I’ve added a new “loyal opponent” to my list (right), the Public Transit blog, which is run by Michael Setty. In truth, Michael’s loyalty as an opponent is somewhat questionable as he is more willing to listen to alternative views than some of his more radical smart-growth allies. (I hope I don’t reduce his credibility among his peers by saying so.)
In any case, he has a recent post comparing Houston and Portland traffic in 1993 and 2003. In a nutshell, data presented in the post show that Houston built more freeways, while Portland built light rail. Yet traffic congestion grew faster in Houston than in Portland. “These data suggest,” Michael mildly observes, “that some of the main beneficiaries of Portland’s transit investments may be the drivers who remain on the road.”
Did Portland’s light-rail lines significantly reduce congestion? At first glance, the data seem to say so. While this may be debated at length for years to come, I think there are some alternative explanations.
I first wondered why the data (which, Michael says, were compiled by someone else) only go back to 1993 when the original source of the data (the Texas Transportation Institute) has numbers going back another decade? I found out when I looked at the earlier numbers. As the table below shows, Houston and Portland present the opposite picture between 1983 and 1993 vs. 1993 and 2003.
Growth in Various Indicators from 1983 to 1993 and 1993 to 2003
| Houston | Portland | Houston | Portland | |
| 83-93 | 83-93 | 93-03 | 93-03 | |
| Lane miles | 50% | 20% | 13% | 4% |
| Transit PM | 91% | 19% | 13% | 83% |
| Delay/traveler | -28% | 560% | 66% | 18% |
| Transit’s share | 42% | -25% | -15% | 26% |
| Per-capita driving | 8% | 41% | 5% | 10% |
The satisfaction generated buy generic levitra due to the enjoyment during copulation is amazing. This legislation means that in order viagra levitra the future, it will continue to be in the future, it will continue to be in the future, and you will not have to face any embarrassing situation. Some canadian viagra pills of the factors include alcohol abuse, emotional, and mental stress, and depression which is all that have negative effects on testosterone levels. Buying medications amerikabulteni.com online doctor viagra and dietary products in this modern time has never been so easy.
Those who want actual numbers can download a detailed spreadsheet showing every year from 1982 through 2003.
In 1983, Houston had the worst congestion in the nation. So it started a major highway expansion program, paid for mainly with tolls. Over the next decade, it built 50 percent more lane miles of freeway. Many of those lane miles were bus or HOV lanes, and Houston greatly expanded its bus service to take advantage of those lanes. As a result, bus ridership grew by 91 percent.
All of this reduced congestion (as measured by hours of delay per rush-hour traveler) by 28 percent. From Houston’s most-congested year of 1985 to its least-congested year of 1992, congestion declined by 52 percent.
Meanwhile, Portland built light rail, opening its first line in 1986. It also added about 20 percent more lane miles of freeway. Between 1983 and 1993, transit use grew by a meager 19 percent while congestion grew by a whopping 560 percent. Portland’s population actually grew slower than Houston’s during this time period; but Portland’s per-capita driving grew much faster than Houston’s: 41 percent vs. 8 percent.
Houston road construction slowed after 1993, so that it built only 13 percent more lane miles in the following decade (compared with 50 percent the decade before). Houston’s transit agency also became enamored with light rail, so transit ridership growth slowed to 14 percent over the next decade (compared with 91 percent the previous decade). Ridership has actually fallen since 2001, when they began building the light rail. Growth in driving has continued unabated, so congestion increased by 63 percent.
Between 93 and 03, Portland built only 4 percent more freeway lanes, but it did manage to rack up an 83-percent increase in transit ridership (still not quite as good as Houston’s 91-percent increase the decade before). Congestion increased by only 18 percent.
Transit clearly did not play a major role in reducing Portland’s congestion growth. Transit only carried 2.3 percent of Portland’s motorized passenger travel in 2003 (which is actually less than the 2.5 percent carried in 1983, but more than 1993’s 1.8 percent). Getting a half-percent of people out of their cars did not make a big difference.
To put it another way: between 1998 and 2003, Portland’s per-capita auto passenger miles declined by 2.7 miles per day. Per-capita transit passenger miles grew by 0.1 miles per day. So something other than transit is responsible for 96 percent of the decline in driving.
The big difference is in the change in driving and per-capita driving. Houston’s per-capita driving grew at a relatively constant rate in both decades. But Portland’s grew by a whopping 41 percent in the first decade and only 10 percent the second decade. Why would that be?
I am sure some smart-growth advocates are quick to credit the slowing growth in per-capita driving to Portland’s pedestrian-friendly design and other land-use policies. But I think a more fundamental explanation is likely. In 1983, Portland was suffering its worst recession since the Great Depression. By 1993, it was booming, with new high-tech factories opening every year. Per-capita driving continued to grow through 1998, but in 2000 another recession began, inspired in part by high housing prices that caused employers to look for more affordable cities to locate their factories and offices.
While some questions are still open to debate, the data clearly show:
- You can build your way out of congestion. Houston reduced congestion by 50 percent with new roads paid for almost entirely out of user fees. Unfortunately, local opposition stopped this highway growth.
- New highways do not induce new traffic. Houston’s traffic growth after building the new roads was about the same as before.
- Transit has an insignificant effect on congestion. Houston grew transit’s market share from a low of 0.6 percent to a recent high of 1.1 percent. Portland grew it from a low of 1.6 percent to a recent high of 2.3 percent. Again, getting 0.5 to 0.7 percent of cars off the road is not going to make much of a difference.
What remains open to question is exactly why Portland’s per-capita driving growth slowed after 1998. I think the economy had something to do with it, but to be honest Portland’s economy did not begin to decline until a couple of years after per-capita driving started dropping. There may be some measurement problems: all of these numbers are the best estimates of state departments of transportation or local transit agencies, but the best may be none too good.
If Portland planners did play a role in driving down per-capita driving from 21.1 miles per day in 1998 to 19.4 miles in 2003, I have to wonder whether it is really a gain. There are many reasons why we value mobility and resent congestion. Instead of giving up on trying to relieve congestion, Portland would have been better off doing what Houston did in the 1980s: cost-effectively meet demand with new systems paid for out of user fees.
What remains open to question is exactly why Portland’s per-capita driving growth slowed after 1998.
jk: Perhaps an influx of new immigrenats who tend to use mass transit for their first years here until they can afford a car?
Didn’t I read that Trimet has been losing market share lately?
thanks
JK
1.You can build your way out of congestion. Houston reduced congestion by 50 percent with new roads ….
Yeeees. All those new anti-Kelo initiatives ought to help this idea along quite well! Condemn away for road ROWs, I say.
2. New highways do not induce new traffic. Houston’s traffic growth after building the new roads was about the same as before. [emphasis added]
Hahahahahaha! Aaahhhh…hahahahahaha!
Randal made a funny:
1. Disparate evidence indicates that the provision of extra road capacity results in a greater volume of traffic. The amount of extra traffic must be heavily dependent on the context…of road schemes, but an appropriate average value is given by an elasticity of traffic volume with respect to travel time of about –0.5 in the short term, and up to –1.0 in the long term. As a result, an average road improvement has induced an additional 10% of base traffic in the short term and 20% in the long term… [emphasis added]
2. Improved roads simply spur additional travel or divert trips from parallel routes, quickly returning a facility to its original congested condition. Traffic is thought to behave more like a gas than a liquid–it expands to fill available space… Claims of induced demand have spawned such cliches as “build it and they will come” and “you can’t pave your way out of traffic congestion.”
The preponderance of empirical evidence to date suggests that the effects of induced demand are substantial. A widely cited study by Hansen and Huang (1997). based on 18 years of data from 14 California metropolitan areas, found that every 10% increase in lane miles was associated with a 9% increase in vehicle miles traveled (VMT) 4 years after road expansion, controlling for other factors. Another study of 70 U.S. metropolitan areas over a 15-year time period concluded that areas investing heavily in road capacity fared no better in easing traffic congestion than areas that did not (Surface Transportation Policy Project, 1998). [emphasis added]
3. A simple and fundamental principle of economics is that consumption increases as goods become more attractive to the consumer. If transportation is viewed as consumable goods, then transportation infrastructure will partly determine its attractiveness to the potential user. Improving the overall attractiveness of a transportation system will increase traffic and therefore ultimately lead to more traffic-related pollution.
Such new roads accelerate the traffic and the motorists save time. The question then arises of how the motorists spend the time saved. The answer to this question is surprisingly simple but is a key to understanding the increase in traffic. The time saved is used to join more traffic which results in additional traffic. This traffic is ignored by conservative traffic experts, although – apart from the direct impacts on the landscape – it is the most important impact of a new road on the environment. There is a technical term for this kind of traffic. It is called induced or generated traffic. [emphasis added]
4. An analysis of US data on lane mileage and VMT by state is conducted. The data are disaggregated by road type (interstates, arterials, and collectors) as well as by urban and rural classifications. Various econometric specifications are tested using a fixed e€ffect cross-sectional time series model and a set of equations by road type (using Zellnero’s seemingly unrelated regression). Lane miles are found to generally have a statistically significant relationship with VMT of about 0.3±0.6 in the short run and between 0.7 and 1.0 in the long run. Elasticities are larger for models with more specific road types. A distributed lag model suggests a reasonable long-term lag structure. About 25% of VMT growth is estimated to be due to lane mile additions assuming historical rates of growth in road capacity. The results strongly support the hypothesis that added lane mileage can induce significant additional travel. [emphases added]
5. Induced travel responses to changes in road capacity have been recognized by transportation planners and economists since the first applications of cost-benefit analysis to road projects in the 1930s. However, interest in determining the magnitude of induced travel responses has grown in recent years in response to policy concerns about the links between highway construction, air quality and urban development patterns, and the desire to ensure efficient resource allocation through more detailed evaluation of highway projects. It has also been suggested that induced travel responses have become more substantial as a result of worsening traffic congestion. [References omitted, emphaisi added]
Thanks for the laugh with your hilarious argument this morning, Randal. Precious.
DS
DS,
Thanks for proving once again that too many planners still believe the myths and don’t keep up with reality. As Robert Cervero (a planner himself) wrote a couple of years ago: “Many induced-demand studies have suffered from methodological problems that, I believe, have distorted their findings.”
Based on his own research on induced demand, he concludes that if you build a new road, it does not lead people to drive MORE. It may lead people to locate new facilities on the road, thus people will drive on THAT road more. But it does not lead to more driving overall.
“Fighting highway projects,” he concludes, “is misguided. The problems people associate with roadsâ€â€e.g., congestion and air pollutionâ€â€are not the fault of road investments per se. These problems stem from the use and mispricing of roads, new and old alike.”
Randal, I included Cervero’s paper that he mentions in this link you provided (#2).
Nonetheless, are you claiming that the Cervero paper in JPL says that [n]ew [roads] not induce new traffic?
DS
4. should read:
Nonetheless, are you claiming that the Cervero paper in Access says that [n]ew [roads] not induce new traffic?
Apologies.
DS
DS,
Read what I wrote. New roads do not induce people to drive more. They may induce people to move homes or businesses to areas convenient to the roads so they can drive on an uncongested road instead of on the congested ones they were driving on before. Thus, the new roads relieve the congestion on the formerly congested roads.
The idea of “induced demand” is that, if you build more roads, people will just drive more, so there is no congestion relief. This is simply untrue, which is why Cervero’s paper says this idea has led to poor investments.
The Pfleiderer and Dieterich paper cited suggests that the extra fuel consumed by induced traffic results in increased pollution. As I understand it, decelerated traffic (congestion) is a terrific polluter, and there is nothing in their argument that shows the pollution from the extra fuel consumed by induced traffic is greater than that generated by the decelerated traffic it replaced. Given the improving air quality in the US for the past couple decades, it may not even matter. The paper also fails to speak to increased productivity resulting from induced traffic. Shouldn’t that be factored into any cost-benefit analysis?
Pfleiderer and Dieterich also indicate that improving public transport also generates increased traffic. I hope those pro-planning aggressors who elect to cite this article in their defense will embrace it in its entirety, lest they show themselves to be cherry-pickers.
And of course, anti-planners know that few, if any, anti-Kelo initiatives seek to prevent takings for road ROW’s.
JK If induced demand did exist, what would be wrong with letting people be free to travel? Or is that too much UNLIKE the old Soviet Union and other enlightened societies?
Thanks
JK
Hi folks. I will introduce myself as a Houstonian who has lived here for 35+ years of my life. Randall, I can’t wait for the ADC conference to be held here next year. I’d love to meet everyone.
I would like to say that with $60 oil and $7+ per MCF of natural gas, Houston is currently booming as it was back in the 1970’s. Our census numbers for the City of Houston were at 1,953,000 in 2000 with Harris County at 3.4 million. The latest unofficial figures, which are of controversy since the City charter requires additional city council seats to be added if the city reaches a population of 2.15 million, are somewhere around 2.3 million. No numbers are available yet for the county.
From purely anecdotal evidence, I can see there is a lot going on in Houston right now. Things are just booming. The roads are definitely more congested than they were just 4 or so years ago.
I am in the process of doing a study on my own of transit patronage for Metro (the local transit agency) from 1997-2007, having asked for public information requests from Metro which I will be getting this week. Even with this crush of new comers from Hurricane Katrina and those looking for new opportunities, the comparatively miniscule patronage numbers don’t look any better for transit. The tram and the buses run at an average speed of 13mph which cannot match the speed or versatility of cars.
As for the light rail system, Metro did the usual transit agency tricks. The cancelled a few dozen bus routes to save operational money for the tram and rerouted many bus routes towards the tram to ensure ridership and show how incredible the tram really is. 4 new BRT lines are in the works and it is quite as though we will see a repeat the exchange of bus routes for trams / BRT. We do have a noisy constituency demanding mass transit, but we still don’t have the density here to justify it and it many people would just get turned off with the amount of time it would take to get anywhere on mass transit even if we blew the money on it.
The road building continues. We have a $3 billion expansion of I-10 (with 2 toll lanes), and none too soon. The county is the main toll road builder, as the City of Houston has given up building roads to a large degree. My main worry about toll roads is not the roads per se, but rather that the County will use toll road monies to cover up other – ahem – **** ups, such as bailing out sports stadium bonds and what not.
I do strongly agree with criticisms of so called “induced demand” by road building. People make their decisions on the margin and if the roads are “free” to use, then they will take the road if it gets them to where they want to go faster. Road pricing does matter as tolls make them reconsider this decision.
I’m buring the midnight oil writing this and I have to get up tomorrow morning. Enjoy your day!
Read what I wrote. New roads do not induce people to drive more.
I did read what you wrote, which is why I responded.
You made a claim that you based on a Cervero paper, and I’m asking you:
do you claim that the Cervero paper you linked to asserts that new roads do not induce people to drive more?
If this is your claim, can you explain why Cervero disagrees with how you characterize his paper and spells out why on pp. 25-26?
DS
I hope those pro-planning aggressors who elect to cite this article in their defense will embrace it in its entirety, lest they show themselves to be cherry-pickers.
There’s no ‘lest’, as there’s no cherry-picking:
And as the other papers show, the intensified land use provides amenities which in turn increase traffic. Not much you can do about nonessential trips if people choose to make them. Hence the increasing talk about congestion pricing, eliminating free parking, etc.
DS
Dan writes: “do you claim that the Cervero paper you linked to asserts that new roads do not induce people to drive more? If this is your claim, can you explain why Cervero disagrees with how you characterize his paper and spells out why on pp. 25-26?”
I said that new roads do not lead people to drive more. They may lead people to relocate and drive more on that road.
Cervero says on pp. 25-26: “Developers often know about highway projects well before they are built, and many begin securing building permits and entitlements early on. Thus, they respond directly to the variable ‘lane miles’ in the path analysis. Building activities also respond to performance (i.e., higher travel speeds). All else being equal, suburbanites prefer to live near fast-moving corridors than snarled ones.”
Cervero goes on to say: “for every 100 percent increase in capacity there’d be an eighty percent increase in travel, reflecting increased travel speeds and land use shifts along improved corridors. However, only around half the increases in speed and growth in building permits was due to the added capacity. Factors like employment and income growth accounted for the other half. Accordingly, the traffic gains that one can attribute to the added capacity is actually around half of eighty percent, or forty percent.”
In short, he says for every 100 percent increase in capacity, you get 40 percent more driving — but that 40 percent is due to “increased travel speeds and land use shifts along improved corridors.” It is NOT due to more per-capita driving.
My interpretation of Cervero is correct. New roads do NOT lead to more per-capita driving. They only relocate driving. This is confirmed by the Houston numbers that I cited in my main post. If you believe that new roads lead to more total driving, then you are misreading Cervero and relying on past studies that Cervero says are misleading.
Come now.
Cervero, in this explanatory paper for his paper I cited earlier, merely made the numbers more precise, for example the elasticity numbers from another paper I cited above went from .8 to .4 in his results.
The .4 is the increased VMT, which negates your point 1. in your original, and the negation of your blunt point 2. Cervero explains in the paper you must either not be reading or think no one else will for Cervero explains how “over the longer run, behavioral shifts are only part of the story” (but they are still part of the story):
DS
On March 7th, 2007, Dan said:
“There’s no ‘lest’, as there’s no cherry-picking:”
Anyone who cites this paper as an argument against road building as a means of reducing congestion, and ignores it when promoting light rail, contradicts their own cherry-picking position. The admonition is valid, and failure to recognize that truth doesn’t change anything.
The admonition is valid only if you can show cherry-picking.
But there’s no need to change the subject, as it’s plain for all to see that the empirical evidence is ignored and the paper used to support a position doesn’t say what the asserter claimed it said.
But feel free to make your ‘ignoring’ point using the paper you like when the topic is appropriate. Provided, of course, that you don’t ignore the paper when you promote auto-oriented solutions.
DS
I think what The Antiplanner is attempting to say is that construction of more roadway capacity and roads that can be traveled faster does not lead to more per-capita TIME being spent behind the wheel. What does increase is MILES DRIVEN PER CAPITA, up by 40% or so for a given roadway as pointed out by Cervero.
In fact, this sort of thing happens all the time in the U.S., as land uses continue to drop in densities and sprawl into the suburbs, “real” gasoline prices have dropped since 1980 until very recently, and incomes have increased. People have been running up much more annual miles driven per capita, up something like 60% since 1980 when gasoline prices peaked in “real” terms.
For more insight into this, see http://www.planetizen.com/node/23032.
The post there pretty much summarizes the trends “driving” growth in overall U.S. VMT traveled and VMT per capita. Additional factors–as The Antiplanner knows quite well–include chronic non-pricing of automobile travel in urban areas, the $300 billion+ annual subsidies to motorists for “free” parking NOT covered in the act of driving documented by Dr. Shoup, and the yet-to-be-calculated damage to the planet and its climate contributed by motor vehicles (something I know Randal is skeptical of).
On March 9th, 2007, Dan said:
“The admonition is valid only if you can show cherry-picking.”
No, Dan. An accusation would have to show cherry picking. An admonition makes no charge, and therefore requires no defense.
“But feel free to make your ‘ignoring’ point using the paper you like when the topic is appropriate. Provided, of course, that you don’t ignore the paper when you promote auto-oriented solutions.”
Come now, Dan.
What ever makes you think I would pay attention to any proviso you attach to the circumstances under which I can “feel free” about anything?
Anti-planner,
Until the quality of your opposition improves, I won’t be posting anymore comments. Rest assured I will be visiting your site daily. Keep up the good work, and best wishes.