This weekend, Los Angeles will close a 10-mile stretch of the 405 freeway for 53 hours so work crews can conduct demolition that will enable widening of the freeway. Locals are referring to the planned closure as “Carmageddon,” anticipating gridlock on nearby roadways that remain open. The hope is that the short-term pain will bring relief of chronic traffic congestion once the widening project’s finished.
Unfortunately for hopeful Angelenos, adding more road space probably won’t relieve congestion. NPR’s Guy Raz spoke to University of Toronto economist Matthew Turner, co-author of a study that analyzed traffic in 228 US cities, who reported that 1% more roads means 1% more driving. Turner also had disappointing news for hopeful public transportation advocates like me: More public transportation doesn’t decrease traffic congestion, either. Here’s a bit of Turner and Raz’s conversation (transcription errors are mine):
Turner: What we find is that as you increase a city’s stock of light rail or bus cars, that there’s no impact on the amount of driving. That sounds sort of surprising, but it’s logically consistent with the other result that we have about the effect of roads on driving. What we find is that as you add roads to a city, those roads get filled up, so there are people waiting to use the capacity …
Raz: It’s almost like there’s a long line of people waiting to use that space on the freeway. The minute you put someone in a bus or a metro train, someone else is just going to take their space on the roads?
Turner: I think that’s right. There’s a lot of people out there trying to decide whether to make one more trip to the store or drive a little further to work. As you add a little bit more capacity, it gets filled up.
Tanya Snyder of Streetsblog Capitol Hill interviewed Turner several weeks ago and read the new paper, soon to be published in the American Economic Review, he co-authored with Gilles Duranton. Snyder writes:
They didn’t find that transit reduces congestion. But that doesn’t mean that metro areas shouldn’t build transit as a way to maximize the efficiency of their transportation networks, they say. Turner said transit is a good way to get more “person-miles” out of roads.
I’d also add that whether or not public transportation reduces congestion, it’s important for people who don’t drive cars. There are many reasons people don’t get behind the wheel – age, physical limitations, dislike of driving stress, the high costs of car ownership, or a desire to consume alcohol, send text messages, or do other things that impair driving ability. Plus, using public transportation instead of driving a car can allow people to get more physical activity as they walk to and from stops.
(Since there’s a common misconception that transit advocates want to take away everyone’s cars, I’ll also add that there are plenty of times when driving a car might be a better option than taking the bus – for instance, when a destination is poorly served by public transportation, when the trip occurs during hours of little or no transit service, when hauling bulky items or squirmy children, or when a person has physical limitations that would make it hard to walk to a stop, stand at a stop, or remain standing in the aisle or a crowded bus or rail car.)
As worthwhile as public transportation is, though, Turner and Duranton’s research found that it doesn’t reduce traffic congestion. Since building more roads doesn’t solve this problem, either, that leaves us with congestion pricing as the solution. Tolls on cars entering central areas during peak times have successfully reduced traffic congestion in London, Singapore, and Stockholm – and Turner told Raz that Stockholm saw a dramatic reduction in peak travel time in response to a relatively low fee. You can read more about these cities’ experiences in this report prepared by KT Analytics for the US Department of Transportation’s Federal Highway Administration.
New York Mayor Michael Bloomberg tried to move toward congestion pricing for Manhattan, but didn’t succeed. As Turner pointed out to Raz, “people don’t want to pay for something that they’re used to getting for free.”
Raz brought up the concern that lower-income households will find it harder to afford tolls; Turner agreed, and added that poor workers tend to have less discretion about when they get to work. (One option for avoiding a toll is to travel before or after peak times, but it’s often higher-paid employees who have the flexibility to work from, say, 7am – 3pm or 10am – 6pm.) The NPR segment doesn’t delve much farther into this issue, but the KT Analytics report notes that London and Stockholm both use toll revenue to fund improvements to public transportation. Better public transportation, perhaps along with subsidized fares, could make it easier for those who can’t afford the tolls to take a train or bus instead – although, again, public transportation still won’t work for every person in every situation.
Given that traffic congestion has large environmental, time, and economic costs, and that other attempted solutions haven’t succeeded, it seems like it would be worthwhile to figure out how to make congestion pricing work.
Always fun seeing further confirmation of Braess’s paradox.
this is not a case of Braess’ paradox
What is the goal of the congestion pricing? To reduce the normal of commuters or to push people to public transportation? I don’t see how either of those is necessarily a net positive.
“Plus, using pubic transportation instead”
You dropped an “L” — thought you might want to know.
*fades back into the ether*
Thanks, WMDKitty!
Superdave, congestion pricing is essentially a market approach to allocating a scarce resource (road space). Because drivers aren’t currently charged to use road space, demand for it is high, and congestion is the result. The idea is that if you start charging for the use of road space, the users who value it most highly will pay for it, and those who value it less will use it less (assuming they have other options, which isn’t always the case).
Problem with congestion charging is that your boss doesn’t let you come in at any other time than for a 9am start and doesn’t let you go home when you want either (they’ll be OK with you staying longer).
In London the congestion charge was brought in to reduce car use and get people using trains.
So people used the trains and the trains couldn’t cope.
So they put congestion charges at commuter hours to cut down on people using the trains.
So they went back to using their cars.
Actually improving the capacity at peak times didn’t occur to them as necessary before congestion charging and still hasn’t popped up as an idea.
Similarly when people were exhorted to cycle to the train station to cut down on car journeys: they changed rolling stock to ones without a guards van or place to put bikes.
Wow: Here in DC, we’re lucky that the federal government allows a lot of employees to work flexible schedules, so a not-insignificant portion of the workforce is getting to their desks at 7am. But as long as “normal business hours” continue to be assumed to be 9am – 5pm, a lot of people will still have to show up at 9. (Telework is another arrangement that can help, and is used by some government agencies, but there are lots of jobs that can’t be done remotely.)
As you point out, transit capacity can’t always increase rapidly if a bunch of people are suddenly shifting from driving to bus or rail (whether due to a new congestion charge, spike in gas prices, or some other factor). Especially when it comes to rail, the costs are high and the lead times long. For istance, this Washington Post article by Lena Sun reports that buying new rail cars for the region’s Metro system in expectation of a planned Metrorail extension will cost $2 billion and require years to complete – “about 10 months to solicit bids, an additional three years for manufacturing and more time for testing.”
Ideally, cities would be able to anticipate increased demand for public transportation several years in advance, and get sufficient funds from their jurisdiction to invest in necessary service expansions. But I’m guessing that doesn’t always happen.
Wow @7:
My understanding is that the central part of the London Underground has been running at virtually full capacity for several years. The length of the stations limits the number of cars per train and the train frequency is limited by the duration of the station stops, the distance between signals and the train stopping distance. Adding more capacity is no trivial matter and basically comes down to building more lines – although I agree that construction of the Crossrail Link is coming about 10 years late.
Recall that the subways in central London are 100 to 160 years old. The circle line was built in the 1860s and originally used steam locomotives. The problem in the UK is that there is no money for infrastructure, just like in the US. (As an example about 15 years ago I took the Channel tunnel from Brussels to London, and once you hit the UK shore you slowed right down from the 300 km/hr speed in France.
On the larger subject I guess the way then is to let the congestion increase till the pain point is reached and people start moving away. (After all Las Vegas has a lot of empty houses)
“My understanding is that the central part of the London Underground has been running at virtually full capacity for several years”
The reason for that was that after being sold off to private interests (well, to a lesser extent before then), the rail network had a large amount of land that was used to hold rolling stock. That land also contained rolling stock.
So when the morning rush hour into London happened, the rolling stock waiting outside London in the yards were used to make trains going in longer than normal. Running back, they’d drop off the extra carriages in the london yards to reduce their weight and wasted space.
As the day wore on, the london yards filled up. During the day, the rolling stock that wasn’t needed was left there.
At the end of the day, the train would pick up extra carriages allowing more people to leave London than enter it and the yards outside London filled with carriages, ready for the next day.
But they were sold off.
Now they have 8 carriages and that’s the case whether it’s full with enough people to fill 12 carriages or whether it would barely fill 2.
I’m a little older and I remember the idle rolling stock.
Lyle, the speed change on the train is due to the cuttings and so forth requiring a smaller train and curtailing reactive suspension. It also pretty much cost the UK the war:
1) You have to transport tanks via road or rail. Rail is far more effective, and you still have to drive your car through cuttings anyway.
2) your train has to be wide enough to carry the tank, therefore it limits the size of your wheelbase
3) Your turret can’t overhang the tank body, it still has to pass through the same tunnel, this limits the size of your turret
4) The size of the turret has to hold the recoil of the main tank gun. this means a smaller turret means a smaller gun
5) German tanks were able to pop one at british tanks at a distance far beyond the range at which they could return the favour
And why the width? Romans. The bloody Romans.
I bet they did it deliberately.
đŸ™‚
Wow – you are probably correct about the ‘spare’ rolling stock being sold off, and certainly selling off the railways did a lot of damage.
At least part of the reason for the small gauge of British railways was that it was virtually the only system that was not state sponsored. Railways were required to fence all lines, build and maintain road overpasses and so on. When weight limits are raised for trucks, the railways are required to carry all the expense of strengthening the bridges. So naturally most railways were built as small as possible.
Perugia isn’t L.A. or New York, but it is an interesting case study.
http://news.nationalgeographic.com/news/energy/2011/01/110126-perugia-italy-energy-minimetro/