How Road Subsidies Might Cause Transit Subsidies

A recent article by Josh Barro admits that cars are subsidized through road spending, but argues that roads are less subsidized per capita because so much of car-related spending is private.  

It seems to me that (even leaving aside the issue I pointed out here, and environmentalist concerns about externalities) this argument overlooks the interrelationship between road subsidies and transit subsidies.

Often, gas tax revenues are used to build new roads and widen existing ones.  When these roads are in newly developing suburbs (or at least facilitate travel to those suburbs), the roads open up those suburbs for development- that is, they shift people and jobs from cities to suburban areas with limited or nonexistent transit service.  

Thus, the new road reduces transit ridership, which means farebox revenues from transit are lower, which means government has to choose between (1) subsidizing transit more heavily to achieve the same level of service or (2) reducing transit service.   Even if government chooses strategy 1, transit riders are effectively less mobile, since they can reach fewer of the region's jobs than they could before the road was built. 

It follows that this type of road spending creates one or more of the following harmful externalities:

1) it harms transit users by effectively reducing the number of available regional destinations (i.e. it effectively reduces transit service) and/or

2) it actually increases transit subsidies by making transit less economical and thus more highly subsidized.

Thus, gas taxes lead to expenses not accounted for by a simple look at the road budget.