Tuesday, January 19, 2010

The power of congestion pricing and bus rapid transit

I recently received an email from John of NY with a couple of book micro-reviews that make some really good points, so I thought I'd pass it along with his permission. The power of congestion-priced toll lanes to move so many more cars is a particular eye-opener:
I have read the book Mobility First: A New Vision for Transportation in a Globally Competitive Twenty-first Century by Sam Staley and Adrian Moore that I mentioned back in November as well as a book put out a few years ago called 21st Century Highways: Innovative Solutions to America's Transportation Needs by Wendell Cox, Alan Pisarski and Ronald Utt. Staley and Moore's book Mobility First was an excellent read! I am sure you will really like it. I can assure you that these authors are very much on the same page as you and I.

They argue for more limited access highway capacity and the use of variable tolling (where the toll increase as use increases) and electronic toll collection ("open road tolling") both to generate revenue for building and maintaining new capacity and to control excess traffic loads. Indeed, they cite a very interesting statistic involving the use of variable tolling on the SR 91 Freeway in Southern California. That roadway, which I personally traveled on back in 1997, has 4 or 5 free lanes and two variable toll lanes on the far left side in each direction. According to the statistic they cited, the two variable-priced toll lanes (which are rarely congested) move as many cars in a day as the 4 or 5 free and often congested lanes. If these statistics are even remotely accurate, it makes a whole lot of sense to use variable tolling whenever new capacity is added.

They also advocate improvements in boulevards and other arterial roadways. These improvements include synchronizing traffic lights and building underpasses or overpasses to move traffic nonstop through major intersections. There whole point is to increase the traffic speed and, hence, the mobility for those who use arterial roads.

They also argue that rail mass transit (including light rail) makes very little sense outside of New York City's Boro of Manhattan and a few other very densely populated places. They argue for expanded Bus Rapid Transit (BRT) which is much cheaper, and more flexible and which can be easily combined with express toll lanes and other improvements for cars. Indeed, the book suggests that your skepticism about much of Houston's proposed light rail system is very well-founded. It would seem that BRT should be used instead, as you have suggested. The book was particularly tough on light-rail - which is very much favored by anti-car elites. The authors suggest that light rail in many ways combines the worst both both bus and rail transit. It is inflexible and often quite expensive, like a new rail line, and it is often slow like traditional bus service, particularly where the light rail line must use public streets (as most light rail systems do at least part of the time).

The book 21st Century Highways, which came out in 2005, reaches many of the same basic conclusions as Mobility First. This book was published under the auspices of the Heritage Foundation think tank based in Washington D.C. To me, this book's most valuable contribution is historical context. It provides an interesting history of highway building in America. In particular, it takes a detailed look at the Federal government's role in highway building and transit funding from before the interstate system right up until about 2004. It makes the important observation that mass transit is not underfunded. Indeed, it is overfunded. Over 20 percent of federal "highway" spending now goes to mass transit projects, even though only about 2 percent of all transportation trips involve the use of mass transit, while over 90 percent involve the automobile.
Which creates its own warped incentives (solution here).

Labels: , , , ,


At 9:21 PM, January 19, 2010, Blogger Alon Levy said...

Over 20 percent of federal "highway" spending now goes to mass transit projects...

...and even after adjusting for this effect, the national highway system only made 67 cents in user fees per dollar that went into it. That's according to Pew's Subsidy Scope, which isn't exactly Light Rail Now.

And, of course, on the state level, it's even more egregious - in Texas, a 51% recovery ratio is high (if you need to adjust for gas taxes that don't go into highways, which I'm not sure you do, then make it 68%). Some roads are at 16% (or 24% if you need to adjust).

At least it's nice that they're not calling for abolishing the NYC subway as O'Toole does, or saying that it emits more CO2 than a Prius as Cox does (which is a bald-faced lie - the NYC subway averages 115 passenger-mpg; if you also put the buses in, then it goes down to about 90).

At 7:49 AM, January 20, 2010, Blogger Tory Gattis said...

If you're arguing both highways and transit should survive on user fees alone, I'm in complete agreement. The gas tax does need to be raised, esp. in Texas.

At 1:39 PM, January 20, 2010, Anonymous Anonymous said...

Maybe... raise the gas tax. I don't think it should be the only funding mechanism to make up the current shortfall.

I'd be in favor of funding that also takes into account the damage done to a road. If your car is 2,000, it does a lot less damage than a 20,000 vehicle. Yet, the gas tax does not even come close to approximating this disparity of use.

I'd like to see the combination of:
-gas tax
-direct user fees for newly constructed roads (HOT lanes)
-yearly vehicle miles tax based on the vehicle weight.

If you drive, you can minimize your exposure to any or all of these three taxes.
-Want to minimize your gas tax? Buy a Prius.
-Want to avoid tolls but not in a hurry to get anywhere? Don't drive in HOT lanes.
-Want to avoid a yearly bill for your vehicle miles? Don't commute in an F350 duelie to work everyday from Conroe to the Medical Center.

At 1:57 PM, January 20, 2010, Blogger Tory Gattis said...

I think all of those suggestions make a lot of sense.

At 1:59 PM, January 20, 2010, Anonymous kjb434 said...

Or, the common sense approach is put all the funding that is being taken by the transit and give it back to highways.

No new funding is needed for roads if they received their money as originally intended. Texas can do the same with our state gas tax. There is way too much of it being taken away to be spent on education (an over funded sector).

Texas could do what Louisiana did in the 90s. The state gas tax in Louisiana used to go in the general fund with the intent on it being appropriated to LaDOTD. In the 90s, governor Foster pushed a bill with big public support that all the gas tax be separated from the general fund and that only the LaDOTD has access. Since then, the states roadways have taken a dramatic turn around.

The money is already there. No new funding is needed. Just some politicians with the testicular fortitude to use common sense.

At 2:02 PM, January 20, 2010, Anonymous kjb434 said...

One more thing. All new lane miles should be told. This will allow a new road to have to be self sufficient.

HOT lanes for freeway expansion should be implemented. It's working extremely well for the Katy and will likely work well on US 290.

At 2:03 PM, January 20, 2010, Anonymous kjb434 said...


It should be:

All new lane miles should be tolled.

At 12:10 AM, January 21, 2010, Blogger Alon Levy said...

No new funding is needed for roads if they received their money as originally intended.

The 67 cents recovery ratio measures total gas tax receipts against highway spending. If you measure just the portion of the gas tax that is locked for highway spending, then the ratio drops to 51 cents. If the general fund couldn't be used for highways and the highway trust fund couldn't be used for non-highway spending, the national highway system would go bankrupt.

But at any rate, why should we send every gas tax cent back to roads? Driving has negative externalities, such as pollution. A pollution tax shouldn't go to funding roads any more than the cigarette tax goes to funding tobacco or the alcohol tax goes to funding beer.

What I'm arguing is that, after adjusting for externalities, all forms of local transportation should be subsidized by the same amount, based on what the local government chooses. It's just that if the gas tax weren't so artificially low, transit ridership would be higher, making large subsidies unnecessary. (Beyond that, it depends on transit agency competence, something American governments are in short supply of.)

At 8:05 AM, January 21, 2010, Blogger Tory Gattis said...

The hitch, as I've pointed out before, is that the basic road network connecting all properties is not optional. It must be constructed no matter what. Even if everybody rode transit, you still need access for police, ambulance, fire trucks, deliveries, construction equipment, etc. So there's a good argument for funding that with property taxes. The question is when does the 'basic' road network end and the higher-capacity highway/freeway network begin, which should be paid for with user fees. I think almost no matter where you put that line, at least 80-90+% of all road lane-miles are that basic network connecting all properties.

At 8:52 AM, January 21, 2010, Anonymous kjb434 said...

"Driving has negative externalities, such as pollution."

If that's what your hanging your hat on, then you'll readily lose the argument.

Reducing road congestion through proper traffic mitigation measures will benefit the environment more than just denying funding for roads and pushing it to transit. The Katy Freeway project alone has helped to reduce pollution. The less time cars are stuck at a standstill the better air pollution is reduced (even if more cars use the freeway. This is the primary goal of any freeway expansion project is to reduce the length of the rush hour. In the cost benefit analysis for the various alternative, emissions reduction plays a role. Emissions reductions are looked at projected traffic increase in the future. This also applies to non-freeway projects.

Improving traffic conditions will benefit air pollution more than throwing a train option in the mix will.

At 9:25 AM, January 21, 2010, Anonymous Appetitus Rationi Pareat said...

Pollution is not the only externality associated with driving. One big reason is because what you are advocating invariably increases sprawl and this contributes to many other problems. It eats up green space and reduces habitat for other animals (in some case bringing possibly dangerous animals in closer contact with humans as happens in California).

It also has a HUGE effect on our foreign policy. The greater the demand for oil in this country, the greater the need for this country to continue with a foreign policy that has given us (and many other places in the world), lots of misery and death. And don't give me this canard about "drill baby drill!" If we put oil wells in the Grand Canyon and Yosemite we still wouldn't have enough oil to meet current demands, let alone future increased demands that would result from your "keep building highways/more sprawl" mindset.

So, I think it is you who lost the argument.

At 9:33 AM, January 21, 2010, Blogger Tory Gattis said...

So plug-in electric hybrids like the ones coming out this year would solve most of the externality problem? And the green space issue of sprawl is insignificant when compared to the amount of greenspace re-created over the last 100 years as 90+% of the population shifted from agriculture to cities.

At 9:47 AM, January 21, 2010, Anonymous Appetitus Rationi Pareat said...


These magical, plug in cars have been talked about for decades. Maybe they will come. Maybe in the future we will run our cars on garbage like Doc Brown's Delorean. I don't know.

I am talking about right now. Right now real men and women are living in this country missing limbs, with horrible burn scars and worse. Sadly those are the lucky ones who actually survived the attacks. And when you throw away all the rhetoric, chest beating and political bs, it comes down to the fact that the wars these people were fighting in were about oil and our access to it.

If I was king, I would put at least a $2 per gallon tax on gasoline and just put that money into a fund to support veterans coming home and the families of the people who were killed over there. If you're going to live 30 miles (or more) from work and drive a giant truck just because you want a five bedroom, four bath McMansion in Katy, then deal with the ultimate consequences of that choice.

At 2:45 PM, January 21, 2010, Blogger Alon Levy said...

Reducing road congestion through proper traffic mitigation measures will benefit the environment more than just denying funding for roads and pushing it to transit.

Except that the Texas Transportation Institute believes that public transit is much more effective in reducing congestion - see table here.

So plug-in electric hybrids like the ones coming out this year would solve most of the externality problem?

Not really. Congestion is still there. And the price issue is absolutely brutal. Even regular hybrids, which cost $22,000 apiece as opposed to $40,000 for PHEVs, aren't selling very well. So far the most consistent adopters of hybrid technology have been bus services - hybrid sales are about 3% of new private car sales in the US, versus 22% for buses. Unlike semiconductor technology, battery technology has not progressed very fast, so it's unlikely the cost of a PHEV will go down significantly in the future.

But you have to admit, the whole issue of electric cars is funny. Supposedly, electrified mass transit is not a good solution to cars' pollution, because it's a 120-year-old technology that lost out to the gas-powered car. Instead, we should be adopting electric cars, a 110-year-old technology that lost out to the gas-powered car.

At 4:09 PM, January 21, 2010, Anonymous kjb434 said...


The oil solution is actually an easy one. And yes, we do have more oil in the US than we know what to do with. If we drill the oil we have we could easily remove ourselves from foreign oil.

The worst projections have oil supplies last well over 100-years at our (worldwide) increase usage rate during the late 90s and early 2000s. That's the worst. Many projections that have been in line with reality would have it last much longer. Why stop using what have given quality of life to more people than any other substance in the world.

At 5:26 PM, January 21, 2010, Anonymous Appetitus Rationi Pareat said...


Where did I say we should quit using it? That is completely and utterly unrealistic given our history and today's technology.

My post was for two reasons: (1) to show that driving and the use of oil has more externalities than simply pollution; and (2) to point out one of the biggest ones in particular that often gets lost in these discussions, the effect our addiction to oil has on our foreign policy and how it is literally measured in blood.

Now, to your points, I haven't seen anyone claim that we have the oil reserves domestically to meet our current oil demand, let alone the future demand if we keep living the way we have been the past 40-50 years. Please provide a link for your claim.

I suspect, even if there is some way to meet the demand, it involves placing oil wells in places that many people in this country (including myself) would never agree to. I'm sorry, I am not going to allow you to drill in the Grand Canyon simply because some people aren't willing to conserve a little. I mean, god-forbid that you can't move out to BFE and commute 40 miles to work everyday.

Again, this is not about ending the demand for oil. That is not going to happen in my lifetime. This is about starting to re-think the policies that have gotten us to where we are and trying to slow the growth of sprawl, freeways and all the negative aspects that go with them.

I realize that may be shocking in Houston, the city of big oil and big sprawl but it's about time we started to turn a corner here before change is literally rammed down our throats by the rest of the country.

At 6:08 PM, January 21, 2010, Blogger Michael said...

>>Pollution is not the only externality associated with driving.

I think the main externality associated with driving is death and serious injury. People don't get injured or killed from "drunk subway riding", or at the very least, not too often. And I think we have something like 50k deaths on US highways alone each year.

At 6:22 PM, January 21, 2010, Anonymous Neal Meyer said...

Appetitus Rationi Pareat:

Do you really want to encourage policies that stop sprawl? Is that your goal?

Answer: The fastest and most surefire way to accomplish that goal is to pursue policies that make people poor. Then people will not be able to afford to rent an apartment or buy a home, nor will they be they afford to buy a car. That will solve your problem.


At 6:31 PM, January 21, 2010, Blogger Tory Gattis said...

Another option is the Pickens plan to shift our vehicles from oil to natural gas, which we do have plenty of domestically. Also much cleaner burning.

Michael, your argument could just as well apply to unhealthy foods. Or boats. Or hiking or rock climbing. Should we ban all those too? Where does personal freedom fit in the equation? And in any case, if people feel cars are too risky, they are more than welcome to move to an apartment near work and walk or ride transit everywhere.

At 9:04 PM, January 21, 2010, Blogger Michael said...

>>Michael, your argument could just as well apply to unhealthy foods. Or boats. Or hiking or rock climbing. Should we ban all those too? Where does personal freedom fit in the equation? And in any case, if people feel cars are too risky, they are more than welcome to move to an apartment near work and walk or ride transit everywhere.

I'm not arguing that we should ban driving. But I would argue that when considering the cost / benefit of various modes of transportation in order to decide which mode to build, we should capture as much data on so-called externalities as possible, and fatality and injury rates should be included as part of the cost. I'm sure economists have a way of putting a price tag on a human life that would be preferable to not including this as a factor at all.

Like it or not we all lose out economically speaking when people like John O'Quinn, or even less famous / economically-productive denizens of our fair city, state, and country, die in car crashes each year - let's face it - probably 1-2 people die on Houston's roadways every day and hundreds are injured every week. How many people have died on the Main Street line or on the 82 bus route? Are you saying fatalities and injury rates should not be included in government cost-benefit analysis?

This is interesting (if a bit dated):
"Dallas was the most dangerous central city, mainly because it averaged 169 traffic fatalities annually, a rate of 1.4 fatalities per 10,000 residents. Houston had the next highest traffic fatality rate among cities with an average of 1.2 per 10,000 residents."

>>Where does personal freedom fit in the equation?

Personal freedom comes in after more of the externalities have been accounted for in the standard cost-benefit comparisons. Otherwise, you are just passing the costs off to someone else or to the taxpayer - which has nothing to do with your "freedom".

At 9:23 PM, January 21, 2010, Blogger Tory Gattis said...

I do agree it should be part of the overall cost-benefit equation, as should the value of peoples' time lost in longer transit trips vs. car trips (depending on congestion).

At 2:02 AM, January 22, 2010, Blogger Alon Levy said...

And yes, we do have more oil in the US than we know what to do with.

That's shale oil, which is more expensive to extract than conventional oil. US oil companies tried to drill for it in the 1970s during the oil crisis, but then oil prices came down and they lost a lot of money on it. Even tar sands, which are much better developed, are still at a stage when only about 15% of the total oil there can be extracted. And both tar sands and oil shale are inherently more difficult to extract in terms of energy return on energy investment.

One of the predictions made by peak oil theory is that as oil becomes scarcer, people will start drilling in more difficult places, where the cost of production is higher. Because all the easy oil in the world has already been tapped. Athabasca isn't Ghawar.

Do you really want to encourage policies that stop sprawl? Is that your goal?

Answer: The fastest and most surefire way to accomplish that goal is to pursue policies that make people poor.

Funny, Hong Kong has limited development of parkland and open space without making people poor - all it's done is externality-tax everything. The other big cities of Europe and East Asia are less successful in limiting sprawl, but they still have much more compact development than the US does - even the ones that are as rich as the richest US cities, such as Paris and London.

Michael, your argument could just as well apply to unhealthy foods. Or boats. Or hiking or rock climbing. Should we ban all those too?

Hiking and rock climbing aren't particularly dangerous, despite the stereotype. Unhealthy food is, and you do see a lot of public health experts propose to either regulate what goes into food (e.g. a trans fat ban) or levy externality taxes (e.g. a soda tax).

Your freedom ends where my nose begins. With unhealthy food the main ill effect is on public health, which is a social good in the abstract. With driving, it's more straightforward: all those cars and heavy factories in LA and Houston are poisoning other people, not just their own users. If New York had had the car usage rate of Singapore or Paris or Tokyo, my lungs would be a lot cleaner than they are right now.

(Yes, CNG is a partial solution, and should be pollution-taxed at a lower rate than diesel and gasoline, which would encourage more transit agencies to switch. But it's not a solution to everything - e.g. it offers no improvement in carbon content. Under a Mankiw-style regime, it'd probably be taxed at a little over $1/gallon, versus more than $3/gallon for gas and $3.50/gallon for diesel.)

At 9:44 PM, January 24, 2010, Anonymous Anonymous said...

"According to the statistic they cited, the two variable-priced toll lanes (which are rarely congested) move as many cars in a day as the 4 or 5 free and often congested lanes."

Are those statistics available somewhere? As stated, I think there's no way that's true. Sure, the HOT lanes may move more cars during the 2-4 hours of congestion. But the rest of the day the free lanes are moving at posted speeds and the HOT lanes are minimally used.

At 10:22 PM, January 24, 2010, Blogger Tory Gattis said...

To clarify, these stats are for the 91 in OC, CA - not the Katy HOT lanes. The 91 has many congested hours every day, including weekends, and I can tell you from personal experience that the free lanes can move agonizingly slow (<5mph).

To answer your question, I'm sure the stats are probably footnoted in the book, but I don't think I have a copy (although I might - I'll have to check).

At 10:33 AM, January 25, 2010, Anonymous Appetitus Rationi Pareat said...

So your ultimate transportation solution is to allow sprawl until we have the same mess they have out in the Los Angeles metro area and then the HOT lanes will be the savior?

I'm sorry. I think we need to rethink what we are doing before we get into that situation, not after.

At 10:08 PM, January 25, 2010, Blogger Tory Gattis said...

Sprawl is not really optional - it's going to happen no matter what. Even NYC, Boston, Chicago, and DC have tremendous sprawl. The question is whether we are going to channel all of the sprawl commuters onto heavy rail lines into one or two major core job centers, or use HOT lanes to distribute them to a dozen+ ones (as Houston is today).

At 4:13 AM, January 26, 2010, Blogger Alon Levy said...

Sprawl is optional - just look at what happens in Hong Kong and Singapore, where it's not subsidized.

Even in the West, where sprawl is subsidized, it's not subsidized to the same extent everywhere, and it doesn't exist to the same extent everywhere. New York sprawls more than Paris, and Houston does more than New York.

Rail is capable of serving suburban job centers - see e.g. La Defense. In Greater Tokyo, the suburbs have a 30-40% transit share overall. The trick is not to build it in the style of American steam trains (or Metro-North, which apart from the upgraded technology hasn't really left the steam era in terms of service).

At 7:48 AM, January 26, 2010, Anonymous Anonymous said...

But Hong Kong and Singapore are essentially island nations, unable to expand outward. If they wanted ANY green space in their countries, they had to preserve it somewhere near/in town. Plus, Hong Kong is pretty mountainous. They just couldn't realistically build (or at least cheaply build-more cheaply than another apartment tower) suburbs in the mountains. The same thing holds in California. Whatever policies existed, Mount Tamalpais or the mountains to the south of San Francisco were much less likely to have neighborhoods than the flat areas of Oakland and Berkeley. In LA, people tend to live in the valleys.

And Paris? It's been growing without automobiles for over a thousand years. It grew up that way. That's history that shaped Paris, not anti-sprawl policies that made it that way.

At 8:58 AM, January 26, 2010, Anonymous Appetitus Rationi Pareat said...


Sprawl may not be "optional" but it can be better controlled with the right policies (as Alon illustrated with his examples). The problem is that, far too often, Houston has failed to control sprawl or mitigate the negative aspects of sprawl. The LA region made the same mistake.

This is not because of history or geography but because of policy failures, failures you seem to want to perpetuate.


La Defense was developed starting in the late 1950's. Much of the building boom there occurred in the 1970's and 1980's (about the same time Uptown Houston boomed). The difference is that, from the start, Paris planned better for the transportation needs going in and out of La Defense. Houston did not.

At 11:04 PM, January 26, 2010, Blogger Alon Levy said...

Anon: outside the Ile de la Cite and the Latin Quarter, Paris only urbanized in the industrial era. The urban area only reached the limits of the present-day city in the late 19th century. The high density and lower-than-US car use came from the M├ętro and urban planning encouraging the rich to live in the city and the poor in the suburbs.

And Hong Kong is mountainous, but other space-constrained cities build in the mountains all the time. The favelas in Rio are all on the hillsides, as are the luxury communities in LA (the people live in the valleys in LA because the valleys don't have multi-acre mansions). Hong Kong made a deliberate decision to avoid that, and instead preserved open space both in the mountains and in flat territory. (And Singapore did not - it just taxed car externalities.)

I'm not denying that geographical constraints can force a city to build upward, as they did in 19th century New York. But they're not the only determinant of density. For example, Toulouse has rates of growth comparable to Houston's and no geographical constraints; its density of 3,700/km^2 is low by French standards but not by American ones, much less Sunbelt ones.


Post a Comment

<< Home