So as you’ve probably noticed if you’ve been reading for awhile, I don’t have a lot of patience for the complaining rail advocates indulge in when they claim that “rail is subjected to higher standards” than highways/any other public investment. I think that’s largely just self-serving nonsense, and, moreover, it’s an empirical statement, so back up the whinge with actual analysis that shows rail projects are held to higher B-C ratios than water infrastructure or new ozone standards if you want me on board. In general, all new spending initiatives get the business in the hurly burly of politics.

However, if you did want to argue about fair play in B-C analysis, there are I think two things about B-C in highway investment that should be challenged.

1. The large numbers game. Many highway improvements get a nice boost on the benefit calculation even if the estimate of delay reductions per person are low simply because some highways serve thousands of drivers and passengers a day. With that client group, a two-minute delay reduction can look pretty good in the benefit column even if it’s doubtful that overall welfare really improves from scattering small time benefits across large groups. It seems like we do hit an asymptote at some point. Look! I save a trillion people a nanosecond!!! YAY ME AND MY PROJECT.

2. The safety benefit. This one is harder to describe, but highway projects get a pretty big benefit bump if they promise safety improvements, and most do. Most transit projects do not claim safety improvements, even if they claim environmental benefits. The question for me: in what world does it make sense to count not killing people as a benefit rather than the baseline expectation of a mobility system? We have a system that kills people, routinely, and we accept this state of affairs because cars are convenient, blaming drunks, bad drivers, and God instead of technology and a system that routinely fails to protect people. And when we want to invest more in that system, we use the fact that it will kill and maim marginally fewer to rationalize that investment to expand the killing system. WOO benefit. If not killing people is such a great thing, how about we not kill even more people and shut the damn thing down entirely, then.

Now, of course, you want to invest in things where many people benefit and where investment protects human life, but I think you really do need to be making substantial benefits on the first to be legitimate. And with the last, we should be thinking differently about how safety should work.