Taxing the miles you drive, or the gas you burn?

I've been thinking about a proposal being floated in California (see this article in the LA Times), ever since I saw a reference to it by Kevin Drum . Joan Borucki, appointed by Arnold Schwarzenegger to be the director of Californians Department of Motor Vehicles, has proposed ending that state's tax on gasoline, and replacing it with a tax on the number of miles driven per year.

Ignoring for the moment the issues surrounding the technology needed to do this, and the privacy concerns, and any questions one might raise about taxing California drivers for miles they drive out of state, what does this look like as a tax?

Let's start with the traditional purpose of gasoline taxes. In most states, and at the federal level, they have been dedicated taxes, directed to funds reserved for building and maintaining roads. That is, they function as user taxes. In this sense, shifting to a (revenue-neutral) mileage tax would seem to accomplish the same objective, if those funds were similarly dedicated. An argument is that the mileage tax functions better as a user tax than does a gasoline tax, because (to some extent) it's the amount you drive, not necessarily the gasoline you burn, that defines the benefits you get from the highway system. On the other hand, less fuel efficient cars are heavier and cause greater wear. For equal mileage driven, larger, heavier cars pay more in tax. Which functions better as a user tax, then, is not immediately clear.

But that's not the only issue to deal with. For one thing, gasoline taxes have other consequences. And shifting to a mileage tax also has other consequences.

First, gasoline taxes raise the price of gasoline relative to other energy sources. So they provide an incentive, even if only a minor one, for the development of automotive fuels other than gasoline. And because these alternative fuels are likely to generate less pollution, a gasoline tax can also be thought of as a tax on the pollution caused by burning gasoline. Indeed, the gas-tax-as-pollution-tax strand may lead to an argument that the gas tax should be increased (and offset by reductions in other taxes in a revenue-neutral way).

Shifting to a mileage tax, of course, shifts away from the pollution-tax aspects of the gas tax. It even, perversely, works in the opposite direction. Unmodified, the mileage tax shifts taxes from people who drive fuel-inefficient vehicles to those who drive fuel-efficient vehicles. Consider two people each driving 15,000 miles a year. If one drives a car that gets 40 miles per gallon, that driver uses 375 gallons of gas per year; if the gas tax is $0.60 per gallon, s/he pays $225 a year in gas taxes. If the second driver gets 15 mpg, that's 1,000 gallons of gas, or $600 per year. Now shift to a mileage tax that's revenue neutral--$825 for 30,000 miles, or $0.0275 per mile. Both drivers now pay $412.50 per year. Note that the effect is to "punish" the driver with the fuel-efficient car, and to (at the margin) discourage fuel conservation, which (at the margin) will lead to increased use of gasoline and increased air pollution.

Second, we know that lower-income people are, other things equal, likely to buy more fuel-efficient vehicles than are higher-income-people. So the mileage tax is also a shift in the tax burden from higher to lower income taxpayers.

The LA Times article suggests that these effects can be adjusted for by varying the per-mile tax based on the fuel efficiency of cars. This (and here I'm not ignoring administrative complexity) would significantly increase the administrative complexity of the system. Any system with multiple tax rates is harder to administer than a system with one tax rate.

[To continue with the administrative problems. California had about 35.5 million people--about 20 million households, about 2.75 million businesses, and about 30 million motor vehicle registrations, and a half a million motorcycles, in 2002, and collected about $3.3 billion in motor fuel taxes. There were about 24,000 gas stations. (All data from the Statistical Abstract of the United States.) I don't know about you, but collecting taxes from 24,000 sources seems easier than collecting taxes from the roughly 20 million to 22 million taxpaying units--households plus businesses with motor vehicles--that'd pay a mileage tax.]

Take all this together, and then think about the shift to a mileage tax. The environmental consequences would be bad. The distributional consequences would be bad (well, in my opinion, anyway). The administrative complexity of collecting the tax would be greater. And in favor? That maybe--maybe--a mileage tax would work better as a user tax.