Transit is in the business of moving people, not relieving congestionJuly 18, 2010 by: Samuel Scheib
As regularly as an equatorial sunrise, Google Alerts sends me news clips from around the world of transit. From these articles I know light rail is making progress in the Carolinas, transit ridership is increasing in car-mad Phoenix, and congressional candidates are talking about transit on the campaign trail. Unfortunately, most of these missives from the transit field carry with them the promise of failure: congestion relief.
Congestion relief is nearly always listed as one among many benefits of using transit: reducing pollution, saving money, reducing VMT, providing mobility to disadvantaged populations. These are all worthy goals but congestion relief is a chimera, something we transit professionals simply cannot provide.
To begin with, congestion is a product of success. Country roads are quiet and bucolic, the exact opposite of city streets simply because city streets contain places people want to go. In this context success does not imply that a place is well-planned and well-built, but rather that it is a place where there are jobs and things people enjoy doing.
Where there is work, shopping, and entertainment, there are people and that is a good thing, but people inevitably lead to congestion whether with horse carts and pedestrians in ancient Athens, bicycles fighting for space on the streets of Beijing, or the automobile-dominated streetscapes of the United States. In the annals of transportation history there has yet to be a vibrant city that did not experience congestion. The uncongested city is not inconceivable—Frank Lloyd Wright produced a detailed scale model of exactly that in 1935 with his Broadacre City—just not buildable. The notion that bigger roads provide for less congestion is specious. If this were true Atlanta and Los Angeles would enjoy free-flowing roads all day.
A new or wider road may not relieve congestion, but why not a transit project? There is the matter of scale. The 2001 National Household Travel Survey reported 91.2% of trips to work were in personal vehicles while transit captured only 4.9% of the mode share—and this including New York, a city that accounts for roughly 25% of all transit trips in the nation. Even a 100% increase in transit use would still account for fewer than 10% of work trips. The gains in auto efficiency caused by such an increase in transit (or by a widened road or new interstate for that matter) would soon be consumed by triple convergence and latent trips, an idea popularized by Anthony Downs in his books Stuck in Traffic and Still Stuck in Traffic. Commuters who would have otherwise traveled at a different time (time shift), on a different path (route shift), or by some other vehicle (mode shift) will take advantage of the new capacity by driving to work during rush hour on the most direct route. If you ever delayed a trip to the store to avoid rush hour traffic, you made a latent trip and those become real trips when additional capacity becomes available.
There is a nasty piece of circular logic at work in the idea that transit makes room on the road: if more people took transit to work, then more road capacity would become available, encouraging more people drive to work. This is the great fallacy of congestion relief, that some drivers will opt to take transit to work so that other drivers will have an easier time driving to work (a concept brilliantly explored by the satirical newspaper The Onion in a piece called “98% of U.S. Commuters Favor Public Transportation for Others”). There are many great reasons to use transit but altruism must rank as the least compelling.
Another reason to shy away from singing the praises of traffic relief is that there is no statistical relationship between transit ridership and congestion. Looking at the 65 most populated cities in the U.S. (excluding NY as an outlier) the Texas Transportation Institute ranks L.A. as the most congested (93 hours of annual delay per driver). L.A. has the most passenger trips in the data set. The second highest number of passenger trips belongs to Chicago, ranked 7th in delay in this set. The third most trips are in Philadelphia, 23rd in delay, then Washington D.C., ranked third in delay. Accounting for population, Philadelphia (23rd in delay) has the highest number of passenger trips per capita with Honolulu coming in second (ranked 43rd in delay). L.A. is 12th in per capita passenger trips.
The utter lack of consistency means that the affect transit has on congestion cannot be measured. In Houston, for example, all the measures TTI uses to index congestion showed an increase in congestion from 2002 to 2007 except the number of rush hours (steady at 7.8). During that same period transit ridership increased from 81.3 million passenger trips to 97.4 million, no doubt in part because of the addition of light rail in 2004. Even with a 20% increase in transit ridership, the congestion indices for Houston remained constant or got worse. There is simply no evidence to support the claim of transit as congestion relief, a notion on which countless transit projects are based.
By touting congestion relief, transit providers are handing our severest critics potent ammunition to use against us. To wit, a certain anti-rail crusader has used a comparison between Atlanta and Portland as strong evidence against light rail. He wrote in 1999 that since 1982 congestion in Portland increased 33 percent compared to 36 percent for Atlanta. He failed to mention that Atlanta’s congestion increased three percentage points more than Portland’s despite Atlanta’s insistence on building roads and Portland’s insistence on building transit projects, but what he said was true. As congestion worsens, transit will not provide relief.
If San Francisco suddenly lost its wonderful BART system and its 300,000 daily trips, arguably some or most of those passengers would drive alone on the most direct path at the most convenient time (peak hour). Those additional trips would certainly, absolutely, add to the congestion of San Francisco’s roads and that if BART then reappeared transit trips would again replace some automobile trips. Similarly, if L.A. had one-quarter as many highways as it does it would be far more congested than it is now. But no one would say that L.A. is not a congested city, nor could this be said of San Francisco despite the presence of an outstanding transit agency.
Transit does have the ability to remove cars from the road and that can be measured by vehicle miles traveled (VMT). That may seem to contradict my previous argument, but VMT and congestion are not interchangeable because congestion is not about cars but rather about perception. A planning professor I know moved to Tallahassee from L.A. and was pleased to find there was no traffic congestion—and surprised to find that his neighbors and colleagues complained about traffic congestion all the time.
All cities with great transit systems—New York, Portland, San Diego, San Francisco, Chicago, Moscow, Paris, London, etc.—also have streets glutted with automobiles during peak travel times. When transit projects are based on the ability to relieve traffic congestion the proponents of the project are setting themselves up for failure because whether or not there is a significant reduction in the number of vehicle trips people view delays during peak travel times as congestion whether in LA or Tallahassee. Transit simply cannot remove enough cars from enough roads to remove the public’s perception that there is traffic congestion. Transit can give people time to read, work, or catch a catnap. It aids the transportation disadvantaged. Most importantly, transit offers commuters the choice not to have to sit, white-knuckled, in traffic, and that is a promise transit can keep.