Ed Glaeser has an interesting post attempting to run the numbers for high speed rail. For my money, the most important point is that the thing that makes high speed rail problematic in most parts of the country is the absence of a tightly concentrated central business district. High speed rail makes sense in the Northeast corridor, because when you get off the train you are very close to where you’re likely to want to be. And if you’re not, you’re connected to a public transit system that can get you there. Trains therefore have a considerable advantage over planes, since you’re spared the time and expense of getting from the airport to downtown. But what happens if you build a supertrain from Houston to Dallas. There is just not enough population or business density in a single spot on either end, or a well-developed public transit system. So you’d have to rent a car on either end and get on the highway, and where you’d want to get to could be 20 miles away. Maybe having a supertrain system would encourage more density in Houston or Dallas, but my sense is that those cities are pretty much what they are. This has all the signs of a costly boondoggle.
That said, there are a lot of places in the country that could probably benefit from fast trains, because of what’s on either end. The corridor from Portland to Vancouver, almost certainly. Boston to DC, absolutely. And substantial parts of the country could benefit from just decent train service–Amtrak is criminally slow, and places like Florida could probably use a system of trains at the speed of your average continental European train. But Texas–and substantial parts of the rest of the country–aren’t going to be able to sustain a high speed rail system until they get a lot denser than they are now. And I don’t see how that’s going to happen.