Richard Green captures the essence of the problem: he liked the train when he had it, but he is good at math. It is much easier to loudly advocate for high-speed rail if you can’t do math very well. Paul Krugman can, in fact, do math extremely well, and that’s why he cherrypicks the one location in the United States where HSR probably makes sense at this point: the US’s Northeast corridor, which is nothing but pavement from Boston to northern Virginia. It’s possible to look at the numbers, realize they are lousy, and still advocate for the project: we do it all the time. But it hurts your finer feelings if you are an analyst at heart.
One of my favorite colleagues at Berkeley is fond of saying “we don’t evaluate rail merely on cost-benefit analysis.” When pressed, she claims that rail somehow transcends the methodology, that there is no way to quantify how it revolutionizes a metro area. I’m the first to acknowledge that cost-benefit is a limited methodology, but honestly: if I ever hit up against a ready-made excuse for building bad projects, it’s the idea that your mode transcends cost-benefit calculus. Yes, I do think rail can have a powerful effect on urban land and human settlement patterns, but not all of those are positive. Rail makes sense in contexts, and it can set contexts, but it’s not like we can necessarily recreate San Francisco or DC using infrastructure alone. If she argued that we do a bad job costing out cost and benefits? Sure, I’ll buy that. But no; the literal interpretation is that what she wants transcends mere analysis. We’d all like to think that.
It seems to me what advocates are saying is that rail provides service quality improvements, and I do think that these are undercounted in cost-benefit analysis. Note that this does not save bad projects from John Kain [1} or Don Pickerell’s  infamous critiques: if you are offering service quality improvements and you’re not getting ridership, you don’t get to count the service quality improvement as a benefit (nor your emissions saved, etc etc).**
If it is the case that high speed rail offers service quality improvements, then it should be possible to quantify how much passengers value those service quality improvements by examining the cross elasticities between airfare and trainfare in countries where air and HSR are competitors. I don’t believe the California HSR plan does so, though the business plan suggests fares will be competitive with Southeast.
See? Those of you who can do math are trying to rough out how even a $34 billion project (let alone an $84 billion project) is going to retire its bond debt at $100 a ticket, and you’re getting a dull pain behind your left eyebrow, because even if you value carbon emissions savings at $200/ton, you don’t actually get a revenue stream from saved carbon emissions. This is what I mean. We all like trains–that’s not the issue. It’s that dull pain behind the eye you can’t get rid of.
I think one of the reasons why we haven’t done service quality calculation: we would have to be frank about the fact that HSR is high-end, luxury service way more like air travel than taking the car. People like me who have money are on planes (and HSR in Europe) all the time: people like Joe the Plumber (yeah, I know, I know) are not. And it’s not smart at this point of the HSR debate to be upfront of what we are proposing to build here: an expensive service for people like me rather than people like my mother in Iowa who never go anywhere and who would faint at the $350 ticket price it would take to get them to Chicago, where they would faint at the prices.
**And one of the irritating things about light rail and heavy rail is that ridership past a given point degrades service quality. One of the reasons I love the Gold Line in LA is that I can always get a seat. Great for me. Bad from an operations standpoint.
Manuscripts that informed this post:
 J. Kain. Deception in dallas: strategic misrepresentation in rail transit promotion and evaluation. Journal of the American Planning Association, 56:184–196, 1990. Journal Article.
 D. H. Pickrell. A desire named streetcar: fantasy and fact in rail transit planning. Journal of Transport Economics and Policy, 19(3):385–482, 1992. Journal Article.
 B. Flyvbjerg, N. Bruzelius, and W. Rothengatter. Mega-Projects and Risks: An Anatomy of Ambition. Cambridge University Press, 2003. Book, Whole.
 Theodore M. Porter. Trust in Numbers: The Pursuit of Objectivity in Science and Public Life., Princeton, N.J.: Princeton University Press, 1995. 311 pages