Ryan Avent Demolishes Ed Glaeser's Attack on HSR (Source: California High Speed Rail Blog)

California High Speed Rail Blog: Ryan Avent Demolishes Ed Glaeser's Attack on HSR

Wednesday, August 5, 2009
Ryan Avent Demolishes Ed Glaeser's Attack on HSR

Harvard economist Ed Glaeser posted the second in his HSR evaluation series for the New York Times' Economix Blog yesterday. There are several problems with his study, particularly his choice of Dallas-Houston as his example to assess HSR costs. Matthew Yglesias criticized this pick as being unrepresentative and not even being part of the official USDOT HSR route map. I wouldn't hang my hat on that latter factor to undermine Glaeser, since the USDOT HSR map will be updated this fall, and will likely include Dallas-Houston, which is part of the "Texas T-Bone" HSR project.

But is Dallas-Houston a representative corridor? Ryan Avent, writing at Streetsblog Capitol Hill, argues it isn't:

Why would he choose this corridor to examine? Why not begin with the most natural place to construct true HSR -- the Northeastern Corridor -- or the state moving fastest toward building its own true HSR network -- California?

Well, Glaeser was able to use Dallas' low share of commuters taking transit to knock the corridor's estimated ridership down by half. Transit's share of commuting in Los Angeles is nearly three times that in Dallas. In San Francisco, transit's share, at 32.2 percent, is more than seven times larger than in Dallas. Presumably this difference had something to do with his choice.


The Texas T-Bone scored pretty low on The Transport Politic's assessment of US HSR routes. Obviously Glaeser has not picked a representative sample. But Avent argues Glaeser's approach is more fundamentally flawed because of how his metrics work:

This is a bad beginning for Glaeser, but it actually gets worse. He presents a formula for determining whether the direct benefits of rail are worth the costs:

Number of Riders times (Benefit per Rider minus Variable Costs per Rider) minus Fixed Costs.

That seems simple, does it not? Perhaps a little oversimplified? But it must be so, says Glaeser:

I’m simplifying, but a formula needs to be simple if interested parties can seriously debate the numbers, and the only way that America is going to get to the right answer on public investments is if numbers trump rhetoric.

But it matters which numbers we're considering, and omission of important variables that planning experts take seriously is not the way to conduct this debate.

The simple fact is that Glaeser's stripped-down formula obscures far more than it reveals. Again, as I mentioned at the beginning, I am hesitant to judge this series a mere one part in, but the way he has begun here is simply irresponsible.

What are his long-term assumptions? How quickly does he think the population of the Dallas and Houston metropolitan areas will grow? What will that population growth do to the number of people living within easy reach of a train station? How will that population growth interact with planned expansions of local transit systems?

How sensitive are his projections of changes in oil prices? Do they take into account the effect of changing demographics on demand for various kinds of housing and transportation?


In short, Glaeser has left an enormous amount of stuff out of his calculations, exactly as I predicted he would. Avent makes the point Morris Brown expected me to make about the cost of doing nothing:

And that brings us to a final point (which, again, Glaeser may ultimately address): What is the proposed alternative?

Is it doing nothing? Then at what point does the rising cost of congestion justify construction of something? Let's say an alternative is new airport capacity; well, how do the costs and benefits there work out, and how does that math change with oil at $150 per barrel?

Or perhaps an alternative is new highway capacity. Can we see a cost-benefit analysis for that, and how that varies with oil prices, congestion levels, and so on? If we assume that drivers will need to pay the full maintenance cost of the highway network already constructed via a user fee (and currently they're coming up well short), what does that do to expected demand for rail?

Even if you accept the numbers that Glaeser uses (and one shouldn't automatically do so), you're left with almost nothing -- an amateurish, back-of-the-envelope analysis for a corridor that's not even part of the current Obama administration plan. What is this supposed to prove, exactly?


As is typical with conservative economists, HSR is treated as if it will exist in a vacuum, unrelated to any other changes in transportation, land use, oil prices, carbon taxes, population growth, or other costs.

I'm with Avent on this - Glaeser's metrics don't hold up. I'm curious to see part three, but I am not any less doubtful than I was after part one.
Posted by Robert Cruickshank at 4:23 PM