent times, so the data is clearly telling us that  simply continuing to invest incrementally in the system the way we have been will only reap the sort of incremental rewards we've seen here, not the significant change that everyone agrees is needed.

Source: Charting Transport

But, build it, and they will come.


Footscray, Flemington, Docklands, Carlton, and South Yarra were more obviously on this list, but the report showed "public transport mode shares of over 50% can be found in pockets of West Footscray, Glenroy, Ormond – Glen Huntly, Murrumbeena."

Glenroy is uniquely well served by both the Upfield and Craigieburn lines. The last three are essentially the same region where the Frankston line diverges from Pakenham-Cranbourne, so this is theoretically one of the best served rail regions in Melbourne. Build it and they will come.

And we HAVE shifted the paradigm for new greenfields development in Melbourne's West with Regional Rail Link. 

Charting Transport continue, "the biggest shifts to public transport in the middle and outer suburbs were in Wyndham Vale, Tarneit, South Morang, Lynbrook/Lyndhurst, Point Cook South, Williams Landing, Rockbank, and Glenroy. That’s almost a roll call of all the new train stations opened between 2011 and 2016."


Source: Charting Transport


I think we should all be crowing about this more, including the advocates for working Melburnians. This is not remotely how we built Noble Park. We seem to be laying the foundations of MUCH better planned communities for tomorrow's 'second Melburnians', and the communities themselves are voting with their feet. Build it and they will come.

I think it's extremely important that we are looking right now at making sure ALL the modifications to Upfield/Craigieburn needed to serve the growth communities in the north are as strategically well forward planned as Regional Rail Link was, and probably gotten underway today in preference to tomorrow. Having built it BEFORE they got there (hello, Caroline Springs station) seems to have been a success factor here.

Or Don't Build it, and They Won't Come

Unlike Sydney (and we've discussed this before), almost nobody is travelling to suburban work destinations in Melbourne by public transport, however the report clearly shows that concentrating jobs in suburban locations properly serviced by heavy rail is a key success factor in reducing car dependence for journey to work.
"There are some isolated pockets of relatively high public transport mode share for journey to work in the suburbs, including
  • 34% in a pocket of Caulfield – North (right next to Caulfield Station),
  • 33% in a pocket of Footscray (includes the site of the new State Trustees office tower near the station),
  • 25% in a pocket of Box Hill near the station, and
  • 17% at the Monash University Clayton campus
  • an area near Camberwell station (26.8% PT mode share)
  • Swinburne University Hawthorn (39.8% PT mode share),
  • a zone including the Coles head office in Tooronga (11.2% PT mode share)"
The commonality here is all too obvious. All are very near heavy rail stations, except Monash which is served by several relatively high frequency buses from three stations. Most are catered for by more than one line, and most are actual branch stations, therefore have an effective "3 spoke" coverage of their immediate region.

Keen readers will note that this was of course one of the key factors we were looking to maximise through the creation of a Wombat outer suburban orbital rail loop to maximise the network for exactly this factor.

Source: Charting Transport


Almost ZERO mode shift to public transport has taken place at workplaces more than 7km from CBD. All most all of the behaviour change of the past 15 years has been amongst CBD or inner city workers.


Source: Charting Transport

The Second Melbourne doesn't work in the CBD.

What particularly annoys me about Infrastructure Vistoria's Donacaster Rail debacle was it failed to consider what could be done from an urban consolidation perspective to make Doncaster a much more effective activity centre in its own right, to give local residents much better options to work and shop locally on roads that are less clogged, but what was modelled was really just existing behaviours.


Source: Charting Transport

The report states "public transport dominates journeys to the CBD, no matter how far away people’s homes are, but the number of such journeys falls away rapidly with home distance from the CBD. Very few people commute from the outer suburbs to the CBD."

So, what about the possibilities Doncaster rail created for all residents who have needs to better access local destinations, rather than just for the probably wealthier portion of that society who work in the CBD? Is it because these latter people probably look a lot like the study authors that their mindset has been dialled so front and centre in to the equation?


Options for connecting Doncaster ORBITALLY to the heavy rail network
 as stage one of a Melbourne outer orbital rail loop, creating new RADIAL catchments
into existing suburban retail, education and employment clusters

We keep extending a radial rail network to suburbs where the data tells us these people do not to a large extent want to travel radially for very far.

Orbital Rail is Needed in Melbourne

The Wombat "Maximal Network Effect" vision for heavy rail in Melbourne

The number of large suburban job centres that are adequately served by the current heavy rail network is very small, and the distribution of jobs across the suburbs is highly dispersed.

The radial public transport network is therefore at its limits to provide anything further for the second Melbourne. While it remains radial, it can only properly transport workers to one destination - the CBD. And while the “new economy” jobs growth is occurring within such a geographically constrained space, if you don’t have access to the CBD, you're getting shut out of the new economy altogether.

So while the network remains radial, it persists as a vehicle of entrenching economic disadvantage, and every cent spent on public transport in this city is actually discriminatory.

For years we have been planning public transport from a "zone one" mindset. But Zone Two is increasingly crying out for the sort of congestion relief that road projects are at their limits of being able to provide. 

Creating effective suburban job centres adequately served by heavy rail is the better mousetrap that we are looking for to solve not merely urban traffic issues, but also address a whole range of social equity and employment access issues that are presently plaguing the second Melbourne.

Every bit of data in this report seems to suggest that creating a secondary rim of suburban job centres around 20kms from the CBD should now be a major policy goal.

Here's a plan for an orbital rail network that will meet this demonstrable NEED. Sod your cost-benefit study, we need to discuss a better future.