Forgive me for talking about high speed rail even though Ryan Avent says I'm not allowed to. I just can't control myself, I'm afraid.
As libertarians go, I'm a big fan of high-speed rail. I think it would be very nice if we had some in the Northeastern United States, where it might actually work, rather than the pathetic Acela that shaves a whole fifteen minutes off the trip between DC and New York. Unfortunately, that is apparently never going to happen, because in the United States, acquiring new rail rights-of-way seems to be virtually impossible. That means that the Acela has to run on existing track, which is not very good for high speed rail because, first, it was not designed for a train that accelerates to hundreds of mph, and second, there are other trains on it that don't go hundreds of mph, which slows everything down.
So we're not going to get true high speed rail in one of the two areas of the country where it stands a decent chance of working. Instead, we're going to get high speed rail in between Dallas and Houston or some other likely location--at least if the HSR part of the stimulus actually has its intended effect.
Ryan Avent defends these sorts of rail links on the grounds that if you view intercity rail as a substitute for air travel rather than car travel, they make sense. And I think that's fair. But I think that Avent underestimates the difficulties in doing this. The northeast corridor is the only place where people use trains as a substitute for cars for relatively short distances (sub 4-5 hours), because you don't need a car when you get there--and also, because road congestion makes car travel dicey, schedule-wise. This generally isn't much of a problem in flyover country.