George Will on transportation

George Will wrote a column at http://www.washingtonpost.com/wp-dyn/content/article/2007/03/09/AR200703...
which to some extent parrots the conventional road lobby wisdom. My thoughts on the relevant parts of his column:

The usual scolds -- environmentalists, urban "planners," enthusiasts for public transit (less than 5 percent of the workforce uses it) -- argue that more highways encourage more driving ("induced demand") and hence are self-defeating.

The "5 percent" argument is a self-fulfilling prophecy. For 40 years (1920s-60s) the federal government supported roads but did not support transit. For 40 more (1960s-2000s), the federal goverment spent 4 or 5 times as much on roads as transit. Now, if you spend more on A than on B, don't you think more people are going to use A? Or to put the question another way: if the government spent 40 years subsidizing transit to the exclusion of roads, and 40 more subsidizing transit to a much greater extent than roads, does anyone still think 90% of commuters would be in their cars?

"But as Ted Balaker and Sam Staley respond in their new book on congestion, " The Road More Traveled," among the 10 largest metropolitan areas, Los Angeles has the least pavement per person; Dallas has twice as much per person and half as much congestion."

First of all, it is pretty sloppy to quote one book's reference to only two cities as gospel truth on this kind of fact-intensive issue. More importantly, Dallas is catching up with Los Angeles in the congestion derby. In 1982, Los Angeles residents lost more than three times as many hours per peak period traveler to congestion (47 to Dallas' 13, according to the Texas Transportation Institute). In 2003, the ratio was only 93-60; congestion rose in both places, but much faster in Dallas. Obviously whatever Dallas is doing is not working.

"Furthermore, when new schools are built because old ones have become congested and then the new ones fill up with children from families attracted by new schools, who argues that building the new ones was a mistake?"

I don't quite get this analogy, for two reasons:

First, since when are new schools built because old ones become congested? In most of America, the old schools become deserted as families move to suburbs for newer, whiter, more upscale schools.

More importantly, there is a huge difference between schools and roads. A child can only use one school at a time; so when a child transfers to another school there is no net increase in school "congestion." By contrast, if you build a new road, a driver may use both the new roads AND other roads, thus increasing congestion on both. For example, suppose that Willville builds the new Sprawl Tollway, extending from a central city to Sprawl Valley (an outer suburb). Before the tollway was built, Mr. X lived and worked in the central city, driving 3 miles to work. After the tollway was built, X's job moved to Sprawl Valley. He now drives a dozen miles to work -3 miles on central city streets, AND 9 more on the Tollway. X drives just as much as he did on central city streets, AND is driving on the Tollway as well, thus congesting BOTH the "old" central city street and the new Tollway.