Revisiting the Fuel Tax–Based Transportation Funding System in the United States

Demand-side responses

The inability of the fuel tax system to price congestion is one of the contributing factors in our nation’s growing congestion problem. Annual delay per auto commuter rose from 19 hr in 1982 to 52 hr in 2014 in the nation’s 15 largest urban areas (Texas Transportation Institute, 2012). This growing rate of congestion imposes real costs to users and the economy as a whole. Using value of travel time (estimated at US$16.79 per hour of person travel and US$86.81 per hour of truck time) and excess fuel consumption (estimated using state average cost per gallon), the Texas Transportation Institute (2012) estimated the total cost of congestion in the nation’s urban areas for 2011 at approximately US$121 billion. In addition to individual users, businesses are also hurt by congestion. In part because of rising congestion, the logistic costs for American businesses rose to 10% of GDP in 2006 (National Surface Transportation Policy and Revenue Study Commission, 2007).

In addition, fuel tax serves as a blunt instrument to deal with congestion problem. The congestion occurs when the number of vehicles on the road increases and slows down the traffic (Parry et al., 2007). “Clearly a fuel tax, which raises driving costs for all regions at all times of day, is a very blunt instrument for alleviating traffic congestion, which is highly specific to rush periods in urban areas” (Parry et al., 2007, p. 379).

As the fuel tax rates are lower than the marginal costs, it sends poor price signals to users regarding the trade-off between using autos and trucks versus using transit or freight rail. Lewis (2008) noted that it should come as no surprise that while our roadways are congested, there is little use of public transit. If fuel taxes reflected true costs, including congestion and environmental costs, it is likely that, over time, some users would shift to transit and freight rail. As Wachs (2003) pointed out, raising the gas tax sends price signals to motorists to use the transportation system more efficiently as it encourages motorists to switch to public transit to save money. Furthermore, while our highway system is experiencing growing levels of congestion, our freight rail system remains relatively uncongested. In 2006, 88% of the freight rail corridors were operating at below practical capacity (National Surface Transportation Policy and Revenue Study Commission, 2007). It should be noted, however, that transit, like the highway system, is underpriced, and thus, transit costs are not fully covered either.

In addition to sending price signals that lead users to shift away from peak period travel, and auto and truck trips to shift to transit and freight rail, respectively, a properly set tax could help reduce the number of less-than-full truckload and empty trips. Robinson (2008) indicated that the introduction of heavy vehicle tolling in Germany, which priced trucks per-kilometer of use of the Autobahn, more closely reflect full internal costs as these tolls are in addition to fuel taxes. Tolling of trucks resulted in about a 20% reduction in the number of empty truck trips. The pricing of trucks was based on maximum rated gross vehicle weight, which meant that empty trucks were charged the same rate per kilometer as fully loaded trucks. This gave rise to load consolidation brokers and led to greater efficiency in truck operation.

Once again, the inability of the fuel tax system to price roads to reflect the true costs imposed by users leads to inefficiencies. Long-distance commuters do not have to pay their full cost of travel, which is a built-in incentive to locate to more distant locations to take advantage of lower home prices and higher acreage, which, in turn, contributes to urban sprawl. According to the national census, from 1970 to 2000, central city density declined by approximately 35% (Langer & Winston, 2008). The trend remains in recent years. A census report of pattern of population change between 2000 and 2010 indicates, “people in larger metro areas often lived farther from city hall” (Wilson, Plane, Mackun, Fischetti, & Goworowska, 2012). It should come as no surprise that during this period, the ability of the fuel tax system to recover the full costs imposed by drivers was significantly eroded.

Considering the fuel tax as a Pigouvian tax to correct the negative externality of pollution, fuel taxes have some effect in reducing petroleum-based fuel consumption and greenhouse gas emissions. One could argue that the polluters-pay principle is in effect as drivers with lower fuel efficiency vehicles do pay more in fuel taxes per vehicle mile. Also, it is possible that increases in fuel taxes may encourage some motorists to purchase more fuel-efficient vehicles and tune their vehicles to get better gas mileage (Wachs, 2003). Hsing (1994) found that the 4.3 cents per gallon (1.1 cents per L) increase was expected to reduce gasoline consumption by 2.0%, which amounted to 53.4 million barrels of oil per year.

While, in the short run, fuel tax increases are likely to have modest effects on fuel consumption, the price elasticity of fuel in the long run is much more significant (Sterner, 2007). Analysts have suggested that the price elasticity for fuel in the United States in the short run is around −0.18, whereas in the long run, it is closer to −1.0 (Sterner et al., 1992). The meta-analysis using Seemingly Unrelated Regression (SUR) approach by Brons, Nijkamp, Pels, and Rietveld (2008) shows that the elasticity of gasoline demand is −0.34 in the short run and −0.84 in the long run. They decompose it into the price elasticity of fuel efficiency, mileage, and car ownership. The estimation of the decomposition shows that price change response is more to fuel efficiency and mileage driven than car ownership. Sterner (2007) estimates that if, for a long period of time, the United States would have applied the tax policy of the European nations with the highest tax on fuel, then U.S. gasoline consumption would have been reduced by approximately 57%. Furthermore, the reduction in CO₂, if all the Organisation for Economic Co-Operation and Development (OECD) countries applied the highest fuel tax rates found in Europe, is estimated to be 8.5 billion tons over a decade (Sterner, 2007).

The fuel tax system promotes less-polluting fuels insofar as it raises the price of petroleum-based fuels, thus creating an incentive to develop and use alternative fuels. The Transportation Research Board (2006) noted, “[S]ubsidies in the form of waivers of excise taxes have been a popular way to promote alternative energy development (e.g., the fuel tax subsidy granted to gasohol)” (p. 14). While one could argue over the degree to which gasohol is a less-polluting fuel than gasoline, the argument that fuel taxes promote the use of less-polluting fuels could be extended to other fuels. As long as there is a tax on petroleum-based fuel, there will be an incentive to develop and use other forms of fuel.

Supply-side responses

Small et al. (1989) explained that, in addition to leading to a more efficient use of transportation capacity by users, an advantage of marginal-cost user charges is that “[T]he resulting revenues provide a tangible signal to public officials as to whether additional investments to provide more or better services are likely to be worthwhile” (p. 9). Thus, by properly pricing the usage of roads during the peak time, public officials and investors are able to get a more accurate picture of the demand for more or better service.

As indicated previously, the fuel tax system does a poor job of accurately pricing the costs of highway use and leads to inefficient overuse of the capacity. Because of this, it also fails to convey price signals to help direct investment to the most worthwhile projects. Underpayment may result in underinvestment in transportation, as the fuel tax system does not raise sufficient funds for economically justifiable projects. Underpayment by users, and the resulting unconstrained demand, may also lead to overproduction in some instances. As stated by Lewis (2008), “Because the absence of congestion pricing encourages peak period demand that would not otherwise arise, the need for highway investment is increased accordingly . . . the federal taxpayer should not be burdened by investment costs that are not economically justified” (p. 30). Thus, lack of congestion pricing can lead not only to excessive congestion but also to excessive highway investment. A study by Winston and Langer (2006) estimated that US$1 of government spending on highways reduced road users’ congestion costs by only 11 cents. From this finding, the study went on to estimate that states would have to spend nearly US$350 billion annually to eliminate congestion costs. These costs, including the need for some expensive road expansions, could largely be reduced if the demand during peak-use periods were reduced.

A study by Boarnet (1997), which estimated county-level production functions, found that while there was strong evidence that congestion reduction can affect county output, there was weaker evidence that street and highway capital stock increases were productive. In their findings and recommendations, the National Surface Transportation Infrastructure Financing Commission (2009) concluded that greater use of pricing mechanisms such as a mileage-based pricing may lead to more efficient investment and reduce the need for additional capacity that may otherwise be built, by shifting demand to off-peak hours and to other transportation modes.

Benefit received

In this section, we begin our analysis with an evaluation of how closely motor fuel taxes, as user fees, adhere to the benefit-received principle. It is important that we remind ourselves what we mean by user fees and benefit-received principle. These are fees collected from those who use a particular service, as opposed to fees collected from the public-at-large. User fees apply to activities that generally provide special benefits to identifiable recipients or beneficiaries, and fees generally vary in proportion to degree of use. The benefit-received principle applies to the quasi-market arrangement that only individuals who receive benefit from a public service pays for the service and the payment share should equal to the benefit share. It is costly to build and maintain the transportation system. The fuel taxes are the payments for using the transportation system. The discussion of whether the fuel taxes adhere to the benefit-received principle from the perspective whether the marginal burden of fuel tax payments aligns with the marginal cost of maintaining the system.

In recent years, the connection between the amount of driving on the roadway and the amount of fuel consumed (and hence fuel taxes paid) has been weakened by the increasing fuel efficiency and the introduction of alternative fuel vehicles. Increasing fuel efficiencies have allowed users with more fuel-efficient vehicles to pay less in taxes per VMT. Furthermore, the introduction of alternative fuel vehicles makes it possible for some users to pay very little or no fuel taxes. The Environmental Protection Agency (2013) reported that 3.1% of the model year 2012 fleet are hybrids. Although plug-in hybrids and electrical vehicles still make up a small minority of all drivers, they nonetheless represent a significant violation of the user-pays-and-benefits principle as a substantial part of their propulsion is powered by electricity, and thus not subject to fuel taxes.

Besides the distance traveled on the roadway, the weight of the vehicle also contributes to the roadway maintenance cost. Wachs (2003) suggested that fuel taxes fare relatively well in recovering maintenance costs: Heavier vehicles pay more in fuel taxes because they are less fuel efficient. However, Small et al. (1989) noted that structural damage to roads is caused mostly by trucks and buses; it is not total weight but weight per axle that is important. (It should be added, however, that total weight is relevant when assessing structural damage to bridges.)

According to the Federal Highway Administration’s (1997) Cost Allocation Study, updated in 2000, the equity ratio (ratio of tax and fee payments by type of vehicle and highway costs imposed by those vehicles) is 0.53 for heavier single-unit trucks (over 25,000 pounds [11,340 kg]) and 0.48 for the heaviest combination trucks (over 80,000 pounds [36,287 kg]). Even for the bread-and-butter five-axle combination units (50,000-80,000 pounds [22,680-36,287 kg]) the equity ratio is 0.83 (Federal Highway Administration, 2000). This means that, while it is true that heavy trucks pay higher transportation taxes and fees than light vehicles, they, in turn, impose substantially higher costs than are recovered through their payments.

The benefit-received principle for fuel taxes is further eroded by the extent that groups of users do not pay their fair share of fuel taxes. This comes in the form of exemptions and outright evasion. Federal law exempts users such as state government, non-profit educational organizations, and emergency vehicles from having to pay fuel taxes (National Surface Transportation Infrastructure Financing Commission, 2009). Subsidies were also provided for users of gasohol until 2012 (National Research Council, 2007). ² In terms of evasion, there are several forms including bootlegging across state lines, diluting the blend, and “daisy chains” (creating a dummy corporation and a fraudulent and complex trail of paperwork; Denison et al., 2000). The National Surface Transportation Infrastructure Financing Commission (2009) noted that evasion remains a problem, even though progress has been made in recent years through legislative changes and increased enforcement. Total fuel tax evasion at the state and federal level may exceed US$1 billion annually or 3.5% of total federal motor fuel tax revenue.

Ability to pay

We conclude our discussion of social equity with an analysis of users’ ability to pay (vertical equity). Vertical equity can be defined as, “Distributing tax burdens fairly across people with different abilities to pay” (Rosen, 2005, p. 576). It is important to note that there is often a trade-off between the benefit-received principle and vertical equity. As mentioned before, since 1983, a portion of federal fuel tax revenues have been spent on mass transit projects. Currently, 2.86 cents of the 18.4 cents federal gas tax goes to the mass transit account. In 2004, highway user fees collected at all levels of government, of which fuel taxes are the main component, transferred US$11 billion to transit projects, which amounted to approximately 10% of total user fee revenue collected at all levels of government (Transportation Research Board, 2006). Thus, individuals who cannot afford to drive, or for other reasons elect not to drive, are still granted access and mobility through mass transit that is funded, in part, by fuel taxes.

Although fuel taxes may aid in providing mobility to disadvantaged groups, fuel taxes, like most consumption taxes, are regressive. The National Surface Transportation Infrastructure Financing Commission (2009) concluded that fuel taxes were “highly regressive,” and more regressive than a general sales tax. The Commission also indicates that higher income groups are more likely to shift to more fuel-efficient vehicles, and thus pay less in fuel taxes. Whitty and Svadlenak (2009) built off the Commission’s assertion that higher income groups are more likely to buy fuel-efficient vehicles and state that, as new vehicles become more fuel efficient, the stratification between what owners of old vehicles and new vehicles pay may become greater without a justification based on road use. Backing the Commission’s claim that fuel taxes are highly regressive are estimates found in a Tax Foundation report by Williams (2007). While Americans earning less than US$40,000 pay between 0.8% and 1.6% of their income in gas taxes, those earning over US$100,000 pay between 0.25% and 0.45% of theirs (Williams, 2007).

Revenue adequacy

As previously stated, the Highway Trust Fund has necessitated billions of dollars in transfers from the federal general fund to keep it solvent (Puro, 2013). These infusions, while significant, only tell part of the story. The National Surface Transportation Infrastructure Financing Commission (2009) provided estimates of average annual capital needs and gap from 2008 to 2035 and stated,

Without changes to current policy, it is estimated that revenues raised by all levels of government for capital investment will total only about one-third of the roughly $200 billion necessary each year to maintain and improve the nation’s highways and transit systems. (p. 3)

It should come as no surprise to many Americans that we have large gaps in financing our transportation system, despite our relatively large tax base. In 2007, there were approximately 205 million licensed drivers in the United States, which represented approximately 68% of the population (Federal Highway Administration, 2009). While this is a sizable tax base, current fuel tax rates are below the rate needed to raise sufficient revenue. To get an idea of the United States’s relatively low fuel tax rates, it is useful to compare our rates with those of other industrialized nations. While Americans currently pay on average US$0.40 per gallon in federal, state, and local fuel taxes, Japanese pay approximately US$2.47 per gallon and the Dutch pay US$5.42 per gallon. ³

Revenue sustainability

Not only is the fuel tax system inadequate for generating sufficient revenue but it is also unsustainable in the long run. A proper user fee should ensure that revenues are self-sustaining and predictable regardless of changes in external factors. Inflation, changes in the price of fuel and in fuel efficiency, and the introduction of alternative fuels, all lead to revenue sustainability problems. The National Surface Transportation Infrastructure Financing Commission (2009) noted,

The current . . . funding structure that relies primarily on taxes imposed on petroleum-derived vehicle fuels is not sustainable in the long term and is likely to erode more quickly than previously thought. This is due in large measure to a drive for greater fuel efficiency, alternative fuels, and new vehicle technology. (p. 7)

In 1975, the average fleet fuel economy was 13.1 miles per gallon (mpg; 5.6 km/L); by 2013, it was 24 mpg (10.2 km/L), an 83% improvement. This significant improvement has caused users on the whole to pay less per VMT (Environmental Protection Agency, 2013). The increase in fuel efficiency is a large reason why today’s overall user fee, taking into account taxes and fees at all level of government, estimated at 3 cents per mile (1.9 cents per km; National Surface Transportation Infrastructure Financing Commission, 2007), is well below the six cents per mile (3.8 cents per km) user fee of the 1960s adjusted for inflation (Morris, 2006).

The trend of increasing fuel efficiency for the overall fleet is expected to continue as average fuel economy for new vehicles accelerates. The Energy Information Administration (2014) estimated that average fuel efficiency for all light-duty vehicles on the road will grow from 21.5 mpg (9.1 km/L) in 2012 to 37.2 mpg (15.8 km/L) by 2040. A report of the National Surface Transportation Infrastructure Financing Commission (2009) found that the current federal surface transportation funding structure, which relies heavily on motor fuel taxes, is not sustainable and is likely to erode more quickly than previously thought due, in large part, to a drive for greater fuel efficiency caused both by heightened concerns over global climate change and by an effort to reduce dependence on foreign energy sources. ³ The commission went on to state that fuel taxes and other user fees account for less than 60% of total transportation system revenue (federal, state, and local), which clearly shows that users do not bear the full cost of their travel.

The fuel tax system fails to perform as an adequate and sustainable revenue source, in part, because it is not indexed to inflation. Although the federal gas tax has been raised many times since its inception, it has not kept pace with inflation. In fact, although the federal gas tax has more than doubled since 1983, its purchasing power remains at approximately 1983 levels (National Surface Transportation Policy and Revenue Study Commission, 2007). This Commission suggests that since 1993, the federal gas tax rate has decreased by 40% when compared with the Producer Price Index for Highway and Street Construction. The corrosive effect of inflation on revenue could be eliminated by indexing fuel taxes to inflation. In its findings and conclusions, the National Surface Transportation Policy and Revenue Commission (2007) noted that a limitation of the fuel tax is that it is not currently able to react to increasing construction costs and adds that this could be remedied by indexing it to the Consumer Price Index or a targeted measure such as the Producer Price Index for Highway and Street Construction.

Small et al. (1989) stated, “The fuel tax has not always proven a reliable source of revenue; since 1973 receipts have fluctuated along with economic conditions and fuel prices” (p. 6). Most notably, fuel tax revenue dwindled during the 1970s, in part, due to the energy crisis (Patashnik, 2000). While price spikes tend to cause a reduction in the amount of fuel consumed and thus fuel tax revenue raised, they also lead to a drive for more fuel-efficient vehicles, thus further reducing the sustainability of fuel taxes. Some analysts have estimated that the average gallons of fuel consumed per vehicle mile by the light-duty fleet could fall by 20% if new regulations or large and sustained fuel price increases drive fuel economy improvements (Transportation Research Board, 2006).

The Environmental Protection Agency (2013) noted that fuel efficiencies have been improving since 2005. Several analysts expect this trend to continue and accelerate. Although this is a positive trend for the environment, it also means lower fuel tax revenues. As users consume less fuel, they pay less in fuel taxes per VMT. Between 1980 and the present, VMT has grown by approximately 100% whereas fuel consumption has grown by about 50% (Sorensen et al., 2009). This is bad news for transportation funding: As demand for road capacity increases and as road damage grows with VMT, revenues needed to address capacity and maintenance needs continuously decline.

Political feasibility

Because the fuel tax system already is in place, we focus not on whether it is politically feasible to implement fuel taxes but rather on the feasibility of future fuel tax increases. As described in the revenue sustainability section, because of fuel efficiency improvements and the effects of inflation, the real value of revenue collected from fuel taxes has been declining. Thus, if we want to maintain and improve the transportation system, it is important to consider whether it is politically feasible to raise fuel taxes. In the past 4 years, only six states increased their fuel tax rate. The greatest increase was in Wyoming, which raised its gasoline excise tax rate from 13 cents per gallon to 23 cents in 2013. The increases in other states were modest, which were on average less than 5 cents per gallon. ⁴ Small et al. (1989) noted, “(D)espite the occasionally severe erosion of real revenues, states have found it politically difficult to raise gasoline taxes” (p. 6). Their assertion that the states have found it politically difficult to raise gasoline taxes can be extended to the federal level, where taxes on gasoline and diesel have not been increased since 1993, despite shortfalls in highway funding. Wachs (2003) asked, “Why is it assumed to be a political liability to raise fuel taxes by a few pennies when fuel prices routinely change by more than that several times every year” (p. 237).

Fuel taxes initially had great public support. In discussing the popularity of state gasoline taxes in the 1930s, Williams (2007) stated, “Gasoline taxes met with little public resistance and in fact became quite popular with the general public. Citizens saw the benefit principle in action, as gas taxes served mostly as user fees, generating revenue for more and better roads” (p. 4). As we documented in the equity section, however, fuel taxes have since moved away from the benefit-received principle and, with this shift, their popularity has dwindled. The 2009 Harris Interactive (2009) poll found that at the state and local level, respondents found the gas tax to be the least fair tax when compared with state income taxes, retail sales taxes, motor vehicle taxes, local property taxes, and taxes on cigarettes, beer, and wine, with only 7% of respondents finding the gas tax “very fair.” At the federal level, the only tax respondents found more unfair than the gas tax was the estate tax, with only 7% of respondents finding the federal gas tax “very fair.” It should be noted, other taxes were also not seen as “very fair,” and taxes of any kind are generally unpopular.

This weakening of public opinion for fuel taxes has coincided with less political support as well. As stated by Sorensen et al. (2009),

With rising anti-tax sentiment among the populace, elected officials have become wary of this politically unpopular task, and the frequency and magnitude of the recent fuel tax increases has been grossly insufficient to maintain comparable purchasing power in terms of real revenue per mile of travel. (p. 2)

As noted before, the federal tax on gasoline and diesel has not been significantly raised since 1993. Thus, it would appear that future attempts to raise fuel taxes may be politically difficult, especially if these increases do not adhere to the user-pays-and-benefits principle.

Political feasibility is typically evaluated by also looking at the tax visibility to taxpayers, the potential for tax exportation, and general political support and public opinion. In our analysis, we also include driver’s privacy and system security, as these issues have been raised when comparing fuel taxes with transportation funding alternatives such as distance-based fees.

Broadly speaking, tax visibility can be defined as the extent to which users are aware of a tax. The popularity of a tax with the public tends to decrease as the tax becomes more visible. When the visibility of a tax is low, it may be tempting for government to increase the tax. Visibility should not be confused with transparency: the extent to which taxpayers know the actual costs they incur as a result of the tax. The National Surface Transportation Infrastructure Financing Commission (2009) noted, “system users are typically unaware of how much they pay in fuel taxes (as distinct from the price of gasoline), such that daily swings in price mask the tax component” (p. 7). As mentioned before, a study prepared for Minnesota Department of Transportation by Fichtner and Riggleman (2007) found that, in general, participants were unaware of what they paid in gas tax. Thus, it would appear that fuel taxes are visible to the extent that users know that they are paying them; however, fuel taxes are not transparent and most users do not know precisely how much they pay. How this translates into the political feasibility of raising fuel taxes is somewhat unclear.

Tax exportation pertains to the potential for nonresidents to pay a tax. Tax exportation increases the popularity of a tax for residents, as residents are able to benefit from revenues derived from taxes paid by others. For the most part, fuel taxes in the United States do not benefit from tax exportation. While foreign tourists and shippers traveling in the United States by auto or truck do pay fuel taxes when they refuel, this is an insignificant portion of total federal fuel tax revenues.

For a user fee system to be politically feasible, the system must ensure drivers’ privacy and provide system security. Ensuring drivers’ privacy entails taking measures so that information collected, kept, and disseminated is limited to the level needed to administer the user fee system and ensure data confidentially. Ensuring system security entails making sure the system is designed with security features to protect it from unauthorized access and improper or illegal use. Little has been written on driver’s privacy and system security as it pertains to the fuel tax system. This is likely because the simplicity of the fuel tax system minimizes any sort of risk to drivers’ privacy or overall system’s security. Furthermore, drivers are not the legal payers of the tax. Fuel taxes are built into the price that drivers pay at the pump and, thus, unlike VMT fees, do not involve the need for any device being installed in vehicles or for driver data to be collected, stored, and sent to remote billing locations.

Administrative feasibility

Administrative feasibility includes the costs of implementation, operation, enforcement, and compliance costs. The fuel tax system has relatively low implementation costs. In listing the attributes that make fuel taxes attractive sources for funding transportation, the National Surface Transportation Policy and Revenue Study Commission (2007) noted, “low administrative and compliance costs” as well as “ease of implementation.”

The fuel tax system also has relatively low operation and enforcement costs. Martin Wachs (2003) notes fuel taxes’ low collection costs and that they are relatively fraud proof, as one of his 12 reasons for why gasoline taxes should be raised. Wachs stated,

Governments have a responsibility to be concerned about the cost of collecting revenues and . . . evasion . . . the fuel tax is unusually efficient in this regard. Whereas traditional manual toll collection, for example, incurs costs that range from 20% to 25% of the revenue produced, the cost of administering the fuel tax is typically only 1% or 2% of the revenue. (p. 239)

Others have estimated the cost to administer and enforce the federal motor fuel taxes to be as low as 0.2% of gross receipts (Peters & Kramer, 2003).

Finally, when evaluating the administrative feasibility of a tax, we consider how difficult it is for the public to pay the tax. Unlike the federal income tax, which costs non-business taxpayers on average 10.7 hr of time and US$129 per year in record keeping, tax planning, form-completion and submission, and other activities (Internal Revenue Service, 2009), fuel taxes have very low compliance costs for the general public. As fuel taxes are collected from wholesalers and passed on to drivers when they refuel, there is no need for the general driving public to keep records of the fuel taxes they pay or complete and submit any forms documenting their payments (unless it concerns documenting work-related mileage for tax purposes).