Vehicle Miles Traveled vs. Pay-at-the-Pump Gas Tax


For some years now, policy wonks have proposed replacing the pay-at-the-pump gas tax with a “vehicle miles traveled” system. The reasons take a few different paths but are mostly centered around the issue of the fairness of “user fees” compared to purchase taxation, an idea made more relevant by the proliferation of non-gas using electric cars.

A few states have already and/or are conducting pilot programs to test the idea and others – and quite possibly the federal government – are looking at how a VMT tax would not only work from a practical standpoint but also part of a drive to modify driving habits.

As the concept is definitely gaining steam – or electricity, as it were – amongst transportation planning types and certain politicians, it may be best to look at a general primer on the subject and the terms and ideas that will soon be coming to a legislative body near you.

Basically, a VMT would be just that – a tax based on how many miles a vehicle moves over the roadways instead of paying a tax every time the gas tank is filled. Proponents say it is intended to be “cost neutral” for the typical driver.

For example, right now – and these are general AVERAGES – Americans drive each personal car about 12,500 miles per year. At a typical miles per gallon figure of 25, that means that – using a national usage-weighted average of about 57 cents a gallon in state and federal gas taxes - a typical motorist will pay about $285 per year, or 2.3 cents per mile, in taxes each year. (It must be stressed that these are national averages – for example, in California the per mile tax works out to more than double at 4.8 cents. Also, these numbers do not apply to diesel fuel usage, though any VMT would include those vehicle types, presumably at a higher per mile rate). Most VMT concepts, depending upon the state, have so far floated a per-mile number in a range close to the current cost as their target.

That is the basic top-line concept– but exactly how it would be implemented is not at all clear.

There are five basic ways the VMT per-mile rates could be calculated:

  • Straight miles – turn the car on, drive 10 miles – no matter where or when – and pay 10 miles in taxes.
  • Highway/major artery miles – no (or a lower) fee for driving around town but the rate would ramp up when one entered a designated freeway and/or major artery.
  • Congestion miles – a lower tax on any driving during “off-peak” hours, but miles driven during typically heavy traffic times for that specific roadway would be taxed at a higher rate.
  • Time miles – Similar to congestion miles, but not tied to specific traffic patterns or roadways, but just the time of the trip – middle of the night lower tax, mornings higher tax, etc.
  • Cordon miles – Driving in certain areas and/or on certain roads would be taxed more while driving elsewhere would be taxed less (this could also be done not just by the mile but also an added flat tax/fee upon entering a particular area or based on how long the vehicle remained in the area).

In theory, the following taxing scenarios could occur on a single 50-mile drive.

Under “straight miles only,” it would be about 50 miles of tax and that would be that.
Under highway/arterial miles, it would be the miles you drove on those designated roadways. The same would hold true for congestion and/or time miles, but the tax rate could vary depending upon when you take the trip.

How cordon mile pricing works is rather intricate. Under a VMT system, a government agency (local, state and/or federal) would pre-determine which areas – not specific roads – have the most problematic traffic and charge additional fees when the driver is in that neighborhood (akin to the charge – about $20 - drivers in London, England now must pay each time they cross the imaginary line into the hub of the city). If in this speculative 50-mile trip, the vehicle passed through one – or even multiple - cordon areas, the additional fees would be added onto the regular rate (cordon taxes could also be imposed as flat per-entry fee or as a stand-alone charge, separate from and/or in addition to any gas tax replacement scheme).

And now for the complicated part – each of these types of fees could be imposed separately or be combined in some fashion depending upon the system installed. For example, a distance plus cordon fee would result in a combination of actual miles traveled with where those miles were traveled; a distance plus time fee could mean that the driver is paying 2 cents a mile when they leave and 4 cents when they arrive, a miles plus congestion model would charge based on which route is taken.

As to cordon fees, they are usually justified not as an overall gas tax replacement but as a way to decrease traffic in highly congested areas and are the most likely “add-on” concept. Most – but not all – cordon proposals discussed so far exempt people who live in the area from paying the fee for that area.

And now for another complicated part - how any VMT would be actually charged is not yet determined. The pilot programs have relied on highly motivated volunteers self-reporting mileages, but it is highly improbable that such a system could work on any large scale as it would rely solely on the memory and honesty of the average taxpayer.

Almost certainly the implementation would involve either a transponder in each vehicle or an app that drivers could download and sync to their car. Either system would track exactly where and when a car was driving and calculate the tax based on whichever fee structure(s) is being used. If a transponder is used, it would simply trace the car moving while the smart phone app would, presumably, track the location of both the car and the specific driver.

Both the app and transponder solutions raise serious data privacy issues . Even VMT proponents admit that is a major stumbling block to eventual implementation, as the chances that the public will embrace allowing the government to know exactly where they are at any given time is minimal, to say the least. There have been various privacy protection proposals made, but, given the interconnected nature of society in general and the VMT concept specifically, it is not clear that any amount of data security assurance would be enough for the public.

And now for another complicated part - beyond the personal liberty issues are the myriad ways the system, once in place, could be modified to benefit certain people, potentially negating its core concept of everyone paying their proper proportional part.

First, there is the income differential issue. Some envision a sliding scale component to the tax, meaning that lower-income residents would actually pay a reduced per mile fee. In other words, if a person receives any public assistance, they could automatically qualify for lower per mile charge. Alternatively, the system could simply compare vehicle registration information to state and/or federal tax return data to determine who gets the discount, etc. (and, yes, that could mean the DMV and the IRS would have to closely “cooperate,” perhaps a scary thought).

Other fee modifications – which could cause the concept to be viewed not strictly as a fair tax but as an attempt to leverage the system to modify public behavior – could include:

  • Fuel efficiency – if it is decided to use the fee to encourage cars that use fewer fossil fuels, then a discount could be applied to cars that getter better mileage and/or some form of the current discount drivers of electrics – which pay no gas taxes - now receive could be kept.
  • Age of vehicle – to weed them out older, more polluting vehicles could face a higher per mile tax.
  • Single car discount – to encourage carpooling and just driving less in general, a household that has only one vehicle could pay a lower rate.
  • Transit discount – a person who regularly uses public transit (say they have an annual bus pass, for example) could also see a different rate.
  • High use rates – to discourage “over driving” – read commuting – the rate could be a sliding one that charges more per mile the more you drive (3 cents for the first 15,000 miles, 4 cents beyond that, etc.)

It should be noted that each of the above concepts would/could be applied to commercial vehicles and diesel trucks under different rate structures and accounting for other factors inherent to those types of uses.

A significant concern is how a VMT would affect people based upon where they choose to live.

If a VMT is implemented as a one-to-one revenue swap, the impacts would be relatively similar to the current gas tax system. However, the likelihood of that being the case is minimal as the VMT system opens up multiple opportunities to guide driving - and, therefore, living - habits that few legislators and bureaucrats and planners would willingly eschew. That means that the VMT will impact city dwellers only slightly, while, conversely, heavily penalizing suburban, exurban, and rural populations and could be seen as another pincer movement in the on-going push to increase urban housing and population densities.

And now for another complicated part – borders. If, for example, the tax at the pump is removed on a state-by-state basis, how would out-of-state drivers be made to pay their “fair share?”

A federal system would seem to solve certain problems associated with the VMT concept, including the ability to properly and accurately apportion the revenue (X percent for the feds, Y for the state, Z for any local taxes, etc.). The federal concept could also simplify the transponder issues by making such tracking devices mandatory, like seat belts for example, for all new cars.

The downsides to any federal effort are obvious, particularly the political firestorm that would erupt in Congress. The legality of a federal law is also in doubt, though this could be wiggled around in the same way the 55 mile-per-hour speed limit (and the 21-year-old drinking age) was imposed.

And none of the above addresses if there would be any modification to how the money – once collected – is spent in theory, the tax income could be divvied up immediately, with, for example, 20 percent going straight to the state, 15 percent to the city, 20 percent to the county, 10 percent dedicated to transit, 20 percent for the feds, etc.

In whatever form the VMT concept presents itself – no matter its possible technical benefits - it will be a hard sell to the public - and very possibly deservedly so.

Thomas Buckley is the former Mayor of Lake Elsinore and a former newspaper reporter. He is currently the operator of a small communications and planning consultancy and can be reached directly at You can read more of his work at

Photo credit: Ted Eytan Flickr under CC 2.0 License.