Do new freeways signify progress?
- Share
Living in the new metropolis - slide show (click)
According to Federal Opposition Leader Tony Abbott, “almost nothing signifies progress more than new roads.” That line is from the speech he gave to the Federal Council of the Liberal Party on Saturday when he announced the Coalition’s Plan to Reduce Traffic Congestion.
The Coalition’s Plan comprises a $4 billion commitment to accelerate construction of three major freeways – Melbourne’s East-West Link, Sydney’s M4 East, and Brisbane’s Gateway Motorway upgrade. These projects should be underway, he says, within twelve months if the Coalition wins office. Further road priorities will be announced before the next election.
Mr Abbott says these freeways are needed because “businesses and consumers are paying more in transportation costs and our cities are becoming less productive.” He’s picked up on the ‘commuting undermines social capital’ meme too, arguing that traffic congestion means parents are away from their families for too long (hmmm, how prescient am I then?).
Now I’m not automatically against all freeways. Cars will be with us for a long time yet. Freeways can potentially remove through-traffic from residential streets, make freight and buses more efficient, and serve trips that can’t easily be captured by public transport.
There can be a place for certain freeways provided they fully recover their costs (and preferably are part of a broader road pricing scheme), the risks to taxpayers are small, and the ratio of benefits to costs is not only positive but better than the realistic alternatives. They should also be accompanied by strong policies to improve the fuel and environmental efficiency of vehicles.
However there are some evident flaws in Mr Abbott’s plan. The idea that building freeways is an effective way of addressing traffic congestion is the obvious one.
Expanding road capacity increases the volume of vehicles that can move between two points even in congested conditions (and that can have value), but it won’t eliminate congestion. There’ll be an initial period of relief but inevitably traffic expands to fill the extra capacity until a point is reached when speeds slow markedly.
Another flaw is it’s not clear if these three projects are even good ones. As the Grattan Institute points out, most of the low hanging fruit in terms of urban transport projects has already been picked. Certainly Melbourne’s proposed $5 billion East-West Link looks bad – Sir Rod Eddington calculated it has a BC ratio of just 0.7!
A key reason the benefits are lower than the costs is a lot of very expensive tunnelling is involved. That implies high tolls would be needed to recover construction and operating costs. Recent experience with failed road tunnels in Sydney and Brisbane suggests private investors won’t be prepared to carry all or even most of the risk.
The East-West Link might consequently have to be funded by the Australian and Victorian Governments (even if the latter effectively “pays” private lenders). Mr Abbott says he’s had “discussions” with the Victorian Government, but just how Mr Baillieu would find at least $3.5 billion (but probably considerably more) isn’t clear.
Nor is it apparent if other projects, like the proposed Melbourne Metro rail tunnel or level crossing eliminations, would have to be delayed to enable the freeway to proceed.
Another issue is whether or not the East-West Link would be tolled. Infrastructure Australia favours road projects that are tolled but it’s not clear if an Abbott Government would continue that policy. Voters don’t like paying tolls and, while Governments don’t like paying either, it’s usually easier to ‘splash the cash’ than offend voters.
Mr Abbott says he’s had “discussions” with Infrastructure Australia too. The Victorian Government is undertaking more detailed work on the business case for the East-West Link and doubtless there’s a lot of effort going into “refining” the BC ratio. I’m disappointed Mr Abbott would commit $1.5 billion of public money to a project whose worth hasn’t yet been established.
All in all, I think Mr Abbott has announced a roads policy, not a congestion policy. And it’s got the hallmarks of populist policy, not rational policy. It reminds me of the same cynical pitch the Greens made to voters at the 2010 Victorian election.
But looking at all of Mr Abbott’s speech, I suspect it’s not only populism. We could be seeing the beginning of a big shift in thinking about cities (and other things too, but cities is of course my focus). The conservative side of politics at both the state and national level may be poised to reject many of the established verities of urban policy.
Please login below to comment, OR simply register here :
Thank you for registering, we have just sent you a confirmation email, which includes your new password to be entered below.
The roads should be tolled, at a rate to amorize their cost of construction and maintenance over their estimated life. If the expected toll is too high, and traffic deterred so that the receipts will be too low to cover this, then the roads should not be built.
I don’t know too much about the other roads, but have had experience with the Gateway Motorway. That was a fairly reasonable road UNTIL the Queensland Roads Authority started an upgrade – which meant chaos. Delays were rampant and journeys from the Gold Coast to the Airport took up to three hours – twice as long as the train! The alleged benefits are taken into account, but rarely, it seems, are the cost of delays to traffic while the road is being “upgraded”. Congestion on the Gateway Motorway could have been removed by tolling the motorway as well as the Gateway Bridge, to the benefit of everybody.
This is a daft promise, and I hope that Tony realises it and deletes it before it becomes enshrined in Liberal Policy. Of considerably more benefit, and rather cheaper, would be all the suggested extensions to Melbourne’s tram system, and the Sydney Parramatta Road/University, the Randwick/University and the North Sydney tramways.
Parker Alan • OAM
Note that in the Netherlands freeways and heavily trafficed main roads are safer because all have parallel or protected footpaths and bikeways. All freeway and main road bridges have separate footpaths and bicycle paths. Added to that all the residential streets have 30 km/hr speed limits and main roads with bicycle lanes have 50 Km /hr speed limits. Australia needs to follow the Netherlands Road safety model and so does the rest of the world.
The Netherlands has a bicycle friendly road systems for all road users and a road death rate per 100,000 population in 2010 that is one half that Australia. That means around 600 road deaths on Australian.
About 1.3 million people die each year as a result of road traffic crashes worldwide. Nearly half of them are “vulnerable road users”: pedestrians, cyclists and motorcyclists. Furthermore, the world’s fleet of passenger cars peaked at 1 billion in 2010 and killed most of the 1.3 million people. The World Health Organization believes that road safety generally and bicycle safety in particular need to given priority in all countries .
Road safety from 1970 to 2010 has greatly improve in Europe bicycling has become much safer in Japan, Switzerland, Germany, the Netherlands and Scandinavia because of their innovative bicycle planning and intermodal bicycle/public transport planning practices.
Australia has the potential to greatly reduce bicycle accidents while increasing bicycle use for trips to work, for business and for recreation while providing healthy exercise. The bicycle can also expand the public transport system by providing 3000 secure bicycle parking at railway stations in Melbourne CD each year for the next decade.
Per-capita metrics for road fatalities don’t really make sense. You should use something like deaths per KM travelled.
The rate of development of health benefitting infrastructure favouring active transport is the most meaningful indicator of ecological progress in the urban environment.
Of course, no one would argue that well designed freeways will not reduce fatality rates for freeway users.
However most urban dwellers need to have their transportation needs met by the rest of the transport system which handles well over 90% of urban trips.
I’m not usually in the habit of citing something from the dark side but it is obvious to anyone that Abbott is just making up “policy” as he wafts along on a tide of his own hot air.
I should have made this observation on previous post, but did you notice that in that Google Map of Houston, more than one third of the surface (more if you count the roads themselves) was car park? That is your typical Sun Belt city for you and an example of where the roads-only policy leads. Houston has the fifth worse traffic congestion in the US despite having (over) built more roads than just about any other major city (apparently Kansas City holds the record–adjusted for population size).
As Michael Owen has described (in his book Conundrum; he gave a lecture in Melbourne earlier this year) the size of car parks is often the biggest barrier to walking in these places. He described staying in a top-name hotel in Phoenix which was spread over a vast area, and how on his first day he started walking from his room to the main hotel office and soon regretted it: he had to walk over bituminized carparks with no protection from the sun such that his shoes started sticking to the almost melted tarmac and he was sweaty and parched by the time he got there. He understood why everyone drives from their room to the hotel lobby–in their cars that are filling the role described in the south as “mobile air conditioning”!
Even Texas, however, is having some painful second thoughts about the policy of cars-to-the-exclusion-of-all-else. They have a HSR project linking DFW, Houston (and Austin, the capital, if I remember correctly?) and were hoping to get some of Obama’s HSR funds. It makes a lot of sense because of the large population within quite modest distances of each other, and the increasing difficulty of roads filling the transport needs.
In my previous post above, when I wrote “They (Texas) have a HSR project linking DFW, Houston …” I meant they have a plan, with who knows what probability of ever getting up. Though Texas is often thought of as exclusively redneck territory, in fact it is a curious mix of progressive-Democrats and neo-Libertarian Republicans. The latter will fight a furious battle against any spending of taxes on anything but roads, however one could not rule it as impossible, especially as any such projects anywhere in the nation, rely upon Federal money. Indeed a lot of the roads were built with Federal subsidies.
Parker Alan • OAM
Hi DR SMITHY,
How to measure mortality rate and fatality risks? I agree that it depends upon
what one uses as a measure of exposure to risk: population, registered vehicles, distance travelled. There is expert opinion on that.
Fatalities per 100,000 population is preferred by the OECD road safety research body IRTAD and the WHO.
There has been considerable debate in the past about that. Those in the health sector prefer the use of population as the denominator, since it permits comparisons with other causes of injury or with diseases. As the health and transport sectors increase their level of co‐operation, fatalities per 100 000 population are becoming more widely used. In the transport sector it has been common, where data are available, to use. In the past
fatalities over distance travelled has traditionally been favoured by road
transport authorities as it implicitly discounts fatality rates if travel is increased.
However fatalities per 100 000 population is now the denominator the most often used, as it expresses the mortality rate or an overall risk of being killed in traffic for the average citizen. It can be compared with other causes of death like heart disease,HIV/Aids, etc. It is a very useful indicator to compare risk in countries with the same level of motorisation.
Fatalities per 10 000 registered vehicles can be seen as an alternative to the previous one, although it differs in that the annual distance travelled is unknown. This indicator can therefore only be used to compare the safety performance between countries with similar traffic and car use characteristics. Indeed In some countries, scrapped vehicles are not systematically removed from the registration database, undermining accuracy.
michael r james @ #5,6,7:
For all its (many) failings, The Australian has some virtues – this is an example of one. Fairfax didn’t pick up on this issue.
Re the Houston image, that’s why I used it – to show the parking lots.
Tony Abbott’s book ‘Battlelines’ has typical right wing memes that cars = personal freedom = good, public transport = central planning = bad.
So I wouldn’t hold my breath waiting for sensible urban transport policy from the Liberals.
More importantly, it measures against something that’s actually relevant to the risk of dying in a car.
I do not understand your implied criticism. If more people are driving further, but the same number of them are dying, why *shouldn’t* the fatality rate be “discounted” ? Similarly, if ten thousand new immigrants arrive and none of them use a car, why *should* the road fatality rate be “discounted” ?
No, it doesn’t. More accurately, it requires the reader to know in advance whether or not the two countries have “the same level of motorisation” and thus whether or not the metric is meaningful or useless.
I suppose per-capita numbers could be useful within the same country for measuring the risk of, say, heat-attack vs car crash, but the user of per-capita metrics for measuring road fatalities *across countries* is disingenuous at best, deceptive at worst. It misleads about the relative risks of driving in one country vs another.
Parker Alan • OAM
Hi drsmithy,
It has also been traditional to fail to produce to measure distance traveled by bicyclists and walkers and people like you will continue to discriminate again bicyclists and walkers. In the Netherlands they measure bicyclists and pedestrian exposure but not in Melbourne or Australia.There is expert opinion on that.
Fatalities per 100,000 population is preferred by the OECD road safety research body
and the WHO. There has been considerable debate in the past about that and there is a consensus that deaths per 100,000 population. Lets end this blatant discrimination.
“People like me” ?
This kind of juvenile, pointless and irrelevant ad hominem completely destroys any credibility you might otherwise have. Take your trolling elsewhere.
If more people are driving further, but the same number of them are dying, why *shouldn’t* the fatality rate be “discounted” ?
Because that would imply that if there was a 30% increase in the distance people were travelling, but only a 10% increase per capita, in the number of deaths as a result of car accidents, that progress had been made. However it ignores many important questions. Why are people driving 30% further? Is it simply because poor planning outcomes have resulted in the sort of city that people can’t easily walk or cycle and therefore drive? If that’s the case then I’d argue both the 30% longer distances, and the 10% increase in road fatalities per capita are negative consequences, of the sorts of policies that favour road building. Why use an measurement that makes that look like there has been an improvement in conditions?
Similarly, if ten thousand new immigrants arrive and none of them use a car, why *should* the road fatality rate be “discounted” ?
For similar reasons to above. How are these new immigrants getting around? Are they not being killed because the communities they live in are well planned and therefore safer?
Per capita indicators raise some of their own questions, but per km travelled indicators almost always favour modes which cover long distances, which inevitably means favouring the car as it is the dominant longer distance mode in virtually all the developed world. It therefore doesn’t make sense to use this indicator, unless you’re looking to measure something like “how safe is this freeway, compared to that freeway” not “how safe is city X’s transport network compared to city Y’s”.
It ignores them with good reason – they’re not relevant. How many other people there are in the country has no direct causal relationship to your risk of death or injury when driving. If you have trouble understanding why, look at the two extreme scenarios: where you are the only person in the country that drives, and where every single person in the country drives.
Because if more people are driving further [in aggregate] but deaths [due to driving] haven’t increased, that *is* “an improvement in conditions” when the relevant measurement is risk incurred by driving somewhere.
Once again, to understand the difference, consider extreme scenarios. In which scenario do you think you have less risk driving: when you are the only driver in the entire country, or when everyone in the country is on the road at the same time ?
But these things are irrelevant if you want to compare the risks of *driving* between different locations. Which is, after all, the only reason you would be comparing the fatality rates of driving.
No, per-km metrics are simply using a measurement that’s actually relevant.
Another example: why should a baby, or a housebound pensioner, be considered to have any relevance to the risk of travelling by car (or bike, or train, or anything else) ? What *causal* relationship exists that needs to be accounted for ?
You have it backwards. This is exactly why a metric like per-km must be used, for the reasons I’ve already highlighted – why should people who never uses any form of transport (car, bus, train, bike, whatever) influence the statistics that are measuring the risk of using those forms of transport ? If you have two countries – one with a stupidly dangerous transport system that hardly anyone uses and another with a quite safe transport system that nearly everyone uses, you could end up with quite similar per-capita fatality rates, but a vastly different level of risk to actually use their transport systems.
There are really only two metrics that make sense when you want to measure the relative risks of different transport methods – how far you go, and how long you’re using them. How many people you’re sharing the country with has zero direct influence, and using it is largely nonsensical.
No, those questions are relevant. Your logic is flawed and implies that just because average journey’s are longer, they are somehow safer.
As you completely ignored the numbers in my previous example, take this scenario. Two cities, exactly the same populations, exactly the same housing numbers, job numbers, universities, schools, etc. Exactly the same numbers of toddlers and house bound pensioners (nice that you switched from the immigrants, who clearly still travel). One is a compact city, three times as dense as the other, which is a sparse city. Well call them Compo and Sparso. As a result of the dispersal the average trip distance in Sparso is three times as long as on Compo, the number of road accidents and road mortalities are also double in Sparso compared to Compo.
Which population is safer on the roads? By any sensible metric you’d see that people are twice as likely to die or be involved in an accident on the roads in Sparso, and therefore the population of Compo is safer. However by the per km metric because there is only double, and not triple the numbers of accidents and deaths, Sparso comes out a head, simply because Sparso is more than three times as spread out.
Its rarely relevant to know how many people per a given distance. What is relevant how many people die on a a particular journey (to work, to shops, etc), or how likely on a given day, under average travel needs they are to be involved in an accident or suffer a fatality. What distance they travel is usually a result of the urban form, and the travel options given to them.
If you have two countries – one with a stupidly dangerous transport system that hardly anyone uses and another with a quite safe transport system that nearly everyone uses, you could end up with quite similar per-capita fatality rates, but a vastly different level of risk to actually use their transport systems.
If that was the case I’d ask how everyone is managing without travelling in the country with a dangerous transport system? Does their urban form limit the need to use their transport system and travel long distances? If so, then the country with the safe but heavily used transport system should look at what they’re doing, whilst they should look at making their own transport system safer for the few that do need to travel.
If you can elaborate a bit on what’s “flawed” about the logic that says if you can travel (say) 150km without death or injury in locale A, but only 100km without death or injury in locale B, that makes locale A safer, I’m all ears.
Yes, it’s easy to construct an example to support your conclusion when you appropriately manipulate the variables to do so. That doesn’t make it valid.
That is to say you’d ignore the point and go off rambling about irrelevant issues.
Look, if it’s not immediately obvious why measuring per-capita isn’t really valid, then I doubt there’s anything I’ll be able to write that convinces you. However, this page covers the reasons why, as does this one from the OECD (it’s focused on driving, but the principle is the same):
“Fatalities per million inhabitants can be compared with other causes of death in a country (heart diseases, cancer, HIV, etc.) however when comparing countries road fatality risks, this indicator looses it relevance if countries do not have the same level of motorisation. Fatalities per billion vehicle-kilometre provides a better measure of fatality risk on road networks, but there is currently no harmonisation in the methodology to calculate distances travelled, and not all countries collect this indicator. “
Yes, it’s easy to construct an example to support your conclusion when you appropriately manipulate the variables to do so. That doesn’t make it valid.
Aherm…
If you have two countries – one with a stupidly dangerous transport system that hardly anyone uses and another with a quite safe transport system that nearly everyone uses, you could end up with quite similar per-capita fatality rates, but a vastly different level of risk to actually use their transport systems.
In which scenario do you think you have less risk driving: when you are the only driver in the entire country, or when everyone in the country is on the road at the same time ?
You’ve laid out quite unrealistic scenario’s, yet you’ve ignored my vastly more likely scenario’s, then claimed I’m the one creating scenarios that aren’t valid? There are lots of cities with low and high urban densities, yet similar populations, job opportunities, etc. Where are the mythical countries where nobody travels?
Also, the OECD article you’ve linked to does not support what you are claiming it does. The NMA article is talking about discussing highway safety. Which, like I said is about the only valid use for it: It therefore doesn’t make sense to use this indicator, unless you’re looking to measure something like “how safe is this freeway, compared to that freeway” not “how safe is city X’s transport network compared to city Y’s”.
If one can forget this squabble over how to measure the slaughter on the roads, one might go back to the full sentence Tony said: “Almost nothing builds confidence more than seeing cranes over our cities and almost nothing signifies progress more than new roads.”
Now that is a beaut reference to Joh Bjelke-Petersen’s quip about how he measured the economic health of Queensland, but one must admit that what he said is believed by a very large number of people. Note, he said “new roads” not “new freeways”. And new roads includes new suburban housing estates, new roads to mines, new roads from rail yards to ports, and many more, not just freeways. So in a sense he is right. New roads do signify progress. And that means that all thinking people have to carefully consider what new roads they would like, and what is the price to be paid for them. $4000M and how many lives? Because journeys rarely start and end on motorways (a much better term than ‘freeways’ – which term is likely to be misinterpreted as ‘toll-free’ where as it means ‘free-flowing’). So additional journeys as a result of new motorways means additional journeys on the feeder and distributor roads, to the detriment of people! Be warned!
Yes. Absolutely. I quite specifically chose extreme scenarios to demonstrate why the fundamental reasoning is wrong. The whole point of the example I used was to take an extreme case and demonstrate tjat your logic breaks down because using the wrong measurement can lead to identical answers (same per-capita fatality rates), but completely incorrect conclusions (that the same trip in the two different locations carries the same risk).
My claim is that per capita metrics are invalid for comparing travel risk (regardless of travel method) between areas with different transport systems. The OECD article absolutely supports the rationale behind my assertion.
No, it makes perfect sense to use it regardless of what the method of travel is, because it’s measuring the risk against a variable that is directly related to the risk of travelling. This is contrast to measuring it against the number of people, which has an indirect (at best – potentially completely irrelevant) relation to the risk.
Once again, the number of people in a country has no direct influence on travel risk. As such, per-capita is an invalid metric to measure travel risk against between locations with different transport systems. Per-km-travelled (or per-hour-travelled) does have a direct relation to travel risk, therefore it is an appropriate metric to use.
If you want to measure the risk of dying of a heart attack or mugging vs a car or train crash within a single population, then per-capita is reasonable. If you want to measure the risk of dying in a car/bus/train/bicycle/boat/blimp/whatever between two separate populations, then per-capita is meaningless, and you need to use something that is relevant, like per-km-travelled or per-hour-travelling.
Yes. Absolutely. I quite specifically chose extreme scenarios to demonstrate why the fundamental reasoning is wrong.
Just as I did, except I didn’t need to point to hypothetical examples that were completely unrealistic, because realistic examples already exist and demonstrate the fault of logic adequately, i.e. residents of Sparso, still twice as likely to die on the roads than the residents of Compo regardless of the fact that Sparso’s roads are safer on a per km metric.
I don’t know what ideological barrow you have to push that requires using per-capita metrics to justify, but the simple fact is that they will not inform you accurately about transport safety in differing locales. Which is why nobody involved in road safety uses it as a measure of success (or lack thereof).
Here is a real-life example. Denmark and Estonia.
Denmark
Population: 5.6m
Cars: 2.2m
2.5 people/car
0.4 cars/person
406 fatalities/yr
49,250,000,000 total km/yr
8800km/person/yr
22386km/car/yr
Road fatalities per 100,000 people: 7.4
Road fatalities per 100,000 vehicles: 14
Road fatalities per billion vehicle-km: 8.2
Sources: CIA world factbook, http://www.infrastructure.gov.au/roads/safety/publications/2009/pdf/rsr_05.pdf, http://en.wikipedia.org/wiki/List_of_countries_by_traffic-related_death_rate
Estonia
Population: 1.3m
Cars: 550,000
2.4 people/car
0.42 cars/person
101 fatalities/yr
5,600,000,000 total km/yr (derived)
4307km/person/yr
10181km/car/yr
Road fatalities per 100,000 people: 7.5
Road fatalities per 100,000 vehicles: 34
Road fatalities per billion vehicle-km: 17.5
Sources: CIA world factbook, http://www.mnt.ee/public/statistika/Maanteamet_AR_2011_ENG.pdf, http://en.wikipedia.org/wiki/List_of_countries_by_traffic-related_death_rate
Which country do you think is safer to go on a driving holiday in ? Denmark or Estonia ?
Parker Alan • OAM
drsmithy still persists in taking an anti public health approach to road safety.
Road traffic Injuries now pose a public health crisis that requires urgent action at at a national an international level. That is why the Director General of the WHO in 2004 said.
“ Too often, road safety is treated as a transportation issues. Not a public heath issue … many countries spend far less effort into understanding and preventing road traffic Injuries than they do to understanding and preventing diseases that do less harm”
This view is soundly based on data from selected bicycle friendly EU countries which have the following 2010 road death rates per 100,000 population: , Sweden 3.0, Netherlands 3.7, Japan 4.3, and Germany 4.7, Denmark 4.5, Switzerland 4.5 France 6.1. Australia’s death rate is higher (6.2) and the US death rate of 10.5 is even higher. The data shows that the US and Australia are not bicycle or pedestrian friendly and have much safer road transport systems for all road users’
This why The European Parliament adopted a resolution in 2011 that “strongly recommends the `responsible authorities to introduce speed limits of 30 km/hr in all residential areas and on single lane roads in urban areas which have no separate cycle lanes “ This resolution is is part of a wide range of measures to halve Europe’s 31,000 annual road fatalities by 2020. (Kock Report 2011)
The latest information from Europe is that a 30 kph limit on local roads and main roads in busy urban areas heavily used by walkers are much safer for all non-motorised users, motorised wheelchairs and electric bicycles.
It is very clear that on the basis of kilometres ridden by bicycle in the Netherlands is still safer even though no one is compelled by law to wear a bicycle helmet. The Dutch government believes that the 30 kph limit is of great benefit to all non-motorised users”
It is assumed that riding a speed limited electric bicycle is as safe as a bicycle given the existence of a lower speed limits.
No, I simply like to measure it in a way that’s both relevant and useful.
The rest of your comments is just a verbose straw man.
Please login below to comment, OR simply register here :
Thank you for registering, we have just sent you a confirmation email, which includes your new password to be entered below.