Higher Gear

To Fund Transportation, Congress Should Tax Travel, Not Fuel

By Paul Sorensen, Liisa Ecola, and Martin Wachs

Paul Sorensen is a RAND operations researcher with expertise in transportation planning. Liisa Ecola is a RAND senior project associate and city planner. Martin Wachs is director of RAND’s Transportation, Space, and Technology Program and professor emeritus of city and regional planning and of civil and environmental engineering at the University of California, Berkeley.

With the pending reauthorization of the federal transportation bill, the U.S. Congress can steer the country toward a better path of financing the U.S. transportation system. We have identified promising options for shifting the country away from its increasingly obsolete method of relying on motor fuel taxes to fund highways and toward a method that would be fairer and more effective.

It is now possible to charge drivers directly for the miles they travel instead of relying on fuel taxes and other indirect fees.

Motor fuel taxes, consisting of gasoline and diesel taxes, have long been the mainstay of highway finance at both the federal and state levels. Taxes on motor fuels are typically levied on a cents-per-gallon basis. This means that they must be periodically raised to keep pace with inflation and improved fuel economy. Elected officials, however, have grown increasingly reluctant to take on this politically unpopular task. As a result, fuel tax receipts have fallen precipitously over recent decades when measured in terms of real dollars per mile of travel, leaving insufficient revenue to maintain, let alone expand, the road network. The anticipated introduction of more fuel-efficient vehicles and alternative fuel options in the coming years, though desirable on other grounds, will accelerate the deterioration of highway revenue.

Traffic moves along the 405 Freeway in Los Angeles in October 2010.


Traffic moves along the 405 Freeway in Los Angeles in October 2010. The freeway facelift has caused so much gridlock that commuters are tweeting their complaints while driving. The $1-billion upgrade requires complete closure at times.

The good news is that the next reauthorization of federal transportation programs presents Congress with the opportunity to start transforming how the nation raises funds for its roads. Harnessing innovative technologies, it is now possible to charge drivers directly for the miles they travel instead of relying on fuel taxes and other indirect fees. Shifting to charges based on vehicle-miles of travel (VMT) would ensure that all drivers pay in proportion to their use of the roads. If Congress fails to make this shift, the country will miss a critical opportunity to develop a more sustainable source of highway funding, to reduce traffic congestion, and to price travel more fairly than has been the case for close to a century.

Mileage-based road use fees could be implemented in various ways, but three options appear to offer the greatest promise: (1) estimating mileage based on a vehicle’s fuel economy and fuel consumption, (2) metering mileage based on a device that combines cellular service with a connection to the onboard diagnostics port, and (3) metering mileage based on a device that contains a global positioning system (GPS) receiver. Significant political, institutional, and cost-related uncertainties make it difficult to determine the optimal configuration of equipment at this juncture. But in its reauthorization of the transportation bill, Congress should set aside funds for activities that could resolve these uncertainties by testing the alternatives in realistic settings, thus setting the stage for implementing VMT fees within the coming decade.

A New Set of Wheels

For nearly a century, motor fuel taxes have been the primary source of funds for constructing and maintaining the nation’s highways. Adopted as an indirect means of charging drivers for the costs of using the highway system, fuel taxes now account for almost two-thirds of highway user fees and about half of highway expenditures. But political resistance to tax increases of any sort, dramatic increases in fuel economy, and the rise of alternative energies (such as biomass and electricity) to power cars have eroded the role of fuel taxes as a reliable revenue source.

One problem is that the federal gas tax was last raised to 18.4 cents per gallon in 1993 and has since lost approximately a third of its value to inflation. Additionally, while the total number of vehicle-miles traveled in the United States has doubled over the past three decades, fuel consumption has risen by only 50 percent during that time. Better fuel economy is good for the economy, energy independence, and reduced air pollution. But better fuel economy also means that motorists drive more miles with each fill-up at the pump and pay substantially less through fuel taxes per mile of driving than they did in past years.

In contrast, VMT fees, enabled by new electronics and communication technologies, could potentially generate a reliable revenue stream while upholding the core principle that users should pay for the system in proportion to their use. VMT fees would be based on the amount of travel rather than the amount or type of fuel. VMT fees could also be set higher for peak-hour travel in congested corridors to help reduce congestion, and the fees could vary with vehicle emission characteristics to encourage more rapid adoption of less polluting vehicles.

Figure 1 — Historical and Forecasted Growth in Miles Traveled by Vehicles Far Outstrip That of Fuel Consumption

Historical and Forecasted Growth in Miles Traveled by Vehicles Far Outstrip That of Fuel Consumption

SOURCE: Highway Statistics 2007, Washington, D.C.: Federal Highway Administration, 2007, Table 5.2.1; Transportation Energy Data Book, Edition 27, Oak Ridge, Tenn.: Oak Ridge National Laboratory, 2008, Table 2.7; Annual Energy Outlook 2009, Washington, D.C.: U.S. Energy Information Administration, 2009, Tables 45, 60, 65, 67.

But shifting to VMT fees would have disadvantages as well. It would entail considerable cost to retro-fit the entire fleet of vehicles on the road. There are privacy concerns related to monitoring travel behavior. It is not apparent that instituting VMT fees or increasing them to keep pace with inflation would ignite less opposition than would an increase in fuel taxes. And administering VMT fees would almost certainly be more costly and complex than collecting fuel taxes.

Nonetheless, the motivations for shifting to VMT fees are strong. Current and projected revenue shortfalls are the most pressing concern. Figure 1 shows that since 1980, growth in vehicle-miles traveled on the nation’s roads has far outpaced growth in fuel consumption. The gap has kept widening, which means the funding has kept tightening. Moreover, the gap will widen further over the foreseeable future, based on the expectation that more fuel-efficient vehicles, along with alternative fuel options, will achieve greater market penetration in the years to come.

We examined what would happen to revenue if fuel taxes were replaced with VMT fees on an initially revenue-neutral basis beginning in 2015. As two bases of comparison, Figure 2 shows one trajectory with fuel taxes fixed at current levels and another with fuel taxes rising five cents per gallon in 2015. The other four lines, bunched together, show the projected revenues from four different VMT revenue streams: a flat fee of roughly 1.1 cents per mile (the amount that produces a revenue-neutral shift in 2015), an upper and lower bound that represent the range of revenue if the miles traveled were to grow 10 percent higher or lower than estimated, and a VMT fee structure in which passenger vehicles would pay slightly lower VMT fees per mile and trucks higher VMT fees per mile (0.8 cents per mile and 3.4 cents per mile, respectively).

Figure 2 — Only a Big Increase in Motor Fuel Taxes Could Compete with Revenues Based Instead on Miles Traveled

Only a Big Increase in Motor Fuel Taxes Could Compete with Revenues Based Instead on Miles Traveled

SOURCE: Computed by authors based on data from Annual Energy Outlook 2009, Washington, D.C.: U.S. Energy Information Agency, 2009, Tables 45, 60, 65, 67.

NOTE: VMT = vehicle-miles of travel.

The projections illustrate why, absent the five-cent increase in fuel taxes, VMT fees would be an attractive alternative revenue source. Even under the most conservative scenario in which the growth in road travel is 10 percent less than expected, VMT fees would still generate roughly 20 percent more revenue by 2030 than would the current rate of fixed fuel taxes. The reason is simple: As vehicle-miles of travel grow, VMT fees grow with them.

A national system of VMT fees would require widespread support from the states. In our inter-actions with transportation officials in states ranging from Texas to Vermont, we found that the states are interested in the potential of VMT fees to offset their declining revenue, and state officials are following current pilot programs using VMT fees with great interest. We also found that states would like the federal government to take the initiative. State officials believe that the federal government should exercise leadership in setting technical standards to prevent the development of multiple and potentially incompatible systems in different states and regions.

Construction workers await the arrival of President Barack Obama at a highway road project.


Construction workers await the arrival of President Barack Obama at a highway road project funded by the American Recovery and Reinvestment Act in Columbus, Ohio, on June 18, 2010.

Three Suggested Models

We used several criteria to identify the most promising options for implementing a national system of VMT fees. The system should be capable of metering mileage traveled across the entire national road network. A system with limited capabilities should come at low cost, whereas a higher-cost system should support more-sophisticated pricing policy options — that is, the ability to vary rates by time or location of travel. The system should allow for effective enforcement, require minimal administrative support from the states, and impose a minimal burden on drivers. Of the 11 system configurations we identified and assessed, 3 offer the greatest promise, providing alternative tradeoffs between cost and capabilities.

The first option is to meter mileage based on fuel consumption. Under this approach, all vehicles would be equipped with an identifier, likely a radio-frequency identification (RFID) tag embedded in the license plate or registration sticker. When a driver fills up at a gas station, electronic readers at the pump would detect the vehicle ID and determine its fuel-economy rating based on the make and model. The expected mileage would then be calculated from the number of gallons purchased, with a corresponding charge added to the fuel price, while fuel taxes (built into the retail price) would be subtracted. Vehicles not yet equipped with an identifier would pay the fuel taxes rather than the mileage charges.

Though offering limited flexibility, this option would likely be the least expensive, given the relatively low cost of RFID technology. It would also provide a fallback system — existing fuel taxes — to charge vehicles lacking their identifiers. This pay-at-the-pump model could also be used to collect fees for most vehicles if a transition to more-sophisticated metering equipment were pursued over the longer term.

States would like the federal government to take the initiative.

The second option is to meter mileage based on a device that would combine cellular service with a connection to the same onboard diagnostics port that has been required in all new vehicles since 1996 to support emissions checks. This approach would make it possible to determine the area of travel with rough accuracy and also to vary rates by vehicle characteristics, state or regional jurisdictions, and dense urban districts subject to congestion tolls. The location data would make it possible to accurately allocate mileage fees among jurisdictions. Fees could be collected through either a pay-at-the-pump model, wireless communication with a central billing agency, or a debit card system that would allow people with privacy concerns to insert prepaid debit cards into the onboard unit. While this technology configuration remains to be demonstrated in the context of road pricing, this option could provide a high level of metering flexibility at lower cost than the third option.

The third option is to meter mileage based on a device that would contain a GPS receiver. The only difference between this and the previous approach is that the onboard unit would rely on GPS, rather than cellular technology, to identify the area or even the specific route of travel. This technology has been demonstrated in prior U.S. trials and is currently being used in several road-pricing programs in other countries. If the price of the equipment can be reduced through large-scale production, and if current privacy concerns associated with using GPS can be overcome, then this option would be even more promising.

All three options face obstacles in terms of cost, administrative complexity, and political acceptability. Each option would be more expensive, possibly much more expensive, than collecting fuel taxes. Moving the point of collection from a small number of entities (fuel wholesalers) to a much larger number (retail fuel stations or individual motorists) would make it harder to detect and to prevent tax evasion. Administration of the system could require new tax collection channels, a new national agency or expanded state powers, cooperation from cellular providers and gas stations not currently involved in collecting fuel taxes, support from the Internal Revenue Service, national technology specifications and certification, and enabling or conforming state legislation. Even with all these, it would still be necessary to either index or periodically raise VMT fees to keep pace with inflation, and there is no indication that such increases would be politically easier to ratify than raising fuel taxes.

Drivers make their way through the five-way intersection of State Highway 22, State Highway 11, State Highway 18, State Highway 44, and U.S. Highway 129 in Gray, Georgia.


Drivers make their way through the five-way intersection of State Highway 22, State Highway 11, State Highway 18, State Highway 44, and U.S. Highway 129 in Gray, Georgia.

Dead End for Fuel Taxes

In considering the public acceptability of VMT fees, the experts we consulted offered two salient observations. First, there is little public understanding of the current challenges in transportation finance and thus of the motivations for a transition to VMT fees. Second, concerns that the government might begin to track personal travel behavior remain a potent obstacle to the acceptance of sophisticated in-vehicle metering equipment. To make for a smoother transition from fuel taxes to VMT fees, concerted public education and outreach will likely be imperative on the issues of finance and privacy.

As for finance, if the people and government of the United States want to continue to rely on user-based financing, then the time to end reliance on motor fuel taxes and introduce a new approach is rapidly nearing. With recent technology improvements, we are now capable of charging drivers directly for their use of the road network. Roads can be priced similarly to a utility by metering consumption and, if desired, charging more during periods of peak demand. Charging road users more precisely for particular trips on particular roads at particular times would make VMT fees fairer and more efficient than fuel taxes, and would provide a much more stable source of revenue as the country pursues greater fuel-efficiency and shifts away from petroleum-based fuels.

With respect to privacy, it is not apparent that metering road use would be any more threatening to privacy than using cell phones, but there is genuine concern that the government would surreptitiously track the travel of each citizen without his or her knowledge. Most technology and policy experts agree that VMT fee systems could be structured so that privacy would be maintained. Travel records could be stored in individual vehicles, not in a central repository, and could be erased once payments are made. It is also possible that many motorists would prefer to forgo privacy protection to have access to detailed bills showing each and every trip so that the charges could be audited and drivers could be sure they are paying for trips they actually made.

It is not apparent that metering road use would be any more threatening to privacy than using cell phones.

One of the most important arguments in favor of VMT fees is a subtle one. Relying on fuel taxes in an era of growing concern about fuel efficiency and greenhouse gas emissions creates an untenable contradiction between public policy goals. Any governmental policies aimed at reducing gasoline and diesel consumption to address environmental or energy security concerns would directly undermine the goal of raising sufficient revenue through fuel taxes to maintain and improve the nation’s vital road network.

If the country succeeds in encouraging the vast majority of truckers and motorists to drive plug-in hybrids, electric vehicles, fuel-cell vehicles, or even solar vehicles, it will still need to pay for road construction and maintenance. Drivers should still be responsible for such costs, even if they drive nonpolluting vehicles. The nation should put an end to fuel taxes and other such programs that discourage government pursuit of energy efficiency over the long term for fear of losing transportation revenue in the short term. Quite simply, it will not be possible to rely on the gas tax as a road user fee when cars are no longer powered by gas.

Congress in the Driver’s Seat

Innovation is afoot. Several countries in Europe have already developed automated systems to charge trucks for distance traveled on major highways, and New Zealand levies distance-based road use charges on diesel-powered trucks and passenger vehicles alike.

In the United States, Oregon and the Puget Sound Regional Council in the Seattle area have conducted trials demonstrating the feasibility of GPS metering equipment for assessing VMT fees, and the University of Iowa has recently completed a dozen similar trials in other parts of the country. The results are quite encouraging, but questions remain. It would be worthwhile to invest in additional trials to evaluate technical options and components, to explore the cost and reliability of alternative collection and enforcement mechanisms, and to expand the experiments to more participants and more states.

As it debates the new transportation bill, Congress should consider ways to prepare for the adoption of direct user fees. Alternatively, Congress could reject such a transition and raise fuel taxes to provide needed revenue, though the revenue issue would then need to be addressed again in five or ten years as the adoption of more fuel-efficient conventional vehicles as well as alternative-fuel vehicles accelerates. Or, continuing in the vein of the recent stimulus bills, Congress could allow the nation to begin paying a greater share of its road and transit bills from general revenues, as most other countries do; however, given the extent of the federal budget deficit, along with competing demands for general revenue, such a course would almost certainly result in a considerable reduction in the size of the federal transportation program.

It is time for Congress to act and for citizens to ensure that it does.

The hope in many quarters is that Congress will seize the opportunity to begin building the architecture of a national system of direct user charges. The early deliberations could address how the system would be configured, what types of pricing policies it would need to support, the appropriate institutional roles for various participants from the public and private sectors, and a plan for making the transition from fuel taxes to VMT fees. Beyond expanding the current trials to demonstrate technology options on a larger scale, Congress might also put an early system of VMT fees into actual practice on a voluntary or limited basis.

To bolster the prospects of implementing a national VMT fee system by 2015, Congress should fund a set of preparatory activities that encompass planning, research, technology development, expanded trials, and education and outreach. Such a massive undertaking would likely require a designated governmental entity, granted the requisite level of authority, to shepherd these efforts.

It is time for Congress to act and for citizens to ensure that it does. The forthcoming debate will indicate whether the nation’s system of governance has the ability to make complex technological choices that can simultaneously improve the efficiency, equity, and environmental performance of surface transportation in this country. square

Related Reading

“After the Motor Fuel Tax: Reshaping Transportation Financing,” Issues in Science and Technology, Vol. 25, No. 4, Summer 2009, pp. 85–88, Martin Wachs. Also available as RAND/EP-20090631.

Implementable Strategies for Shifting to Direct Usage-Based Charges for Transportation Funding, National Academy of Sciences, Transportation Research Board, National Cooperative Highway Research Program, Web-only document 143, 2009, 150 pp., Paul Sorensen, Liisa Ecola, Martin Wachs, Max Donath, Lee Munnich, Betty Serian. Also available as RAND/RP-1395.

Share your thoughts on Facebook.