Time for an Extreme Makeover (National Transportation Policy)
by Randal O’Toole, September 24, 2008
posted in Antiplanner Transportation, News commentary
“We need an extreme makeover of national transportation policy,” Robert Puentes of the Brookings Institution recently testified before the Senate Committee on Banking, Housing and Urban Affairs. Considering that this country has thrown more than $100 billion down the rail transit rathole and gotten virtually nothing in return, it is hard to argue with that.
Unfortunately, Puentes has the opposite in mind. After all, his testimony is titled, “Strengthening the Ability of Public Transit to Reduce Our Dependence on Foreign Oil.” His argument is filled with errors, references to shoddy research, and undocumented assumptions about the magical abilities of rail transit to solve all our problems.
Puentes presents his thesis in three parts:
1. High energy prices “have driven millions of commuters to mass transit.”
2. “Yet most metro areas are beset with limited transit.”
3. The tens of billions of dollars we have spent on transit in recent years “are not having the effect they could” because we don’t have enough transit-oriented developments.
Let’s look at each of these points.
First, the evidence we have about new transit riders is from the American Public Transportation Association’s quarterly reports on transit ridership. For the first quarter of 2008, APTA report 86 million more transit trips than the first quarter of 2007. For the second quarter, APTA reported an increase of 140 million transit trips.
Trips, of course, are not commuters. Assuming every single new trip was made by someone commuting to work, and adjusting for the number of work days each month, APTA data indicate that there were, at most, 1.5 million new transit commuters in May, but less than 1.1 million in June and well under a million in each of the preceding months. So, okay, maybe we can say that 1.1 or 1.5 million is “millions.” But it is a slight stretch since “millions” implies at least 2 million.
Second, Puentes argues that most urban areas have “limited transit.” What does he mean by that? Well, he says, “54 of the 100 largest metros do not have any rail transit service.” So, in Puentes’ mind at least, adequate transit means rail transit.
In fact, the Antiplanner has argued that rail transit is limited transit. First, it is so expensive that it limits the expansion of transit services in areas that don’t have rails. If we had spent just 20 percent of the more than $100 billion spent building rail transit since 1992 on bus improvements instead, our cities would have much better (and much more flexible) transit service today.
Second, rail transit only goes where the rails go. Which brings up Puentes’ third point: we need to build “denser, walkable, and transit-friendly communities” to make transit work better. In other words, instead of bringing buses to the people, Puentes wants to build rail lines and then bring the people to the rails.
This is, of course, standard smart-growth cant. Puentes backs it up with studies like this one, which claims to show that increasing densities by 1,000 housing units per square mile reduces per household driving by almost 1,200 miles per year.
There are huge flaws in this claim and the research that purportedly backs it up. Most important, the claim ignores the “self-selection” problem, which is that people who want to drive less tend to live in denser, more transit-friendly neighborhoods, while people who want to drive more tend to live in lower density, auto-friendly neighborhoods.
The research paper actually presents data that demonstrates this, but then misinterprets those data. Families with children, the paper notes, tend to live in lower-density areas and to drive more. Wealthier families also tend to live in lower-density areas and to drive more.
But the paper confuses cause and effect. Instead of saying, “families with children and wealthier families tend to drive more, so they choose to settle in lower-density neighborhoods,” the paper says, “families with children and wealthier families choose to settle in lower-density neighborhoods, and that forces them to drive more.”
A second flaw in this reasoning is the study’s focus on household densities and household travel. Households tend to be smaller in higher-density neighborhoods (fewer children, remember?). But, guess what, children benefit from mobility almost as much as adults. When measured on a per-capita basis, the effects of density on driving are much smaller than indicated by the study.
After all, Americans drive a per-capita average of 10,000 miles per year. So a difference of 1,200 miles per household is nothing if the households that drive less have an average of 12 percent fewer members.
Nevertheless, Puentes seizes on this and similarly flawed studies to argue that we need to reshape national transportation policy and turn it into a national land-use policy that would demand denser cities and more transit-oriented developments.
The Antiplanner has addressed this previously with the above graph showing, first, that the relationship between density and driving is weak and, second, to the extent that there is a relationship, large increases in density — 25 percent or more — would be needed to get even small reductions — 5 percent or less — in driving.
Finally, there are the implied assumptions behind Puentes’ arguments, most important the assumptions that transit saves energy and that the best way to save more energy is to invest in more transit. The Antiplanner has thoroughly debunked these assumptions before.
Puentes does make one good point: transportation spending today has “almost no focus on outcomes or performance.” As a result, “billions of federal transportation dollars are disbursed without meaningful direction.” Of course, real performance standards would probably not support Puentes’ rail transit or land-use agendas.
Those performance standards could include the cost per BTU of saving energy, the cost per ton of CO2-equivalent greenhouse gases, the cost per hour of congestion relief. No matter what standard you use (except the standard of “most money spent on least effective projects”), rail transit will fail almost every time. Similarly, I suspect that alternatives to high-density development, such as making single-family homes more energy efficient, will also prove more cost efficient than promoting wholesale reconstruction of our urban areas to higher densities.
The Antiplanner has no objection if someone wants to live in higher densities. If there really is a growing demand for higher density living, as New Urbanists say, then government should get out of the way and let developers meet that demand. But any proposals to build more rail transit or to arbitrarily increase the percentage of people who live in higher-density housing must be met with skepticism and careful analyses to insure that these really are the best ways to save energy and meet other worthwhile objectives.