As if anybody didn’t realize it before, it’s now obvious that the Texas Transportation Institute, despite its prestige, is intellectually bankrupt. David Alpert at Greater Greater Washington says it better than I could:
The Texas Transportation Institute today released the final version of their report on congestion, which ranks the DC area tied for first with Chicago in hours wasted in traffic. Unfortunately, the report’s methodology completely misleads as to the seriousness of traffic, and TTI is pushing the wrong policy solutions.
The TTI report narrowly looks at only one factor: how fast traffic moves. Consider two hypothetical cities. In Denseopolis, people live within 2 miles of work on average, but the roads are fairly clogged and drivers can only go about 20 miles per hour. However, it only takes an average of 6 minutes to get to work, which isn’t bad.
On the other hand, in Sprawlville, people live about 30 miles from work on average, but there are lots and lots of fast-moving freeways, so people can drive 60 mph. That means it takes 30 minutes to get to work.
Which city is more congested? By TTI’s methods, it’s Denseopolis. But it’s the people of Sprawlville who spend more time commuting, and thus have less time to be with their families and for recreation.
Sadly, despite CEOs for Cities pointing out these methodological problems last year, TTI went ahead and finalized its report without fixing them (PDFs). TTI ranks Portland as worse than Nashville, with a Travel Time Index (TTI) of 1.23 for Nashville and worse TTI of 1.15 for Portland. However, because of greater sprawl, Nashville commuters spend an average of 268 hours per year commuting, while the average Portland commuter spends 193 hours per year.
What does this mean for public policy and the Washington region? TTI’s data is often used to justify spending money on new freeway capacity, since congestion sounds bad. TTI even promotes this approach. Tim Lomax, a co-author of the report, told the Post’s Ashley Halsey III, “You can do little things like stagger work hours, fix traffic-light timing and clear wrecks faster, but in the end, there’s a need for more capacity.”
This logical fallacy is very similar to that of per passenger-mile costs. Michael Lewyn has also discussed these issues in the past.
I was a Transportation/Civil Engineer and it will never cease to amaze me at how accepting media typically is of pronouncements from civil engineers. It is just so amazing that ASCE thinks that all of our nation’s infrastructure needs to be replaced or that a group of transportation engineers from Texas A&M thinks we need to build more freeways. SO SHOCKING.
Actually TTI for Portland is 1.23 and 1.15 for Nashville. Confused me when I read the article.
And what that means is that the ratio of peak hour travel time/free flow travel time is higher in Portland. To use the example in the story
Sprawlville TTI =30min/30min=1
Denseopolis TTI =6min/2min=3
that is just a ratio of travel times nothing about total commute times. Again very shocking that a bunch of transportation engineers would always interpret the results to mean you need to build more freeways.
Now as a graduate student in urban economics what bothers me about both of the reports cited in the article is they are not considering the fact that congestion/commuting costs are acting as a price.
CEOs for cities completely ignores the fact that a big reason that dense cities are dense is because of the per distance price of travel and so New York and Chicago are partially dense because of the high levels of congestion which TTI “finds”. Depending on price elasticities high per distance transport costs could possibly lead to lower total spending levels on transport and thus we could end up with the paradox of high congestion cities actually having lower commute times.
In my Job market paper I reject the use of TTI as a congestion measure for the exact opposite reason as CEOs for cities. Which is because it is likely to be over estimating congestion in disperse cities. The peak period travel time is essentially estimated as PPT=f(total vehicle miles traveled)/g(lane miles). Anything that actually lowers the unit price of travel or actual congestion other than an increase in lane miles of infrastructure could be expected to increase VMT and thus TTI, would be interpreted as an increase in congestion by the institute. Therefore no matter what else is working in a city to lower transport/congestion costs, Texas transportation institute’s response will always be to build more freeways.
Hey Stephen, it would be nice if you actually link to the whole GGW article.
Hey Stephen, it would be nice if you actually link to the whole GGW article.
Oops! Sorry about that…
Thank you sir!
This logical fallacy is very similar to that of per passenger-mile costs.
I don’t think your objection makes any sense. Passenger-miles are the standard unit for transportation cost-benefit analysis because transportation costs and benefits increase with both the number of persons and the distance of travel. You say “the problem with measure at the rate of distance traveled is that the purpose of transit is not to travel long distances.” But that is irrelevant. People wouldn’t be willing to pay the higher costs of longer trips if they didn’t get a commensurate benefit in return. To use your example of commuting, the typical reason people are willing to pay the higher costs of a longer commute is cheaper housing. The typical reason people are willing to pay the higher costs of a longer shopping trip is lower prices and greater selection. The typical reason people are willing to pay the higher costs of a longer trip to school is that the farther school is better than closer schools, or charges less for tuition, or somesuch. And so on.
I agree that “total costs per person” are also a measure of benefit. And the fact that we choose to spend far, far more per person on cars than transit demonstrates just how how much more valuable cars are to us than transit.
The fact that we choose to spend far, far more per person on high-fructose corn syrup than raw sugar demonstrates just how much more valuable HFCS is to us than sugar.
HFCS is for sissies and philistines. Men of taste know that the most valuable thing in America is a military toilet seat.
I’m going to say the same thing I said on Streetsblog: reading the TTI’s various congestion tables for large cities, I honestly don’t see a bias against transit cities and for sprawl. Los Angeles ranked first until this year, and now ranks third. Houston, once held as proof that building freeways cuts congestion, ranks fourth. At the other end of land use, New York ranks thirteenth, with the Solutions to Congestion Problems table citing transit savings equal to nearly half of the no-transit congestion level.
The main conclusion from the TTI tables is that large cities have more per-capita congestion than small cities.
Yes, it does. Is there a point to that observation?
First of all, what are they defining as “cities”? Is it city limits, or are we talking about metro areas? And also, you’re citing outcomes, but I’m more curious about the methodology.
The definition is census-defined urban areas.
I don’t remember the methodology right now, though it’s available online on the website (or at least used to be). As far as I remember, they’re comparing travel times to what travel times should be if there were no congestion, with different congestion-free speeds for freeways, arterials, and local roads. Do not quote me on this.
Oh wow…it’s more serious than I thought…
So that would be a no, then. One would have thought that a proponent of markets would understand that if people choose to spend more money on product X than product Y, it’s generally because product X provides more value to them. Why do you think people are willing to spend $800 a month to run a car if they would get the same benefit from an $80 monthly transit pass? Why do you think people are willing to put up with the higher cost and longer travel time of a 30-mile commute if they would get an equal benefit from a 1-mile commute?
Jim654,
“And the fact that we choose to spend far, far more per person on cars than transit demonstrates just how much more valuable cars are to us than transit.”
Government is so involved in our transportation system, urban planning/land use, and in the sugar market,that the outcomes in these markets probably (almost certainly) in no way reflect the population’s market preferences.
Government is heavily involved in every aspect of our economy. There are no perfect markets. If people wanted to change the degree or type of government involvement in the transportation and land use markets, they would. That’s how democracy works.
If you seriously think you can make the case that government involvement in the transportation market works in favor of automobiles rather than mass transit, please do so. The fact that government subsidizes mass transit at a vastly higher rate than automobiles very strongly suggests otherwise. An $80 monthly transit pass would cost more like $250 if it were priced at market rates.
“An $80 monthly transit pass would cost more like $250 if it were priced at market rates.”
And how much did you pay in tolls to drive around your local streets lately? You can’t at the same time make the revealed preferences argument for cars while brushing aside the fact that this is a very distorted market. As many people have pointed out above, that’s not a logical argument.
And how much did you pay in tolls to drive around your local streets lately?
Subsidies to local roads are on the order of 1 cent per passenger-mile, or maybe 3% of the total cost of driving. Subsidies to mass transit are on the order of 70 cents per passenger-mile, or maybe 70% of the total cost of mass transit. You cannot seriously claim that subsidies distort the market in favor of cars over mass transit.
I don’t know, but my point is that it’s definitely not as clear-cut as you make it sound. Right now, we have a system where a huge network of subsidized roads exist. That makes driving much more attractive, since this network makes getting around much easier. If instead we had huge networks of rapid transit and much smaller, skeleton networks of roads, I would bet that the subsidies would look much different.