It's possible to figure the environmental impact of public transportation in two different ways. So does it make more sense to focus on the average environmental impact of an additional rider (which basically entails dividing the total impact of public transportation by the number of riders) or the marginal environmental impact (which would mean calculating how much more energy is required or pollution is created when one more person gets on the bus)?

Often, the Lantern makes the case—as do many "cradle-to-grave" environmental analyses—that we should look at the average. (Consider, for example, last week's discussion of takeout, in which we were talking about averaging out the energy use of the restaurant across every meal.) Here's why: In many cases, looking at the marginal environmental burden of one consumer gives you a value close to zero. For example, as long as a plane has an empty seat, you aren't responsible for any of its emissions when you fly. So fly all you want! But in truth, each consumer contributes to the demand for a new restaurant or a new flight. With that in mind, it seems to make sense to focus on the average.

But in the case of public transportation, the Lantern makes an exception. We've already made the case that it's necessary to run off-peak buses and trains if you also want people to use public transportation at rush hour. If that's the case, then those off-peak buses will run whether there are five passengers or 25. It certainly makes sense to have these lonely buses run as infrequently as possible. But as long as some need to stay on the road to keep a public transit system functioning, you might as well hop aboard.