There is a push by some politicians and traffic engineers to enact a tax on vehicle miles traveled. Some traffic economists such as Randal O’Toole have proposed such taxes (Ending Congestion by Refinancing Highways, 2012, Policy Analysis 695, Cato Institute).

This would be a tax per distance traveled. In order to tax vehicle travel comprehensively, cars would carry transponders, devices that  receive and send radio signals. The car would transmit its location and perhaps other data such as its speed, time of day, and direction. The tax could be based on distance traveled, locations traversed, and the dates and times. Oregon and several other US states have already started to test such systems.

Vehicle tracking with satellites could be expensive, and could invade the privacy of drivers and passengers. The system could be used to track speed and automatically send traffic citations. The freedom to travel implies the privacy of transit. Other problems would include evasion by tampering with the devices, and hacking into the systems. In this era of massive surveillance, many would not trust that the data would not invade on privacy.

One advantage of universal tracking of vehicle travel is that it could facilitate taxing, not just distance, but also congestion. The system would charge vehicle owners a payment just high enough to prevent what would otherwise be congestion at that location and time. But congestion pricing can also be accomplished with tolls on the road, automatically paid by in-vehicle devices, and without disclosing the identity of the owner or driver.

As well stated by Randal O’Toole (p. 5),

“If tolls increase as the usage rate increases, and the maximum tolls are high enough that actual flows never exceed the maximum capacities, then road capacities are nearly doubled for those hours that flows would otherwise break down into stop-and go traffic.”

Another social cost, emissions, can be charged for with remote sensing devices combined with cameras. Therefore the remaining social cost is the wear of the pavement.

Heavy vehicles create much more wear than lighter ones, so a tax on miles traveled by such vehicles is justified. Indeed, a vehicle travel tax has been implemented for trucks in some European countries. But weight-based payments doe not require tracking; it is sufficient to record the odometer at the beginning and end of the year, with high fines for tampering.

Now consider the benefits of vehicle travel. There are two beneficiaries. The first is, of course, the travelers and the enterprises and customers of carried freight. The other beneficiaries are the owners of the land served. Better streets and transit make locations more attractive and productive, expanding the demand to be located there, and so increasing the land rent and land value. This rent increase is an implicit subsidy if not paid back. A land-value tax can pay for the street maintenance and prevent the subsidy.

However, one of the aims of the advocates of milage taxes is the reduction of property taxes. O’Toole, for example, states as a benefit of milage taxes that “property taxes that are spent on roads should... be repealed,” with no discussion of the increase in property values due to infrastructure.  

Consider how the owners of private streets cover the cost. The typical homeowners association with its own streets charge the member residents a monthly assessment. This is the private equivalent to a tax on the sites. If it is a large association, and there is traffic congestion, they would also put in tolls. The association would not bother keeping track of where the cars go.

Gasoline taxes have paid for road maintenance, but greater miles per gallon, electric cars, and not indexing the tax to inflation, have reduced this tax base. If land value taxes and charges on emissions and congestion are sufficient to pay the costs of roads, then neither gasoline nor milage taxes are justified.

Carbon taxes are most efficiently applied on carbon outputs - emissions - than carbon inputs. A tax on emissions creates an efficient price that covers all costs, while tax on gasoline distorts the market price of the input.

It is most equitable and efficient to place a charge directly on those who cause a cost and receive a benefit. A relatively light vehicle traveling one mile on a road does not have an inherent cost other than trivial wear of the pavement. There should not be a charge on mere travel. The charge should be on the costs imposed by that vehicle, namely congestion and pollution. The benefit to the riders is offset by the higher rent they pay to live work in the locations affected. Thus the tapping of the generated land value is equitable, efficient, and sufficient. There is no need to impose an intrusive vehicle-miles-traveled tax.