The road space problem
Russell Degnan

As the debate over Swanston Street rears its ugly head again, the advocacy groups for their particular mode are out in force, to try and asign a few precious feet of street space to their own uses.

The biggest casualty will be the tour buses, but they are, perhaps, the most unfairly treated. Spencer Street station is a significant downgrade from their presently central position. Presuming noone is willing to create a proper bus park along Swanston Street, space next to Federation Square or in an adjoining street should be made for them.

But the tour buses are but a small, mostly immobile, part of a larger problem, as the photo below shows quite clearly.

Even without tour buses, Swanston Street is packed with delivery vehicles at all hours, taking advantage of the ease of parking. Taxis, similarly, take advantage of their access, both for parking and through traffic, and of course (as has been acknowledged for almost 100 years), Swanston Street has too many tram routes for efficient running. The street was never going to be a "walk", though you can walk across it, almost at will. The current configuration is better than what it was, but is a terrible mess.

Robert Doyle's plan to return traffic is either naive or mad. As can be seen in old videos the street has changed significantly in its post traffic days. Apart from the reduction in direct end-to-end through traffic, the footpaths are now substantially wider, street trees and cafes filling the space. Returning traffic under that configuration (two lanes and parking), means either reducing the footpath width, at great cost to the major use of the street (pedestrians), slowing tram speeds further with extra traffic between intersections, or removing parking.

Which oddly enough, makes Swanston Street the same problem facing the shopping strips of inner Melbourne.

The problem of space is obvious at a design level, but too quickly forgotten at a strategic one. At a recent forum, speakers were regularly clapped for promoting "bicycle lanes on every street, protected by parking spaces", "dedicated tram lines", and while it was left unsaid, I am sure they would also be applauded for advocating "increased footpath width and better urban design". I like these things too, but the underlying theme was that the lack of these things occurs because there is a vast road lobby, conspiring to thwart alternative modes, and better livability.

There isn't. On many of these streets, VicRoads is responsible for both improving tram speeds, and on-road bicycle lanes. They have, where possible, dedicated lanes to both trams and bicycles. It just isn't possible on many streets. Not without removing something. And the biggest obstacle to removing something will always be the users of whatever that something is.

Sydney Road is a classic example here. The road is just 1 chain (20m) wide. A 3.5m tram line, 2m bicycle lane, and (narrow) 2.5m footpath leaves just 2m for traffic and parked cars. Even without parking, unless we want to revert to widespread one-way traffic there just isn't space for dedicated lanes for all modes.

Some modes need to be mixed, and the best configuration is a complex negotiation between stakeholders, not feel-good statements of intent of no practical value.

For myself, the priority should be giving p/t dedicated lanes, and increasing pedestrian access. The former because they are the most efficient (and equitable) movers of people across reasonable distances, and the latter because local people should have first access to their environment, and pedestrians are, predominantly, locals. Bicycles are great, but they can share streets, particularly when those streets are slow moving. What makes cycling unsafe now is not traffic per se, but the need to weave in and out, merging with faster traffic at multiple points, or having faster traffic merge with them, pushed from lane to road, and back with little warning for drivers behind.

Parking is a luxury, the value of any particular space marginal to a business, in comparison to the pedestrian traffic, and its removal from streets moderately easily made up for in off-street solutions as the market dictates. This may inconvenience delivery drivers, but few can be parking directly outside their destinations already, and off-street parking could make provision for them.

On Sydney Road, and other inner city roads, that gives 3.5m for trams, 3.5m for cars (now speed limited to 30 or 40 - if they ever reached that speed) and bicycles, and 3m for pedestrians. At intersections, hook turns should be the norm, and light cycles short (an elderly pedestrian crossing and no more). Bicycles should take the whole lane in designated areas, not the left-most edge, as the time savings for traffic behind are minimal.

In the city, with wider streets and already blessed with wide footpaths, a 2m bicycle lane (and even some parking) is possible, though the advent of universal super stops puts further strain on the limited resource.

But we do need to have a debate over space. The extended clearways plan is, in many ways, the most radical change to Melbourne's urban environment. It strongly favours moving modes over local areas, and implemented with little consultation with councils, businesses, or residents. The feedback they did get was almost universally negative, fobbed over with reference a "the silent majority" who remain either ambivalent or unconvinced. The ideas aren't bad necessarily, but we should be cynical of proposals for greater efficiency that run afoul of induced demand, and of advertisements that mysteriously add an extra lane to the configuration of the bulk of Melbourne's roads.

And we should ask what role in our urban system strip shopping centres serve, above and beyond funnels for vast numbers of CBD commuters.

Sterner Matters 19th December, 2008 13:06:22   [#] [0 comments] 

Ratings Update - December 2008
Russell Degnan

The rating system has had another tweak. There has always been the need for a balance between sensitivity to particular tests, and the need to shift the rating quickly when circumstances change. To bypass this problem, I have implemented protection to the rating.

Essentially, any single game change to the rating is divided in half. One half changing the rating, the other added to the long term trend, which is then added (again, by halves) if the trend matches games that follow. As usual, blogging about history can wait, but it has made a fascinating change to the current ratings:

Australia 1266.48 (1st)
South Africa 1188.62 (2nd)
India 1160.06 (3rd)
Sri Lanka 1107.36 (4th)
England 1100.18 (5th)
Pakistan 1064.46 (6th)
New Zealand 981.8 (7th)
West Indies 917.89 (8th)
Bangladesh 595.29 (9th)
Zimbabwe 542.57 (10th)

There is, however, little change to the expected margins in the tests to come:
Australia (over South Africa) by 89 runs
New Zealand (over West India) by 82 runs
India (over England) by 87 runs

Forthcoming Series

Bangladesh (542.57) v Sri Lanka (1107.36) - 2 Tests

Expected Margin: Sri Lanka by 232 runs

As usual, I missed a whirlwind stop of Bangladesh. Sri Lanka have made a habit of thrashing the struggling minnows, and this could hardly be expected to change. Both sides are on an upward trajectory ratings wise, so the margin may matter, even if the result is a foregone conclusion.

Idle Summers 17th December, 2008 12:45:07   [#] [0 comments] 

Monday Melbourne: CLXIX, December 2008
Russell Degnan

I don't put up enough pictures of Flinders Street Station. Taken December 2004

Melbourne Town 17th December, 2008 12:29:53   [#] [0 comments] 

Ratings - December 2008
Russell Degnan

Australia v New Zealand

Opening Ratings: Aus: 1427.55 NZ: 1021.68
Expected Margin: Australia by 177 runs
1st Test: Australia by 149 runs
2nd Test: Australia by innings and 62 runs
Closing Ratings: Aus: 1436.33 NZ: 1010.55

Claims the this, or that aspect of international cricket is killing the game are common, and, in the past I've given them short shrift. Partly, because the key series are as strong as ever. Australia, England, India, and to a lesser extent, South Africa and Sri Lanka have strong sides and high interest (albeit not always at the ground). But there are problems elsewhere, as this series highlight. It is practically impossible to assess Australian performances because they were on another level to the Kiwis. Even in Brisbane, rolled twice for low scores, they never seemed in trouble. I asked which players will enhance their reputation, and the answer became Clarke and Johnson, but neither can rest on their laurels, because the selectors won't be rating games against New Zealand that highly.

On the other side of the Tasman things are bad, as they slip into an ever more populated abyss for wretched cricket sides. The three best sides of the 1980s are all rubbish, New Zealand and Pakistan have been decimated by the ICL, the West Indies by ineptness, and a culture dedicated to wasting talent. For political reasons, neither Pakistan nor Zimbabwe have played test cricket, though Zimbabwe are also not playing because they are complete rubbish. Bangladesh escape that fate because they vote with India. With only a few exceptions, poor teams have been a constant throughout test history. But never before have they been so poor, relatively speaking, nor have they declined so dramatically from positions of relative strength, and nor have there been so many short, meaningless series between teams. Test cricket will live on, but it may yet not live on everywhere it currently resides.

Things may look up for New Zealand. If they can retain their current lineup, then they have players who can play. Ryder is talented, if brainless, Taylor is a genuine talent with similar issues, O'Brien a funny blogger and capable bowler, Southee capable of everything but inconsistent, and there is something about Flynn that bodes well. Oram, Franklin, McCullum and Vettori will provide a decent core if their batsmen could just put their heads down and score. But their inability to carry an innings out will haunt them until they learn.

South Africa v Bangladesh

Opening Ratings: SAf: 1284.20 Ban: 374.79
Expected Margin: South Africa by 353 runs
1st Test: South Africa by innings and 129 runs
2nd Test: South Africa by innings and 48 runs
Closing Ratings: SAf: 1283.37 Ban: 376.03

Another pointless waste of time between two mismatched sides. There is so little to take from this from either side. South Africa scored runs and took wickets, yes, but any bowler or batsman who didn't is probably unlucky. Mushfiqur Rahim and Junaid Siddique made some handy runs for Bangladesh, but given neither averaged over 30 that is clutching at straws. Shakib Al Hasan however, did take 11 wickets at 20.81, which given South Africa only batted twice is practically a 10 wicket haul. The ratings of neither side changed, which says it all really.

Forthcoming Series:

New Zealand (1010.55) v West Indies (882.56) - 2 Tests

Expected Margin: New Zealand by 80 runs

Two sides struggling to maintain respect for their cricket, fighting officially for second last on the test table (the position of Bangladesh, like statistics versus them, now routinely excluded from discussion). Why a series between two sides in this predicament is two test is beyond me, as both could use a proper work-out, but I suppose that is the way of it. New Zealand should win, they are at home, where they have always been hard to beat. The Kiwis are not a good side, but nor are the West Indies, being erratic, dependent on Chanderpaul and Sarwan for runs, and wishing they could depend on someone for wickets. But, New Zealand are also faltering, badly, have a brittle batting lineup, and haven't shown any ability to win games even while in front. An interesting proposition then.

Australia (1436.33) v South Africa (1283.37) - 3 Tests

Expected Margin: Australia by 85 runs

Don't pay much heed to that margin, this series could be anything. South Africa have had a year to remember, sweeping minnows aside with ease, but also playing well away, to draw with India and beat England. Their batting, for so long only capable of pretty 50s unless Smith and Kallis played well, has routinely scored big hundreds, even as Kallis has lost form. Their bowling, always before an unimaginative right arm pace attack, is still an unimaginative right arm pace attack, but with real menace in Steyn and Morkel. They could be rubbish against Australia. Just three years ago they played well only to lose sessions every time they mattered, such that the eventual margin flattered Australia. You suspect, against an Australian side less inclined to always win, that they could surprise. But don't suspect Australia will take them lightly. Even if Australians don't share the same sense of rivalry, they do like playing a tough unit, and will rise to the contest.

But, are Australia good enough. If South Africa are as stable as they've been since re-entering test cricket, Australia are as unstable as they've been since before then. Almost every player is worrying. Even as Katich has cemented one opening position, Hayden has looked less certain in the other. Ponting lacks that hunger for runs that marked him before, his captaincy (and legacy) being questioned every time he plays. Hussey is solid, but may also be in a form slump. Clarke has been scoring runs, but has yet to prove himself resilient in the face of a collapse. Symonds lurches from crisis to crisis, struggling throughout the early summer. Haddin hasn't yet settled with bat or gloves, drops catches, lets through byes, and seems uncertain keeping on poor pitches, with only one knock in favourable conditions to tally, Watson is probably in great form, but may not play, and hasn't yet proven his worth if he does. Lee has proven his worth, then regressed to bad habits, bowling loose spells on two lengths, tempered only by moments of genuine skill. Clark has bowled a good line and length, but is either too slow, too good or too unlucky, with batsmen generally content to play him out given his support. Johnson is improving rapidly, and may well be the key, with a seeming knack for taking important wickets, even while bowling complete rubbish. And Krejza, has one test, a big bag of wickets, and an injury - so your guess is as good as mine.

Clearly Australia are not what they were, they dropped an almost record 174.71 points across 11 games following the win in Sydney. South Africa have steadily added almost 100 points over the same period. Often, the gaining side is primed for a fall, and vice versa. That, home advantage, a history of dashed expections, and a tendency for their opposition to start a series slowly may work in Australia's favour. Australia will probably win, but some reputations (and possibly careers) will be built or destroyed on both sides.

India (3rd) 1228.95
Sri Lanka (4th) 1171.12
England (5th) 1164.62
Pakistan (6th) 1104.19
Zimbabwe (9th) 497.45

Idle Summers 11th December, 2008 15:13:44   [#] [0 comments] 

The same thing. But five times more expensive
Russell Degnan

There was a certain level of excitement amongst planning types when the Brumby government announced that the Eddington report was going to be responded to via a whole new transport plan. The hope was that, in light of the extensive public consultation that preceded the EWNLA, and the limited public consultation that succeeded it, that the government was going to do something new. In the past month it has become increasingly obvious that yesterday's release of The Victorian Transport Plan was going to be as dull as its name.

Or at least, as dull as a plan can be that promises to spend $50 billion, but with the exception of a few big roads and a few small rail extensions, not until they've probably left government.

Two years ago, I noted that the, then new, Meeting Our Transport Challenges, was predominantly a shopping list. Earlier this year, Carlo Carli in defending MOTC, argued that the shopping list needed to be bought earlier and extended, in line with unexpected increases in population and in public transport patronage.

And that, in essence, is what the VTP is: a list of projects, some significant freeways sourced from VicRoads never-ending vault of necessary connections, some from the significantly more modest list of whomever does public transport forward planning (if anyone), some from the Victorian Freight and Logistics Council (the occasionally strained reasoning behind which can be found in the also released Freight Futures) and a couple - the biggest ones - from the so-called Eddington Report.

Not surprisingly, almost every option was already on the table in some form, either announced previously, or discussed vigorously. This may be because they are the only reasonable ways of organising Melbourne's transport systems. Or not.

On the question of how best to cross the Maribyrnong, we get two answers. A shortened version of Eddington's road tunnel, from Dynon/Footscray Roads to Sunshine/Geelong Roads, shorter than necessary, and undoubtedly to be later extended to the Western Ring Road. The Eastern Freeway extension has been shelved, for traffic and economic reasons, but will undoubtedly return in a decade. Similarly, the worrying plan to grade separate sections of Hoddle Street is a mere placeholder for a North-South linkage study, and probable plans for an underground inner ring road.

And we'll get half a rail tunnel, connecting Footscray, Parkville, the CBD and Domain, under the ostentatious moniker of building a "metro" system. For $4.5 billion this will improve travel for around 20,000 commuters - about $225,000 per person - but may include a freight tunnel. Uncongesting the rail system is a worthwhile goal, but I'm not sure this is good value. Not when you consider you could subsidise $200 million worth of travel per year, and build 200km of tram track for the same price.

There is much more in the project list, some fairly nice, like grade separations at rail crossings, and some slightly bizarre, like the plan to accommodate new industries using brown coal from 2015. Climate change be damned, but it's only $9 million.

I could go on, but my real criticisms of this plan run deeper than the projects themselves, and onto the planning process itself.


Cycling and walking were, as usual given lip service but no place at the table. There is a promise to release a Cycle Strategy, and a $10 (or is it $100) million increase in funds. But the substantial problem remains. Transport modelling looks at network node connections; transport predictions look at movements across local boundaries. They are, by definition, long trips. Important trips, to be sure, but only half the trips done, and I might add, only half the congestion.

There is little to no mention of strategically planning for shorter trips, beyond gestures at new bike paths, and inner city bike hire (as if only the inner city could be riding bikes). Localised congestion on roads remains the preserve of councils and VicRoads engineers. Whether land-use plans that actively encourage higher densities and concentrated commercial development will exacerbate these problems is also not discussed.

Other departmental documents indicate an absence of understanding what they might be trying to achieve. In the above report, a consultancy and the ABS were paid to generate a table informing us that almost all walking trips start and end in the same statistical local area, that cycling is more common near the CBD, and that demographics matter. Anybody could have told you that without a report. In any case, they only looked at the percentage of cyclists in each demographic, not even the relative percentage of cyclists between groups.

What needs to be asked, is what percentage of people could take a form of transport, and what percentage do. And what are the characteristics of those groups. Because until potential transport choice is addressed at the right scale, planned changes in mode share will be the result of lucky guesses and unforeseeable change, not policy.

Managing road space

Reading a transport plan is invariably grating. Lifeless, shop-worn phrases spill forth, rarely checked by statistical fact or insight, and weighed down by allusions to a better, blander future. And then right at the end, something a bit different. Eddington occasionally went beyond his remit in his report, and none more-so when he argued for congestion charging (amongst other things):

The Government should re-evaluate its current road tolling policy to ensure that the long term benefits of new road investments can be fully realised (including public transport priority, improved cycling opportunities, road network balance and improved local amenity).

Varying the response from carefully reworded support, the government takes on the role of managerial supervisor dressing down an employee for undermining their position. They are also, mostly, wrong.

First, the Victorian Government does not toll existing roads.

This is not the first body to suggest congestion charging, though the VCEC got a more polite response. In no case have they adequately explained why - beyond their fear of leaving themselves open politically. Given Melbourne already has a widely used system of electronic tolling, it is a mistake to think that congestion charging couldn't substantial improve the efficiency of our roads, and provide a much-needed revenue stream for other transport improvements.

Secondly, roads are only tolled if they are beyond budget capacity.

This is something of a furphy, as numerous people have pointed out. The government is at least as well placed to take on large debts, and if it can be financed via tolling a subset of tax-payers, you can rest assured it can be financed by all of them. Secondly, private companies need to buffer themselves with a risk premium that leads to inefficient tolling, and lower than optimal road use. At best, there is some level of fairness in only charging road users, but in that case, why apply it to only new projects?

Thirdly, the Government does not close other roads to force people onto toll roads and won't compromise public transport on or around toll roads.

The first part is a framing issue. Eddington (like myself) argued that roads running parallel to freeways should be downgraded, to improve local amenity through lane closures, wider footpaths, bike lanes, and landscaping. The government interprets this as "forcing people onto toll roads". Road space is a public asset that could be used for many other things, other than funneling traffic from the outer suburbs. For inner city residents, the benefit of any tunnels built through their locale is amenity improvements. To not only deny them that option, but to also frame the debate away from its consideration is disappointing to say the least.

I'm at a loss to understand the second part of the response however. It seems to reply to a single (perhaps poorly phrased line): "ensure that the long term benefits of new road investments can be fully realised" . The commercial implications clause that prevents an airport public transport link may be an enduring sore-point for a government that loves its ribbon cutting, but it certainly wasn't what Eddington meant.

The VTP, like so many before it, continues to under-utilise economic theory in understanding efficient of public space, favouring road traffic indiscriminately, and congesting streets unnecessarily. The clearways plan, while ostensibly to improve public transport flows, will undoubtedly have an induced demand effect, quickly negating any gains. Similarly, unless substantially more infrastructure is built than necessary, any improvement in transport (private or public) will also suffer from induced demand, and resolve to congested conditions.

Integrating land-use planning

The great irony in this problem, is that under "What you told us" the government heard that we wanted "ongoing integration between transport and land-use planning". Unfortunately, a few glib comments aside, there is little evidence for it.

The recently released Melbourne @ 5 million had two core components: six Central Activities Districts at Box Hill, Broadmeadows, Dandenong, Footscray, Frankston and Ringwood; and three employment corridors through the outer west, from Caulfield to Dandenong, and from Monash to Heidelberg. Both are laudable enough by themselves, but they need substantial support to work. There is no mention of that support, nor, to the extent that some projects would support, no explanation of how they are supporting this plan.

Transport to any area of the city can be defined by two things: its accessibility and its capacity. The CBD is very accessible - a large population of people can get there quickly - coupled with a high capacity. Although train lines and (perhaps one day) SmartBuses run through the six CADs, they are otherwise accessible mainly by automobile. This is a problem, because capacity is quickly (and already) reached using automobile traffic. Those centres and employment corridors will almost certainly need substantial infrastructure improvements. I say almost certainly, because in the absence of any targets for growth, and in the absence of any transport needs assessment, we don't actually know. The absence is, itself, a clear indication that there is no ongoing integration between land-use and transport planning.

If anything, the building of the rail tunnel demonstrates a clear repudiation of last week's land-use plan, by massively increasing the accessibility and capacity of the CBD, and propelling it onwards in its dominant trajectory. A dominant CBD is not necessarily a good thing. It leads to longer commutes, disparity of wealth and services, and needs more expensive infrastructure to fight congestion. Economically, it may be worth the expense, but when a land use plan says one thing, and a transport plan another, it isn't integration.

Planning for greater efficiency

This is an old school plan in many ways. There are lots of words (and advertising dollars) spilt on current transport fads, but fundamentally it is about big road and rail projects. SmartBuses have been downgraded, trams extensions are spoken of in the past tense, cycling awaits it strategy. But the big people movers are planned, and awaiting implementation.

Doing something with our rail system is, without question, a good thing. Because, as Paul Mees never fails to point out, the operation is a mess. The rail system grew organically, and it grew, for the most part, a long time ago. But because of this, there are lots of niggling issues that affect operations, and prevent it getting anything like best practice. Single lines in unfortunate places and numerous level crossings are being removed, if sometimes slowly, but there is a wider problem. Widely spaced and poorly integrated outer suburban stations run reasonably fast with express trains, but are under-utilised for short trips. Closely spaced and very slow inner suburban stations with trains stopping all stations are well used, but get in the way of express trains. Add in regional rail and a plan to introduce more freight traffic and this plan will do little to eke out extra capacity.

There needs to be plans put in place to move forward. Inner city metro trains in Europe run across short distances, and therefore stop often, but we have an unstable mix in a vast city. Moreover, there is substantial doubling up on several lines, with trams running parallel with trains, both probably subsidised when they could run a profit. Integration of services also needs a clean separation of goals. The rail system, uniquely able to carry large numbers of people very quickly, should be configured to do just that, but it needs careful, long-term planning, not a few big projects.

How I would do it can wait for another post (shortly), but using the train system as we do is an analog signal in a digital world (albeit with a smaller bandwidth jump).

Federal funding

This is, to me, one of the strange oddities of the report is the insistence that projects will go ahead, contingent on Federal funding. There are certainly benefits to saying that. It allows the government to pull out of projects for political reasons, then blame the commonwealth for lack of funds. And it takes debt off the books, which allows them to show a balanced budget.

But unless the fund provided from Building Australia Fund have been quarantined in relation to the Commonwealth Grants Commission (and googling this for an hour provided no indication either way), then the money will never really exist. An extra billion on the state budget will be treated as revenue by the CGC when it comes time to divide GST revenue between states. Because all states will get some infrastructure money, and it will be handed out over a number of years the impact won't look large, but it will exist.

And hence, nor will these projects really be funded "by the Commonwealth". Any budget shortfall as a result of an increase in assessed revenue needs to be paid for via increased borrowing, increased taxes, or decreased services, the same as it normally would. The difference is that (most) Commonwealth funding appears as lower "income", not as an expense (payment of debt). A difference in political terms, but not on the budget. It is wrong to pretend that these projects are somehow dependent on government funds; almost as wrong as abdicating responsibility for transport infrastructure to federal control in fact.

Sterner Matters 10th December, 2008 13:33:06   [#] [0 comments] 

Monday Melbourne: CLXVIII, December 2008
Russell Degnan

Federation Square Atrium. Taken December 2004

Melbourne Town 10th December, 2008 13:31:37   [#] [0 comments]