Arguments over transportation policy often run as follows:
HIGHWAY SUPPORTER: Highways pay for themselves! Buses/trains don't! So highways good and everything else bad bad bad!
TRANSIT SUPPORTER: But highways create bad externalities like pollution and climate change! So if highways were taxed at their true cost gas would cost a zillion billion cajillion dollars per gallon! (followed by numerous counterarguments and counter-counterarguments that I won't bore you with, except as written below...)
It seems to me that these arguments miss one point: even if the highway system as a whole pays for itself, the system is so chock full of cross-subsidies that each individual road doesn't (except for toll roads).
Instead, one group of motorists is taxed to pay for projects that benefit another group of motorists. For example, the interstate highway system was built from gas taxes paid by drivers on existing, pre-interstate roads.
Were these drivers the beneficiaries of the interstates? In rural areas, probably yes. But in urban areas, only some of the urban drivers were helped by the interstates, while others were not.
Imagine, if you will, the city of Townsville (which could just as easily be Detroit, St. Louis etc). Townsville drivers pay a tax on their gasoline, which is used not to repair Townsville roads but to build an interstate that opens up the suburb of Sprawlville for development. 10 percent of Townsville residents move to Sprawlville; 1 percent of Townsville residents have their home destroyed to build the interstate.
The 10 percent who move to Sprawlville get a windfall; they pay only 10 percent of the gas tax while the interstate is being built, but get most of the benefits (since they get new homes they like). The 1 percent who have their homes destroyed are harmed quite a bit, because they lose their homes and neighborhoods (and usually are not compensated, especially if they are renters).
The other 89 percent (the people who want to stay in Townsville) may on balance have been harmed as well. Since they don't go to Sprawville they might not benefit very much from the highway. Also, their existing neighborhoods often declined because of the highway(s). When the 10 percent moved to suburbia, the 89 percent suddenly had vacant housing in their neighborhoods, housing that was usually filled by poorer people, often causing declining property values, worse schools, and increased crime. And if Sprawlville was a separate city, Townsville's tax base started to decline, causing higher taxes, making Townville less desirable. So eventually many of the other 89 percent was forced out of Townsville by the highway, just as surely as if they had lost their homes to eminent domain.
In other words, users of new roads benefit from taxes paid by users of existing roads (who are not always the same people). The only way to avoid this problem is to make new roads toll roads and to close them down if they don't break even.
Comments
Michael, you're late to the party
these arguments have been going on for a long time - decades. The latest and greatest "solutions" are beyond toll roads. It involves reading license plate tags and charging per mile by vehicle weight. Roads could be leased to companies similar to how some utilities are operated, price regulated, etc. Transportation, especially motorized, is a really kind of a natual monopoly so it makes some sense.
Even with these options, there are always detractors. What about externalities? Air pollution, groundwater pollution from tires, emergency services provided, etc. These could be priced into the charges, society willing. Although we know society is almost never *willing* to pay for such things.
Total privatization: possible, not likely. Nobody wants abandoned roads, though my speculation is that is a straw man argument since if not financially performing, it would just be sold to someone who bids lower, like a real estate project (which is exactly what it is, really). Plus, there are a whole host of people who basically think it's un-american to sell "our" roads to international consortiums of investors (read "foreigners" in a stump speech).
The entire argument would be incomplete if we didn't metion that transit is not close to paying for itself either (direct or externalities). The difference between the two modes is that given current land use patterns, one is a viable business model and one isn't.
Most people aren't going to be happy with any of the proposed options which is why we are probably stuck where we are.
I was actually thinking about writing on CP's points
... but I figured I'd write a blog post instead of a book! In particular, it occurred to me that even if a transit system broke even, the same sorts of cross-subsidy issues might be involved, since fares paid by users of old transit lines would probably be used to subsidize newer lines. (In fact, that arguably occurs today- I would guess that a new rail line is probably more subsidized per passenger than an old bus route).
Shared resources
It seems to be a much more general problem tied to developing large and ultimately shared resources such as cities. A more recent discussion as property tax revenues have proportionally declined with such measures as Prop 13 in California is that of impact fees as a way to finance new growth. Here, though, new residents in effect pay to join the party, and although a nexus has to be established linking new infrastructure need to new growth so as to not overcharge the new project, it opens up some of the same issues of who pays, benefits which I would say is endemic to large infrastructure bundles new and old such as cities.