Some variable cost differences. Broadly speaking, bus-based projects that use portions of existing roadway will be much cheaper than building rail for those same segments would be. Beyond that, costs for bus vs. rail projects can be hard to compare. Capital costs for rail include vehicles, while a busway is sometimes run with an existing bus fleet. Certain bus-rail comparisons in certain corridors may turn up significant differences in operating cost that may be valid in that situation, but need to be checked carefully to ensure that they assume the same factors on both sides.Well, sure it's hard, and we need to be careful. But let's focus on ongoing maintenance and operations. Is there really any disagreement on the fact that train cars last longer than buses, that railroads last longer than asphalt roads, and that maintenance per passenger is cheaper for rail infrastructure, both vehicles and roads? Similarly, is there any disagreement that trains are cheaper to operate per passenger, in terms of energy and personnel?
As I wrote last month, some time in the next century we will get to a point where it will be difficult to build more infrastructure, and the pace of construction will drop precipitously. The infrastructure we have then will be more or less what we will have for the following hundred years. The easier it is to maintain and the less energy it requires to operate, the more energy will be available for other purposes.
This is one of the reasons why I get so frustrated with organizations like the Los Angeles Bus Riders' Union that promote buses over trains. Less extreme but still frustrating are the Pratt Institute and the Straphangers Campaign here in New York, and the Institute for Transportation Development Policy which is based here but operates worldwide. There are a few arguments they make, some more defensible than others.
The most obviously bogus argument is that we should never spend anything on capital; all the money should go to running buses as cheaply as possible. If you take that to its extreme the buses will eventually fall apart, so nobody really believes that we shouldn't spend anything on capital. What remains is to find the right balance between capital and operating costs.
Those who argue that we should be spending less on capital owe us an explanation for why we should do this while the government is not spending less on car infrastructure. After all, it's the relative value of transit that will ultimately drive the mode shifts necessary to accomplish most of our goals, and if we stop investing in transit while others continue to invest in roads, the relative value of transit will decline.
The most defensible argument is that we should be investing in transit expansion, but it should be "BRT" (something better than a regular mixed-traffic local bus, to be eternally negotiated downward) and not rail. These arguments clearly ignore the long-term cost of maintenance and operations.
Suppose we could get a network of exclusive busways in all the corridors of the Commute Plan (for instance, PDF), or for the same price we could get exclusive rail of some kind in six of the eleven corridors (I'm just pulling that number out of my ass). And then we run out of oil, and only the rich can drive. People are left using the transit system that's built. Which would you rather have: the rail one that will last for years, or the roads that will get more and more potholed as the years go on?