I have just started reading the book “Still Stuck in Traffic: Coping with Peak-Hour Traffic Congestion” by Anthony Downs, and it promises to be quite an interesting insight into what causes traffic congestion, and what can (and can’t) be done to alleviate it. While I have just started the book, and therefore it’s likely there will be future posts as I make my way through it, I just can’t help but quote an excellent tract on the futility of trying to build your way out of congestion by expanding the capacity of the road network.
[A] major approach would be to immensely expand the capacity of the road system. True, a basic tenet of this book is that a region cannot “build its way out of congestion” once peak-hour congestion has appeared there. Yet in theory, enough additional road space could be created so that all those who wanted to drive rapidly during peak hours could do so simultaneously. But in reality, there are too many people seeking to use the roads at the same time each day for this approach to work without enormous financial and environmental costs. key routes would have to be widened so much that huge portions of the entire region would be turned into giant concrete slabs. This would destroy thousands of properties along present roadways, wreaking havoc with trees, open space, and many other aspects of the physical environment. It would also be enormously expensive. In fact, no society could afford all the costs involved. And after all that road space was built, much of it would stand empty most of the time when far smaller percentages of drivers wanted to move. This approach is equivalent to building a big enough parking lot in a shopping center to hold the cars of all the consumers who want to shop during Christmas week and then watching that lot stand empty most of the year. Thus expanding road capacity enough so that peak hour rationing of road space is not necessary is totally impractical.
Another difficulty is that greater capacity might encourage many people who have currently chosen not to travel on roads during peak hours to start doing so. In the long run, it might also encourage more urban development in the region as developers take advantage of the greater mobility provided by larger roads. These forms of “induced demand” would offset much of the hoped-for gain in capacity from building more roads.
Why is this idea so difficult for road engineers and politicians to get their heads around? Why are we going to spend $800 million on widening the Northwest Motorway when it’s only ever congested at peak times, under the belief that adding an extra lane to each side of it will somehow avoid being subject to induced demand and end up just as congested as it is now within a few years?
Interestingly, Anthony Downs also argues that expanding public transport capacity will also struggle to reduce peak hour traffic congestion, while road pricing is currently impractical, and likely to remain inequitable and politically unfeasible. I await with interest to see what he does suggest that might make a difference to congestion levels, but in the meanwhile I think it’s useful to keep in mind a comment made by Jarret Walker from Humantransit.org, that perhaps we’re looking to solve the wrong problem – reducing congestion rather than improving accessibility. I tend to agree with such a statement, that after all what improving transport is all about is increasing accessibility – making it easier for people to do what they want to do. And that’s where public transport investment, particularly investment in public transport that has priority over congested traffic (be it bus lanes, railways lines or busways) can make a difference. The roads might remain just as congested as before, but if people now have a faster and more convenient alternative, then we have made a positive difference.
That’s a fundamental reason why I think the solution to Auckland’s transport problems must come through better public transport rather than more road building. Building more roads is unlikely to significantly improve accessibility, because the gains of increased capacity will be lost to induced demand. However, because public transport is inherently more space efficient, improvements to public transport will give people real and long lasting accessibility improvements.
I don’t think anyone believes that peak time congestion CAN be fixed. Congestion is nature’s way of incenting people to travel off-peak if they can.
What adding lanes to existing roads or building new roads can do is increase the number of vehicles that are needed to create congestion. In that case, Auckland’s motorways have been spectacularly successful. They’ve allowed Auckland to expand from 300k people in 1950 when the first motorway sections were built to its current population of 1.4 million. I don’t think that could have been achieved with public transport alone since a large percentage of trips aren’t convenient by public transport. I think a proper rail system will be essential to support future population growth, but it needs to complement road upgrades rather than replace them.
This all assumes that urban growth (or “sprawl”) is a good thing. Most commenters here have either chosen to live in Auckland or were born in Auckland and have chosen not to move to a smaller city. This is likely to be because big cities offer more opportunities and better services than small ones. Commenters seem to have mixed feelings… I think most people are opposed to sprawl, but on the other hand there is support for Hamilton to Auckland commuter rail which would essentially expand Auckland’s sprawl out to Hamilton.
Obi, Steven Joyce seems to think that congestion can be fixed through building motorways in Auckland: http://transportblog.co.nz/2010/05/06/rons-in-parliament/
There is a bit of an argument that you have to “run fast to stand still”, although Braess’s Paradox suggests that expanding capacity can actually make congestion worse. The Downs-Thomson paradox also suggests something similar.
“This approach is equivalent to building a big enough parking lot in a shopping center to hold the cars of all the consumers who want to shop during Christmas week and then watching that lot stand empty most of the year. ”
Um, pretty sure this IS Auckland’s parking policy…
So true….. when will we learn?
After being in the states for a week or so it makes me realise how much better we actually are with parking, especially in town where at least there is a charge for it. In Vegas there every hotel has huge parking buildings that are completely free for anyone to use, they also have their own private roads that in some cases are more than 3 lanes wide in each direction. The actual roads themeselves are massivily wide, in some cases they are around 14 lanes wide with about 9 in one direction and 5 in the other, it makes it pretty hard to navigate when you are on one side of the road then all of a sudden you need to be on the other. Traffic did seem to flow though.
I think of Vegas as everything I don’t like about the US all thrown together in one place.
Despite all of that it is still a great place to visit.
We will never “learn” (i.e. change permanently) because it is a conflict between short-term gain and long-term pain, and between personal gain and societal gain. And since we humans look for the short-term and personal gain first, we will always have to advocate and fight for the better solution. There won’t be some process or period after which there will be a broad consensus for the better outcome, and all our transport and settlement projects automatically end up on the good end of the scale of sustainability.
May sound cynical, but isn’t – because you CAN win that fight. You just can’t expect it to STAY won.
I think there might be an exception to that, which is the possibility that the long term ‘unsustainability’ filters down into direct short term impacts on peoples daily lives. Perhaps the most likely example would be an ongoing increase in the cost of oil fuels without any suitable mass replacement, this would lead to a step change in transport pretty quickly. For example if petrol prices trebled to $5 a litre and electric cars etc were still ephemeral then transport planning, policy and behaviour would shift almost overnight.
Jeff Rubin: “The market can take us to some pretty Green places”…
Yes, but what I was talking about is that we may well overcome the current unsustainable practices (win THIS fight).
But then somebody will invent this crazy teleport that everyone just goes crazy for because it is so fashionable to eat breakfast in Tokio and then an hour later go shopping in Paris. Meanwhile, some scientists make noises that all that teleporting is slowly wearing away our dimensional lining, and soon we will have hyperspace bleeding through, wearing away our reality. The scientists are hushed up or disclaimed as alarmists by the teleport industry, and the average punter goes “Ah, can’t be so bad – I certainly don’t want to miss my evening sauna in Finland just because of some crackpots with unproven theories…”
Okay, this was intentionally farcial – but the point I am making is that we will end up screwing up something else and have to start solving the next mess as soon as we have sorted this one out
You paint such a happy picture of the world Max, I’ll wake up a nihilist…
Don’t you go disparging my teleport idea, I’m sure placing one in Britomart and the Uni Library will have an excellent BCR…
Congestion is a way of rationing a scarce resource. Most should know the historical experience in doing that. Imagine if you had to queue to use electricity, phone lines or buy food, or in fact think about where and when this happened.
Pricing is a better way to ration such resources. Concerns about those who cannot afford can be addressed by other transfers, but severe congestion can be fixed.
Moderate congestion is efficient, it is seen in queues at shops or airports and the like (or the half hour long peaks in a handful of spots in places like Napier or Nelson), severe congestion is when a network no longer fulfills it core function efficiently (traffic moving slower than cycling or walking).
With efficient road pricing, new road capacity will still be built, but only when congestion is over extended periods and road users are willing to pay the higher price – in which case as long as the new capacity is affordable from the long run revenue generated from the road users, it is justified.
There are a couple of other issues that are important to consider when thinking about road pricing (aside from equity concerns), and they are practicality and efficiency.
In terms of practicality, the question is how you do it. If you just price main arterials then people will find other routes, pushing more vehicles onto local roads. This was the fatal flaw of the idea to toll the Western Ring Route. So you technically would have to price each and every road, which sounds like a practical nightmare.
The second issue is efficiency. How do you avoid half the money raised disappearing into admin costs? Or if you’ve got an “all roads” system, how crazily expensive would that be to operate?
Both those issues could be solved by technical GPS solutions which track when and where your vehicle goes, but that sounds all a bit “Big Brother” to me.
You price all roads, on a distance basis. It is not a practical nightmare, RUC does it now, albeit very simply (but EROADS has a more sophisticated unit). The Swiss do it now for trucks with a slightly more sophisticated unit. All that needs to be done is for a GNSS based unit on board every vehicle, that measures distance correlated between a digital tachometer and on board map. It calculates the charge and spits out the total regularly every time the vehicle is refueled via a mobile data connection. The units would cost around $300 each, but could be built in to all new vehicles to start off with, and so would be optional (either pay fuel tax or distance charge). None of this is technically difficult at all.
The Netherlands was about to do this until the general election was called there. The project manager for this project told me around a month ago that operating costs were estimated to be 7.6% of total revenue. The goal is to drop it to 5%. Nevertheless, the economic benefits are enormous. The UK road pricing evaluation study calculated that congestion could be halved, and the economic benefits were in the order of tens of billions per annum. In NZ it is probably more like hundreds of millions.
The biggest misconception is that GPS tracking is needed. GPS is a broadcast signal from a constellation of satellites, the GPS receiver matches those signals and figures out where you are. The unit doesn’t communicate with anything, but for road pricing all that is needed is enough to determine what distance has been travelled and on what types of road (classified for charging purposes). All that need be communicated is the charging data itself, not the real time location (who cares?). Privacy is protected by leaving the basis for charging data in the unit, but allowing it to be interrogated IF you dispute the charge. Interestingly one US study showed that many people WANTED a breakdown of their trips with one road pricing trial, like having a phone bill and knowing why you are paying. Frankly, as long as there is adequate data protection in place, and people have the choice, it shouldn’t be an issue.
That is a good point, it would be similar to a phone bill and I don’t have a problem with Vodafone keeping track of who I’ve called as long as that information is between me and them… Sensible regulations would ensure that…
How are fuel taxes paid at the pump not currently a per-km travelled form of road pricing..? Or is the focus on charging various levels on specific roads at specific times rather than distance based taxes and would fuel taxes be removed..?
Fuel taxes have two major limitations. Firstly they are invisible, people care about the cost of fuel, but treat road use as “free”. Secondly, you can’t differentiate by time and location. That’s where you seriously unlock the benefits. It will take time, but here’s one approach.
NZ already has RUC, it wouldn’t be hard to move this to a more sophisticated platform for existing RUC payers as a first step. Make all new RUCable vehicles have the new electronic system as mandatory so that:
1. There might be a different price by type of road based on maintenance costs (a simple heavy seal, light seal, unsealed differential).
2. You may have a regional variation based on regional maintenance cost differences (e.g. Canterbury is cheap, Northland is not). You could even charge differentially if there was a parallel railway, so that those roads are charged the actual costs of the route, which would have interesting results.
3. Have a night-time discount as a start of encouraging off peak driving.
It might take up to 10 years to get all existing RUC vehicles on a new platform, with charging varying by road type, a little for location, and a blunt peak/off peak differential. By then the road freight issue on pricing will have been addressed. You’ll also have all diesel car owners used to a basic form of road pricing.
After all heavy vehicles are on the electronic RUC, and you get light diesels on it, then it becomes mandatory for electrics and alternative fuel vehicles (get rid of tax on LPG) and have it initially optional for petrol vehicles. Pay fuel tax or pay distance charging. Then you can either make pricing compulsory on new vehicles or increase fuel taxes to incentivise the shift. Again, could be a 10 year transition.
Ultimately fuel taxes would go, meaning you no longer need to have a refund mechanism for off-road users (which does cost money to administer), but once all vehicles are on it you can charge for congestion, and be far more dynamic. The possibilities are enormous, and yes it can be done wrong, but the key thing is that as a tool it would make a huge difference to how road use is perceived and the choices people make. With a 20 year transition, it would also mean new road building might be assessed on the basis of demand AFTER road pricing is introduced fully, which might mean some major capacity increases never happen.
For road pricing to work (as theory intends it) I think that the price needs to be not just by distance travelled and the standard of road used, but it must also include a time/location/conditions responsive congestion charge to account for the externalised disbenefits that driving on congested roads creates for everyone else.
Furthermore, the pricing would need to transparent and immediately apparent to ensure that people’s driving behaviour is responsive to the price. At the least this would mean a charge when refuelling, but ideally people would have some means to know exactly what the charge is before and during the use of any congested route.
I can’t believe I saw the word mandatory in one of your posts Liberty..! My eyes almost popped out of my head, you’re not still President of the Libnz are you..? You might be impeached..!
I think a long, long introduction period is important and may, just may mean it can be introduced in a budget with the agreement of both major parties…
I might write an editorial to the herald, do you mind if I use some of these ideas LS..?
Jeremy: Bear in mind that I’d simply regard road pricing as a property owner charging for services. It isn’t inconsistent with my principles that state owned infrastructure be priced efficiently, as long as the surplus is either reinvested or returned as a dividend.
Feel free to use the ideas, i’m simply disappointed that pricing is off the agenda for local and central government. For example, in Auckland rates funding of local roads could be replaced with pricing.
There seems to be a large tradeoff between pricing road use as close to cost as possible, and making it transparent so people know how much a journey will cost.
The private sector is not very good at transparent pricing, and once again you need govt intervention to ensure pricing is clear for all.
One quite useful outcome of adopting a time-variable pricing scheme is that people (like me) who mainly use their cars on the weekend would probably actually end up paying less for using roads than is the case at the moment – because of the shift away from a blanket petrol tax.
I think road pricing can be sold if it includes the removal of fuel taxes and 40% of rates…
One good thing about fuel tax is it gives extra punishment to those with ineffecient (SUV) vehicles, and pushes people towards more efficient, smaller cars. Any wholesale shift to road pricing would increase costs for small cars, and lower them for SUV’s. However with future large oil price increases this may not be so much of a problem.
SUV are heavier and you could factor that in your per-km charges…
“That is a good point, it would be similar to a phone bill and I don’t have a problem with Vodafone keeping track of who I’ve called as long as that information is between me and them… Sensible regulations would ensure that…”
Folks, I don’t want to seem like a conspiracy theorist: But if the data is there, it WILL eventually be used for further uses, be they beneficial or invasive. As soon as the government has a way of knowing where everyone drives, they WILL want to use it for other things. Law enforcement (“If you have done nothing wrong, you have nothing to fear…”) will be first, but certainly not the only thing.
We need a much stronger committment to pricacy, both institutional AND hardwired into the devices, or we will end up in a pretty depressing/scary future.
Ten years ago National tried selling road pricing by including the removal of fuel taxes and 40% of rates and including a return on capital investment which would have put road and rail on a level economic playing field. The left portrayed it as privatisation by stealth and won the argument. So here we are today stuck with a petrol tax under Joyce’s direct control.