The Heritage Foundation

Backgrounder #1650 on Smart Growth

May 2, 2003

May 2, 2003 | Backgrounder on Smart Growth

Tolls and Surface Transportation Reauthorization

Ending obsolete restrictions on tolls is a sensible compromise between congressional road-building "hawks" and the Bush Administration. Led by House Transportation Committee Chairman Don Young (R-AK), the hawks have been impressed by testimony on the extent to which road building has lagged across the nation. Seeing an impending "congestion crisis," they want to hike the gas tax by over 50 percent to fund a 60 percent expansion of road and transit funding over the next six years. The Administration, on the other hand, flatly rejects any gas tax increase, but its status quo stance offers little prospect of relief from worsening congestion.

Both sides are overlooking an alternative--easing half-century-old federal restrictions on tolling--that could enable the road builders to get much of what they want via toll roads while also accommodating the Administration's firm stand against higher taxes.

This alternative would require repeal of the federal restrictions on tolling, which date from the 1950s when tolls meant delays from queuing at toll plazas. At the time, the gas tax seemed to offer almost unlimited, easily collected funds for roads. Since the 1990s, technology has made it possible to identify motor vehicles by using advanced radio and video systems, ending the need for on-site toll collection.

As a user-pays system, tolls also offer a way to measure demand for new highway investment. If tolls are set flexibly, they indicate motorists' willingness to pay for the cost of the service, and variable tolls can be used to optimize traffic flow and prevent breakdown into inefficient stop-and-go driving.


A major political battle looms over the reauthorization of federal funding for surface transportation for the next six years. The Bush Administration has proposed no increase in the gasoline tax of 18.4 cents per gallon or the diesel fuel tax of 22.4 cents per gallon, both of which fund the program via the highway trust fund. The Administration has proposed a $41.2 billion per year program, a 13 percent increase over the previous six years' $36.3 billion per year. However, House Transportation Committee leaders have proposed a massive $62.5 billion per year federal aid program for road and rail based on an immediate 5.4 cents (29 percent) increase in gasoline and diesel fuel taxes as well as subsequent annual increases based on the Consumer Price Indexes.

These proposals (especially the big gas tax increase plan proposed by the committee leaders) could be pre-bargaining positions, and the final reauthorization number will likely be some kind of compromise. However, inside-the-Beltway skirmishes over the size of federal aid and the federal gas tax could distract attention from better--and vitally needed--methods of improving the transport system.

Besides generating insufficient revenue, fuel taxes have a doubtful future as the principal source of funding for roads. Higher gas taxes face strong political opposition, and technology is steadily eroding the revenue yield of fuel taxes. The internal combustion engine is being made more efficient and able to use less fuel per mile traveled, while new power plants (hybrid gas-electric) and new fuels (ethanol, natural gas, and eventually hydrogen) complicate taxation.

In recent years, the American public has repeatedly resisted gas tax increases at the ballot box. They do not believe they get value for their gas tax money because they do not see freer-flowing roads resulting from higher taxes. A considerable proportion of federal gas tax revenues is siphoned off into unproductive rail transit projects, used to fund state and federal bureaucracies, and simply spent on political "pork." After maintenance, operations, transit, and "administration," little is left for expanding capacity and alleviating congestion. Traffic has grown at approximately the same rate as the economy, but total lane-miles of roadway has lagged significantly behind. The inevitable result has been busier roads and longer congestion-related delays.

Furthermore, the gas tax is a flawed price mechanism. It costs the motorist approximately the same in gas taxes to drive during peak periods, when road space is very scarce and expensive, as it does during non-congested periods. The U.S. Department of Transportation (USDOT) estimates that the typical per-trip cost of supplying a road trip in peak hours is 30 cents per mile.1 This is 15 times the typical amount of fuel tax collected. Marginal costs of uncongested off-peak and rural road trips, by contrast, are often being overcharged with the gas tax, which works out to 2 cents per mile.

Pavement and bridge costs are also poorly priced with the diesel fuel tax because fuel consumption, and hence tax paid, increases less than proportionately with truck weight, while stresses on roads and bridges increase disproportionately with weight. (Cars do negligible damage to road pavement and do not stress bridges.)


The nation certainly needs a more effective highway system, including new highways and extra lanes on existing highways. Congestion has been getting worse under the current federal aid program and policies.

The USDOT summarized the situation in its most recent comprehensive report to Congress:2

  • In 2000, trips requiring 20 minutes in free-flow traffic took 30.2 minutes during average peak periods, whereas the same trip took only 25.8 minutes in 1987. Therefore, the congestion "penalty" was 10.2 minutes in 2000 compared to 5.8 minutes in 1987, for an increase of 76 percent in 13 years.
  • The percentage of "penalty" time from congestion rose from 56 percent to 78 percent in urban areas with populations over 3 million, from 23 percent to 53 percent in cities between 2 million and 3 million, from 13 percent to 32 percent in cities between 500,000 and 2 million, and from 5 percent to 21 percent in cities of less than 500,000.
  • The average length of "rush hour" (morning and afternoon combined) increased from 5 hours to 5.3 hours or by 7 percent.
  • In larger cities, 40 percent of daily traffic is now under congested conditions, compared to 35 percent in 1987.

The mismatch between roadway capacity and traffic demand is clearly serious and growing. Congestion also represents a growing economic burden, currently estimated at $100 billion.3 Business as usual is no longer an acceptable option.

The USDOT estimates that maintaining the highway system at its present performance level will require $76 billion per year in capital spending, compared to actual spending of $65 billion in 2000. To improve performance--reduce congestion, improve safety, and enhance productivity--will cost an estimated $107 billion per year. These estimates translate into an 11 percent increase over projected spending simply to "maintain" the existing system and a 57 percent increase to "improve" it.4

The current funding structure has other problems. However, it is not clear that simply increasing federal gas-tax funded programs will produce increased highway capital spending. Of the $65 billion in highway capital expenditures in 2000, the federal government provided just $26 billion. The remaining 60 percent came from tolls and state and local taxes.5 States and localities have considerable flexibility in how they spend their own money and may respond to increased federal grants with reduced spending from their own resources.

Similarly, increased federal grants may sometimes be at the expense of--rather than in addition to--toll-financed projects. For example, the notorious Big Dig (I-93) in Boston and the Wilson bridge reconstruction on the Washington Beltway together are absorbing $17 billion in federal and state tax money. Both megaprojects were eminently suitable for toll financing, so the $17 billion could have been used elsewhere for road projects of greater productivity. Moreover, toll financing of these facilities would likely have encouraged some financial discipline and curbed the tendency to engage in "pork-barrel" spending.


Given the weaknesses of financing surface transportation through fuel taxes, there is a strong case for encouraging movement to direct-use charging--tolls. Traditionally, the problem with tolls has been collection. The toll plaza, with its multiple stopping lanes and toll booths manned by collectors, is an expensive operation for the road operator and a nuisance to motorists. However, the advent of the windshield-mounted transponder in the 1990s allows automatic toll collection by radio signal without requiring vehicles to slow or stop.

Along with other advanced technologies to classify vehicles and record their license plate numbers, toll collection is being reorganized as an economical and hassle-free process. Over half the tolls in America are now collected electronically, and the days of cash toll collection are numbered, with most new toll roads incorporating automated tolling and old toll roads gradually being retrofitted.

Tolls have numerous advantages:

  • Tolling is fair--the user pays.
  • Tolling is a flexible pricing method that enables charges to be tuned to cover costs and rationally allocate scarce resources.
  • Tolling, when applied flexibly, can efficiently manage lane loading to enable the free flow of traffic.
  • Tolling is generally less unpopular than new or higher taxes. A majority of people (50 percent to 80 percent) accept the fairness of the user-pays tolling and see a direct connection between a new road facility and the toll--a connection often hidden with taxes.
  • Tolling can raise funds independently of national and state budgets, allowing urgent roadwork even under tight budgets.
  • Tolling can be conducted by state toll authorities, state DOTs, counties, cities, special-purpose multi-jurisdiction agencies, nonprofits, or for-profit investors--it is organizationally flexible.
  • Technology has overcome the worst aspect of tolling--the inefficiency of stopping to pay and the irritating backups at toll plazas.


The lingering bias against tolling in important sections of federal highway law dates back to the beginning of the interstate highway system in 1956. The major proposition in 1956 was to fully finance the interstate highway system using fuel taxes and make the entire system toll-free. It was claimed that an increase in the gasoline-diesel tax of 1 cent per gallon would finance construction of 30,000 miles of new toll-free interstates and pay off the debts and de-toll the 2,500 miles of toll roads already within the interstate system.

However, the 1956 Federal Aid Highway Act, while providing for the retirement of toll-road debt and de-tolling, did not establish a schedule. No money was ever appropriated, and there was no de-tolling under the federal act, although there have been isolated de-tollings at purely state initiative, as in Connecticut.

There are now no serious proposals to de-toll any of the toll segments of the interstate highway system. The provisions of the 1956 legislation have long since been forgotten. The fuel tax is struggling to produce the revenues needed for maintenance and improvements of the current system.

It is time to move on. It makes sense to reshape the law to accept tolling as a full and legitimate method of financing roads, perhaps with the details to be decided at the state or local level of government.

However, U.S. law continues to obstruct toll financing. One section of the U.S. Code is even titled "Freedom from tolls,"6 as if tolls were a kind of affliction or oppression instead of the price paid for the use of a scarce commodity: road space.

Tolls have an honorable history, financing the Pennsylvania Turnpike (the country's first great cross-state expressway) and many other great national roads such as the New Jersey Turnpike; the Garden State Parkway; the Tollways of the Chicago area; the New York State Thruway; major highways of Massachusetts, Maine, New Hampshire, Delaware, Oklahoma, and Kansas; and the great trucking route between Philadelphia and the Midwest through Pennsylvania, Ohio, Indiana, and Illinois.

Tolls have financed construction and maintenance of most of the nation's great crossings, including such national icons as the Golden Gate Bridge; the San Francisco-Oakland Bay Bridge; the George Washington Bridge (the world's busiest bridge); and most of the crossings between the boroughs of New York City linking Manhattan, Queens, Brooklyn, Long Island, and the rest of America. Tolls have financed great tunnels such as the Lincoln, Holland, Queens-Midtown, and Brooklyn-Battery Tunnels in New York City, three tunnels in Boston, two in Baltimore, and one in Hampton Roads, Virginia. Important toll facilities exist in about half the states of the Union.

Tolls generate about $5.4 billion a year, or 6.7 percent of total highway user charges as defined by the USDOT.7 Toll roads constitute about 4,900 miles (9 percent) of America's 56,000 miles of interstates and other freeways and expressways.

Roads and their bridges and tunnels have to be financed one way or another, and tolls are an eminently sensible and fair way of paying. It is unfortunate that U.S. law should perpetuate the existing bias against a legitimate means of road payment that has produced wonderfully productive results for the nation.

For example, Section 301 of the U.S. Code reads: "Except as provided in section 129 of this title with respect to certain toll bridges and toll tunnels, all highways constructed under the provisions of this title shall be free from tolls of all kinds."8 State and local authorities that initiate and operate roads should have the discretion to levy tolls, use taxes, or even franchise road operations. The general presumption against tolling of roads contained in Section 301 is wrong and should be removed.

Section 129 maintains the presumption that citizens of the states need to be protected against tolls authorized by responsible state and local authorities. It directs that the Secretary of Transportation "shall permit" federal participation in "reconstruction or replacement or conversion of the (untolled) bridge or tunnel to a toll facility" but then imposes a long list of conditions for the Secretary's permission.9 It also says that the Secretary "shall permit" reconstruction of a toll-free federally aided highway to a toll facility but perplexingly excludes any "highway on the Interstate System" from consideration.10 If tunnels, bridges, and non-interstate federally aided highways can tap into toll revenues for reconstruction, why can states not convert interstate highways from free to toll?11

No rationale remains for this legal and regulatory straitjacket. The interstate highway system began as a mixture of toll and non-tolled roads in fairly equal proportions. Key segments of the interstate system remain toll roads, and over half of America's toll revenues are collected on interstate facilities. New toll facilities (for example, I-185 in Greenville, South Carolina) have gained interstate designation.

The unwarranted presumption against tolling is reflected in some provisions of Title 23. The Territorial Highway Program under Section 215, which provides U.S. support for roads in the Virgin Islands, Guam, American Samoa, and the Marianas, also requires that, as a condition of federal funding for roads in these territories, the governor of such territories "not impose any toll, or permit any toll to be charged for use by vehicles or persons of any portion of the facilities constructed or operated under the provisions of this section."12 Identical anti-toll boilerplate is in Section 218 (Alaska Highway financing) and Section 212 (Inter-American Highway).

In other words, the U.S. government quite sensibly requires a local commitment to maintain the road facilities it funds or helps fund in these places, but then blindly blocks the most logical way for the local government to fund that maintenance--tolls.

Other complex provisions in current highway law do not explicitly prohibit tolls, but they do tie them up in convoluted legalisms and bureaucratic departmental permitting processes seemingly designed to discourage tolls. Section 119 (Interstate Maintenance Program) contains a series of legal hurdles:

The Secretary may approve a project pursuant to this subsection on a toll road only if such road is subject to a Secretarial agreement provided for in section 129 or continued in effect by section 1012(d) of the Intermodal Surface Transportation Efficiency Act of 1991 (105 Stat. 1939) and not voided by the Secretary under section 120(c) of the Surface Transportation and Uniform Relocation Assistance Act of 1987 (101 Stat. 159).13

U.S. law leaves it entirely up to state and local governments to decide what kinds of fuel taxes, license fees, sales taxes, weight-distance truck charges, special district taxes, developer contributions, vehicular property taxes, and other levies to impose for funding roads and how to impose them. U.S. law asserts no interest in these non-toll road user charges. Yet, as shown, U.S. law asserts extensive controls over whether or not and under what conditions tolls may be levied.

Logically, if tolls, locally imposed, are at least as legitimate a funding mechanism as local taxes, U.S. law should leave decisions as to when, whether, and what kinds of tolls to levy for supporting roads to the discretion of state and local governments. Tolls should be an option to be exercised at local initiative, just as local governments are answerable to local voters for the results of other levies.

Strangely, these remaining old anti-toll provisions in Title 23 are matched by a number of pro-toll provisions in recent sections (for example Section 183, Secured Loans) designed to encourage innovative financing. Toll revenue is explicitly mentioned as a legitimate revenue stream for servicing debt incurred for road projects. One section recognizes the value of a toll revenue stream and even implies that it should be a factor in favor of projects.

This inconsistency within Title 23 makes no sense and should be cleaned up in the reauthorization by recognizing tolls as a legitimate means for states and local communities to finance and manage the roads they need.


People naturally prefer a free road to a toll road, all else being equal. But in the present financial predicament of governments, all else is rarely equal. The usual alternatives to a toll road are no road or increased taxes.

Tolls are most unpopular on an existing road. People think they have paid for the road with their fuel taxes even though a road is never truly "paid for." A road begins to deteriorate on the first day it is exposed to the sun and the rain and the pounding of vehicles, and it needs continuing maintenance and periodic reconstruction. Yet it is an uphill political fight to toll an existing free road.

Tolls are most acceptable when they finance a completely new road, but it is important that one part of a metro region not see itself as burdened with tolls while other areas are getting major new roads for free. A regional policy for concentrating gas tax money on maintenance and reconstruction, with all new capacity being tolled, may have a better chance of political acceptance than a random policy that tolls some major new roads and leaves others free.

HOT Lanes
Many jurisdictions around America are implementing or planning special or managed toll lanes within what are otherwise freeways. Those in operation include:

  • The 91 Express Lanes in Orange County in the greater Los Angeles, California, area
    Tolls range between $1.00 and $4.25 for a 10-mile trip, depending on the time of day and the level of congestion in the free lanes. The four toll lanes are separated from free lanes by yellow plastic guideposts. By metering traffic with a flexible toll to prevent the lanes from becoming overloaded, the operator is able to guarantee a free flow of traffic at 65 to 70 miles per hour (mph) while the unmanaged traffic in the adjacent free lanes bogs down in stop-and-go, averaging 20 to 30 mph for many hours. During peak hours, the managed lanes carry more traffic per lane than the free lanes because they keep moving.
  • Two reversible HOT (High Occupancy Toll) lanes in the median on an eight-mile stretch of I-15 north of San Diego, California
    Traffic travels south into San Diego in the mornings and north in the afternoons. If carpoolers are not making full use of the HOT lanes, solo drivers are allowed in for a toll. The actual toll rates are adjusted as frequently as every six minutes by a simple computer program and announced to motorists by a variable message sign on the roadside before the entry point. "Metering" entry by price ensures that traffic flows freely at all times. The project is so popular that the local government is extending it in stages to 20 miles and widening it to three lanes, with a moveable central barrier to allow adjustable operations in two directions or all three lanes in one direction.
  • Reversible single HOT lanes on I-10 (Katy Freeway) and US-290 (Northwest Freeway) in the Houston, Texas, area
    These HOT lanes began operations as HOV (High Occupancy Vehicle) lanes, but HOV-2 (two occupants required) resulted in overcrowding and stop-and-go conditions during rush hours, and HOV-3 (three occupants required) left the lanes with major spare capacity while the excluded two-occupant vehicles added to congestion in the free lanes. Converting to HOT lanes allowed a controlled number of two-occupant vehicles to utilize the spare capacity for a toll while also optimizing use of the HOT lanes and relieving congestion. Construction is currently underway on I-10 to replace the single reversible HOT lane with four toll lanes on the busiest stretch, between I-610 and TX-6.

Similar toll express lanes are under construction on the busiest part of the Lyndon B. Johnson Freeway (I-635) in Dallas, where a network of managed lanes are planned over the metro area. Managed lanes with tolls are being studied and designed in the San Francisco Bay area; on I-95 in Miami, Florida; on three highways in Denver, Colorado; on eight freeways in Seattle, Washington; on the Washington, D.C., Beltway (I-495); and elsewhere. These projects are assisted by the Value Pricing Program in the Transportation Equity Act for the 21st Century, the current federal surface transportation law.

The Reason Institute has proposed that ineffectual HOV lanes in eight of the busiest metro areas be used as the basis for a comprehensive and interconnected premium-lane network in each area. The proposal involves adding 2,730 lane-miles to the existing 1,700 lane-miles of HOV and building 300 new direct connector ramps at interchanges at a cost of $44 billion. About a quarter of the new lanes would have to be built centrally elevated. Buses and large transit vans travel for free, but cars would pay a toll for the privilege of driving in a managed lane where free-flow conditions are guaranteed.

It is estimated that tolls could cover $29 billion, or two-thirds of the total capital cost, of the Reason Institute's proposal. If transit were made to pay its way, whole networks could be completely self-supporting with tolls.14

Truck Tollways
Other special toll lanes have been proposed for trucks. Two investor groups are vying for the right to rebuild 325 miles of I-81 in Virginia in return for the right to toll. One of the plans would provide special truck toll lanes.

Another proposal is to build separate truck lanes or "toll truckways" on major trucking routes. The special lanes would be separated from regular traffic, allowing heavier, longer combinations, operating at higher speeds than in mixed traffic. According to modeling, these truckways would produce such major increases in trucking productivity that trucking companies would be prepared to pay substantial tolls.15 Such toll truckways, with their own staging areas for makeup and breakdown of long combinations and direct connector ramps to avoid mixing with cars, address safety concerns that have produced an impasse in rationalizing the nation's patchwork of truck size and weight regulations.

Tolls on Reconstructed Highways
Tolls may also be acceptable on major reconstruction and widening of existing interstates, especially if financed by a mix of toll revenue and grant funding. This has been proposed for I-70 between St. Louis and Kansas City.

Bridges have been tolled when reconstructed or enlarged. Examples include the Coleman Bridge in Yorktown, Virginia, and the Tacoma Narrows Bridge in Washington State. The Wilson Bridge on the Washington Beltway, currently under reconstruction, should be tolled since it will provide a greatly improved level of service to area commuters. Other toll-financeable bridges in design include a new bridge over the Mississippi River between St. Louis and southern Illinois and two new bridges over the Ohio River in the Louisville, Kentucky, area.


Reauthorization of federal funding for surface transportation presents Congress and the Administration with a perfect opportunity to effect long-needed and eminently practical changes in U.S. transportation policy. Specifically, surface transportation reauthorization should explicitly:

  • Allow states to authorize or impose tolls on newly constructed HOT or toll express lanes on interstates or other federal aid roads;
  • Allow states to convert HOV lanes to HOT lanes;
  • Allow states to build segregated toll truckways within the rights of way of interstate highways;
  • Authorize bond financing of for-profit highway projects on the same basis as for government authorities and not-for-profits, with the same rules on tax exemption;
  • Continue the Value Pricing Program, which has been the basis for many positive developments in highway funding and management, producing greater returns than programs costing a hundred times as much;
  • Encourage states to pursue telematic or in-vehicle computer-based tolling using global positioning system (GPS) equipment and other positioning devices as outlined in the multi-state supported report;16 and
  • Require that the planning process for major highway projects involving federal funds include a toll financing feasibility study, examining full and partial toll funding of the project and the potential impact of variable toll rates on traffic management.17


If the United States wants to maximize its resources for improving our highways, then all projects that could be financed with tolls rather than with highway trust fund grants should be toll-financed, leaving trust funds and tax money to support rebuilding, maintaining, and operating roads that can be financed only with such grants. Highway trust fund money is being wasted on projects like the Wilson Bridge, which are eminently toll-financeable. Wasting trust funds leaves less money for road projects that are not amenable to tolling.

Relaxing obsolete legal restrictions on tolling in U.S. law and encouraging toll financing could bridge the gap between the highway investment advocated by Transportation Committee Chairman Young and the opposition of the Bush Administration to higher gasoline taxes. A combination of new toll roads and bridges, toll express-lane networks, and toll truckways could realistically attract $21 billion per year of private capital into toll projects, covering the difference between the $41.2 billion per year proposed by the Bush Administration and the $62.5 billion per year proposed by House Transportation Committee leaders. The gap should not--and need not--be filled with higher taxes.

Peter Samuel is editor of Toll Roads Newsletter and an adjunct scholar at the Reason Public Policy Institute.

1. "The Long-term Value of Value Pricing in Metropolitan Areas," Transportation Quarterly, Vol. 56, No. 3 (Summer 2002), p. 19.

2. U.S. Department of Transportation, "2002 Status of the Nation's Highways, Bridges, and Transit: Conditions & Performance," p. ES-6.

3. Texas Transportation Institute, "2002 Annual Urban Mobility Report," Texas A&M University, College Station, at

4. U.S. Department of Transportation, "2002 Status of the Nation's Highways, Bridges, and Transit," p. ES-14.

5. Ibid., Exhibit 6-6, p. 6-9.

6. U.S. Code, Vol. 23, Chap. 3, Sec. 301.

7. U.S. Department of Transportation, "2002 Status of the Nation's Highways, Bridges, and Transit," Exhibit 6-2, p. 6-4.

8. U.S. Code, Vol. 23, Sec. 301 (emphasis added).

9. U.S. Code, Vol. 23, Sec. 129(a) (1) (C).

10. U.S. Code, Vol. 23, Sec. 129(a)(1)(D).

11. A very limited number of interstate tolls are allowed under two pilot programs (value pricing and interstate tolls), but only after a complex selection process.

12. U.S. Code, Vol. 23, Section 215(c)(2).

13. U.S. Code, Vol. 23, Section 119.

14. See Robert W. Poole, Jr., and C. Kenneth Orski, "HOT Networks: A New Plan for Congestion Relief and Better Transit," Reason Institute Policy Study No. 305, February 2003, at

15. See Peter Samuel, Robert W. Poole, Jr., and José Holguin-Veras, "Toll Truckways: A New Path Toward Safer and More Efficient Freight Transportation," Reason Institute Policy Study No. 294, June 2002, at

16. David Forkenbrock and Jon G. Kuhl, "A New Approach to Assessing Road User Charges," University of Iowa, Public Policy Center, 2002.

17. Such a requirement would not mandate the use of tolls, but it would enlighten the public debate about the options. Texas has a similar requirement that all major highway projects conduct a toll feasibility study. Texas is also examining the option of special managed lanes, including toll lanes, on major projects.

About the Author

Related Issues: Smart Growth