The user-pays principle already ended. That sentence sounds like a verdict, and it is meant as one, because the comparison between the Highway Trust Fund and general fund financing of roads is not really a contest between two live options. In August 2016, one of them has stopped working in practice, and the other has already taken its place, quietly, year after year, through transfers that Congress authorizes and describes in the language of the old system. The honest question is no longer whether roads should be paid for by users or by everyone. It is whether Congress should rebuild a user-fee system that can actually pay for the network, or admit that the network is funded from general revenue and design that funding openly. The deciding factor is where the reader believes discipline should come from: whether transportation spending should be constrained by what drivers pay, or determined by what the transportation network needs. Everything in this comparison follows from that choice, because the reader’s preference will track whichever risk they consider greater, the risk of a network starved by a frozen tax, or the risk of a budget freed from the discipline of the pump.

The statutory identity of the institution at the center of this story is worth stating plainly at the outset. The Highway Trust Fund was created by the Highway Revenue Act of 1956 (P.L. 84-627; 70 Stat. 387), signed on June 29, 1956, as the financing arm of the Interstate Highway System. The Act imposed a three-cent-per-gallon federal tax on gasoline and dedicated the proceeds to highway construction, a design that married the spending program to its revenue source in a single legislative bargain. For the first half century of its life, that bargain held in the form its authors intended: drivers paid at the pump, the money flowed into a dedicated account, and the account paid for roads. What happened afterward, and why the bargain broke, and what should replace it, is the substance of this comparison.

The Design Congress Built in 1956

To compare the trust fund with general fund financing, it helps to start with what each arrangement actually is, because the two are not merely two bank accounts. They are two different theories of how public spending should be disciplined, and each carries its own logic, its own political economy, and its own failure modes.

What the Highway Trust Fund actually is

The Highway Trust Fund is a dedicated federal account that collects specific excise taxes, mostly on motor fuels, and spends them on surface transportation. In August 2016, the federal gasoline tax stood at 18.4 cents per gallon and the diesel tax at 24.4 cents per gallon, rates that have not moved since October 1, 1993. A tenth of a cent of each goes to the Leaking Underground Storage Tank trust fund, which means the Highway Trust Fund itself receives 18.3 cents per gallon of gasoline and 24.3 cents per gallon of diesel. Of the gasoline tax, 15.44 cents flows to the Highway Account and 2.86 cents to the Mass Transit Account; of the diesel tax, 21.44 cents goes to the Highway Account and 2.86 cents to the Mass Transit Account. These are the rates as they stood at the article date, and every figure in this comparison is tied to the sources and periods given alongside it.

The design rests on what transportation economists and legislators call the user-pays principle, the idea that the people who benefit from the road network should finance it in rough proportion to their use. The fuel tax was chosen as the measuring instrument because, in an era when nearly every vehicle burned gasoline or diesel, gallons consumed tracked miles traveled closely enough to serve as a workable proxy. The heavier the vehicle and the more it traveled, the more fuel it burned, and the more it paid. That proxy relationship was the genius of the 1956 bargain: it made the financing feel fair without requiring any new bureaucracy to measure road use, because the pump already did the measuring.

What does the user-pays principle actually require?

The user-pays principle requires that beneficiaries fund a service in proportion to their use, so the price of use constrains how much gets built. For highways, that meant fuel taxes rising and falling with use, receipts staying dedicated, and authorized spending tracking receipts through the Byrd rule at 26 U.S.C. Section 9503(d).

That definition matters because it sets the test the rest of this article applies. A financing system satisfies the user-pays principle only when three conditions hold together: the revenue comes from the users of the facility, the revenue is dedicated to the facility so it cannot be diverted to unrelated purposes, and the revenue disciplines the spending so that authorizations cannot permanently outrun receipts. The 1956 design satisfied all three. The initial three-cent gasoline tax fell on motorists, the Highway Revenue Act dedicated the receipts to the new account, and the pay-as-you-go mechanism that Senator Harry Byrd added as an amendment required the Secretary of Transportation to cut state apportionments pro rata if unfunded authorizations exceeded projected receipts for the next four years. That third condition was the enforcement arm of the principle, and its quiet weakening is one of the central facts this article documents.

The pay-as-you-go bargain and the Byrd rule

The 1956 arrangement added a second element that made the user-pays design unusually strict. The Highway Revenue Act included what became known as the Byrd amendment, named for Chairman Harry Byrd, which wrote a pay-as-you-go discipline directly into the financing law. In its current codified form at 26 U.S.C. Section 9503(d), the provision says that if unfunded highway authorizations exceed projected receipts for the next four years, the Secretary of Transportation must reduce state apportionments on a pro rata basis. In plain terms, Congress wrote a rule that spending could not outrun the fuel tax for long: if the money was not there, the authorizations had to shrink. That mechanism was the fiscal conscience of the trust fund, and it deserves attention because its later history explains so much of what followed. The Byrd rule was designed to force a choice. Either the tax rate would rise to support the spending Congress wanted, or the spending would fall to match the revenue the tax produced. For decades, that choice kept the system honest, at least in the accounting sense. The full story of the 1956 highway act and its financing bargain is told in the guide to the Federal-Aid Highway Act of 1956, which traces how the construction program and its revenue title were built together.

What general fund financing means

General fund financing is the ordinary way the federal government pays for most of what it does. Income taxes, corporate taxes, and other broad-based revenues flow into the Treasury’s general accounts, and Congress appropriates money from those accounts to programs through the annual appropriations process. There is no dedicated revenue source, no user fee, and no automatic link between what a program’s beneficiaries pay and what the program spends. A dollar of defense spending, a dollar of education spending, and a dollar of highway spending all come from the same pool, and each must compete with the others every year in the appropriations committees and in the broader budget negotiation.

The defenders of general fund financing argue that this is exactly as it should be. Roads, in this view, are a public good whose benefits spill far beyond the drivers who use them: freight moves the goods that fill every store, emergency vehicles travel the same pavement, and the economic activity the network enables is taxed through income and corporate taxes anyway. Why, the argument runs, should a nurse commuting by bus, or a retiree who no longer drives, be exempt from paying for a network that delivers their groceries and their ambulance? The critics answer that severing the link between use and payment removes the only discipline that ever kept transportation spending honest, and turns highway funding into one more claimant in the annual scramble for appropriations, subject to the moods of deficit politics rather than the measured reality of the pump.

The comparison at its simplest

Strip away the rhetoric and the comparison reduces to a question about discipline. Under the trust fund model, the constraint on spending is supposed to be revenue: the program spends what its dedicated taxes produce, and when the two diverge, something must give, either the rate or the spending level. Under general fund financing, the constraint on spending is politics: the program spends what the appropriations process delivers, competing against every other national priority, with deficits absorbing whatever the revenue system does not cover. Each model disciplines spending, but each disciplines it through a different mechanism, and each fails in a different way. The trust fund fails when its revenue instrument stops tracking use, because then the discipline becomes a straitjacket that starves the program. General fund financing fails when the appropriations process undervalues the program, because then the discipline becomes a popularity contest that the road network can lose. Which failure mode frightens the reader more is the beginning of the deciding factor, and it will return at the end of this comparison as the named ground on which the verdict rests.

The trust in trust fund

The word trust does a great deal of work in this debate, and most of it is psychological. A federal trust fund is not a trust in the private-law sense. There is no trustee with fiduciary duties, no beneficiary who can sue, and no lockbox that Congress cannot open. It is an accounting mechanism within the Treasury: designated revenues are credited to the fund’s accounts, designated spending is charged against them, and Congress retains full authority to change the revenues, the spending, or the designation itself by ordinary statute. The Highway Trust Fund’s Highway Account and Mass Transit Account are sub-ledgers of this kind, created by law and alterable by law. Understanding this matters because so much of the political argument proceeds as if the trust fund were a vault whose integrity must be defended against raiders. The integrity that matters is not physical but procedural: the Byrd amendment’s rule that spending must track receipts was the mechanism that gave the dedication its force, and once Congress began filling gaps with general revenue, the dedication became a labeling convention rather than a constraint. The label still shapes behavior, because legislators, journalists, and voters treat trust fund spending as paid for and therefore legitimate, which is precisely why the label’s survival after the discipline’s death deserves scrutiny.

Why 1956 chose the driver over the taxpayer

The 1956 bargain looks inevitable in retrospect, but it was a choice, and the reasons for it illuminate the whole comparison. The Interstate Highway System was the largest public works program in American history to that point, and financing it from general revenue would have forced it to compete, dollar for dollar, with defense and every other claim on the Treasury in the annual budget process. The user-fee model offered an escape from that competition: a dedicated tax, paid by the drivers who would use the roads, flowing into a dedicated fund, spent on a defined program, with the Byrd rule guaranteeing that the program could not outgrow its revenue. That self-containment was the political genius of the arrangement. It assembled a coalition of truckers, automakers, construction firms, state highway departments, and motorists around a proposition each could accept: pay in proportion to use, and what is paid builds the roads driven. The choice also reflected the technology of the moment. In 1956, the fuel tax was a nearly perfect user fee, because the vehicle fleet was homogeneous in the relevant respect: everything burned gasoline or diesel, so gallons tracked use with an accuracy no other available measure could match. Fiscal design decisions made in 1956 were still constraining policy decades later, not because anyone in 1956 was shortsighted, but because a financing mechanism that works brilliantly for its own era becomes infrastructure itself, and infrastructure is hard to replace.

The apportionment pipeline: from pump to pavement

The comparison between trust fund and general fund financing becomes more concrete when one follows a dollar from the pump to the pavement, because the two models move money through different channels and each channel shapes what the money can do. Under the trust fund model, the journey begins at the refinery gate and the import terminal, where the federal excise is collected from a small number of taxpayers high in the distribution chain. That collection point is one of the fuel tax’s great administrative virtues: rather than chasing hundreds of millions of drivers, the Treasury collects from a few thousand firms, which keeps compliance costs low and evasion difficult. The receipts flow into the Highway Trust Fund’s accounts, and from there the Federal Highway Administration distributes them to the states through apportionment formulas written into the authorizing statute. The states then obligate the funds to specific projects, matching federal dollars with their own revenue, and the money becomes asphalt, bridges, and transit grants.

Each step in that pipeline reflects a design choice with consequences. Collecting high in the chain makes the tax invisible to the driver, who experiences it only as a component of the pump price, which is politically convenient but democratically opaque: few voters can say what share of their fuel dollar goes to Washington. Apportioning by statutory formula rather than by annual appropriation gives states the predictability they need for multi-year construction planning, but it also locks in distributional bargains struck in earlier decades, because formulas are rewritten only in reauthorizations. Requiring state matching funds stretches federal dollars across more projects, but it also means that the federal program’s reach depends on state fiscal capacity, so poorer states can struggle to claim the apportionments the formula offers them. The general fund alternative would change this pipeline at every stage. Revenue would be collected through the income tax system, appropriated annually to transportation programs, and distributed through mechanisms Congress chose each year. That would make transportation funding more visible and more contestable, subjecting it to the same scrutiny as other spending, but it would also make it less predictable, because a highway project that takes five years to build cannot be planned against appropriations that are renegotiated every twelve months. The pipeline, in short, is not neutral plumbing. It determines who decides, how far ahead they can plan, and how visible the costs are to the people who pay them.

The federal-state partnership and its tensions

The trust fund was never a purely federal program, and the partnership with the states shapes every financing question. Washington collects the fuel tax and apportions the proceeds, but the states select the projects, supply matching funds, own and maintain the roads, and live with the consequences of federal decisions. This division was deliberate: the 1956 Act created a national system built by state highway departments, marrying federal financing to state execution. The partnership’s great strength is local knowledge; the states know their corridors, their bridges, and their needs better than any federal agency could. Its great tension is that the two levels of government face different incentives around the trust fund’s solvency. States plan against apportionments and experience a shortfall as an immediate threat to projects already underway, which makes them fierce defenders of the transfers that keep the money flowing. The federal legislators who authorize those transfers experience the shortfall as an abstract budget problem, solvable with general revenue that no voter will trace back to them. The result is a partnership in which the level of government closest to the pavement has the least power over the financing, and the level with the most power is the most insulated from the consequences.

The matching requirement deepens the tension. Federal aid typically requires states to contribute their own funds alongside, which means the federal program’s benefits flow disproportionately to states that can afford the match, while the fuel taxes that finance the federal share are collected wherever fuel is burned. Defenders of matching argue that it disciplines state project selection, forcing states to put their own money behind their priorities, and that it multiplies the federal investment. The counterargument is that it biases the program toward wealthier states and toward capital construction over maintenance, since matching is easier to arrange for ribbon-cutting projects than for the unglamorous work of preservation. Under general fund financing, these tensions would not disappear; they would migrate. Annual appropriations would give Congress more direct control over which states and projects benefit, which the partnership’s defenders see as a threat to the formula-driven predictability that lets states plan, and which its critics see as an overdue correction to formulas that entrench old bargains. The partnership question is thus inseparable from the financing question: whoever pays for the roads will, sooner or later, decide how the money is divided, and the states know it.

The committee split that froze the rate

One reason the fuel tax rate froze while spending grew is buried in congressional procedure: the committees that authorize transportation spending are not the committees that write tax law, and the two operate on different calendars, under different pressures, with different memberships. Surface transportation authorizations come from the public works committees, whose members are elected to deliver projects and whose institutional interest lies in generous, predictable programs. Tax rates are the province of the revenue committees, whose members absorb the political cost of every increase and whose institutional interest lies in protecting the broader tax system from narrow raids. This division of labor made sense in 1956, when the bargain was struck as a package: the revenue title and the spending title moved together, and each committee’s work disciplined the other’s. It makes less sense when the package comes apart, as it did after 1993. The authorizers kept writing generous programs, the revenue writers kept declining to raise the rate, and no single committee owned the gap between them. The transfers that filled the gap were typically attached to larger legislative vehicles, negotiated by leadership rather than by either committee alone, which meant the financing decision was made everywhere and nowhere. The structural consequence is that the committees with the strongest incentive to fix the revenue base lack the jurisdiction to do it, and the committees with the jurisdiction lack the incentive. Any reform that hopes to restore the link must reckon with that separation, whether by indexing the rate so Congress need not act, or by moving transportation to general funding so the appropriations process owns the whole decision. For readers who want to trace how money moves from the fuel tax through apportionment to the states, the explainer on federal highway funding mechanics walks through the machinery step by step.

How much of each gallon’s federal tax reaches the Highway Trust Fund?

Of the 18.4-cent federal gasoline tax in effect in August 2016, the Highway Trust Fund receives 18.3 cents; the remaining tenth of a cent goes to the Leaking Underground Storage Tank fund. The trust fund share divides into 15.44 cents for the Highway Account and 2.86 cents for the Mass Transit Account.

How the Fuel Tax Lost Its Force

If the trust fund was a machine for converting gallons into pavement, then the story of the last quarter century is the story of that machine seizing up. Three distinct causes compounded to erode the revenue base, each one understandable on its own, each one defended by its own constituency, and together forming a squeeze that no single policy adjustment can reverse. Before examining them, it is worth noting the reauthorizations that kept spending moving while the revenue side froze: the Intermodal Surface Transportation Efficiency Act of 1991 (P.L. 102-240, signed December 18, 1991), the Transportation Equity Act for the 21st Century of 1998 (P.L. 105-178, signed June 9, 1998), the Safe, Accountable, Flexible, Efficient Transportation Equity Act of 2005 (P.L. 109-59, signed August 10, 2005), the Moving Ahead for Progress in the 21st Century Act of 2012 (P.L. 112-141, signed July 6, 2012), and the Fixing America’s Surface Transportation Act of 2015 (P.L. 114-94, signed December 4, 2015). Five times across twenty-four years, Congress reauthorized surface transportation spending, expanded its purposes, and left the fuel tax rate untouched. The longer history of infrastructure legislation shows how the program grew in ambition while its financing stood still, and the gap between those two lines is the subject of this section.

The rate that stopped moving

The federal gasoline tax was last raised on October 1, 1993, by the Omnibus Budget Reconciliation Act of 1993 (P.L. 103-66), which added 4.3 cents per gallon and brought the rate to the 18.4 cents that still stood in August 2016. To grasp what twenty-three years of a frozen nominal rate means, consider the two ways of measuring what inflation did to it. The Congressional Research Service, in its analysis of the tax (CRS R48472), found that the gasoline tax lost approximately 73 percent of its purchasing power measured against highway construction-cost inflation. That is the more punishing yardstick, because the price of asphalt, concrete, steel, and labor rose faster than the general price level across this period. Measured against the Consumer Price Index instead, the Congressional Budget Office and CRS analyses put the 18.3 cents that reaches the trust fund at roughly 29 to 31.7 cents in inflation-adjusted terms, which is a loss of about 37 to 40 percent of purchasing power. Both measures describe the same reality from different angles: a tax fixed in nominal cents buys steadily less road over time. The statement that the frozen rate cut the tax’s real value by more than half is valid only when tied to highway construction costs; tied to the Consumer Price Index, the decline is closer to two-fifths. Precision matters here because the choice of inflation measure is itself contested, and each side of the funding debate reaches for the yardstick that flatters its case.

The deeper point is structural. A per-gallon excise fixed in nominal cents is, by construction, a declining tax in real terms. Every other broad-based federal revenue source grows with the economy: income tax receipts rise with wages, payroll taxes rise with employment, and even ad valorem taxes rise with prices. A cents-per-gallon levy does none of those things. It sits still while everything around it moves, which means that maintaining its real value requires Congress to act affirmatively, raising the rate every few years, in full public view, against organized opposition. The 1983 increase under the Surface Transportation Assistance Act (P.L. 97-424, Title V, signed January 6, 1983) shows what such action looked like when it happened: Section 511 of that act raised the gasoline and diesel tax from 4 cents to 9 cents per gallon, effective April 1, 1983, and dedicated one penny of the five-cent increase to a newly created Mass Transit Account, the arrangement remembered as the Great Compromise. The 1993 increase shows the same pattern on a smaller scale. And then, after 1993, the pattern stopped. The reason it stopped is not mysterious. Raising a visible tax on every driver is among the least pleasant votes a legislator can cast, and the procedural path for tax legislation, which must move through the revenue committees under rules described in the guide to how tax bills move through Congress, concentrates the political cost on the members who carry it. So Congress took the easier path: it kept authorizing spending, left the rate alone, and let inflation do the quiet work of cutting the tax. That choice, repeated across five reauthorizations, was the first of the three causes.

Why does the 1993 rate freeze matter more than any single spending decision?

The freeze matters because a per-unit excise that never changes loses value against rising prices automatically, while each spending decision is debated and can be reversed. Between October 1993 and August 2016 the rate sat at 18.4 cents as construction costs climbed, so the erosion compounded without a single vote being cast.

Cars that sip instead of gulp

The second cause is one that most Americans would call good news, which is precisely what makes it politically difficult to discuss. The vehicle fleet became substantially more fuel efficient. According to the Environmental Protection Agency’s Light-Duty Automotive Technology, Carbon Dioxide Emissions, and Fuel Economy Trends report (EPA-420-S-16-001), the real-world fuel economy of new light-duty vehicles rose from 20.9 miles per gallon in model year 1993 to 24.8 miles per gallon in model year 2015, an increase of 18.7 percent, and the fleet was 28 percent more efficient than it had been in model year 2004. Those figures describe new vehicles; the on-road fleet turns over slowly, so the revenue effect arrived with a lag, but it arrived all the same. A driver who travels the same miles in a vehicle that burns fewer gallons pays less fuel tax for the same road use. Multiply that driver by tens of millions, and the trust fund’s revenue per mile of travel falls even as the miles keep coming.

This is the paradox at the heart of the erosion story. Fuel economy standards were designed to reduce oil consumption, cut emissions, and save drivers money, and by those measures they succeeded. But the trust fund’s revenue model assumed that gallons and miles would move together, and efficiency broke that assumption. Every mile-per-gallon gained is a small tax cut delivered automatically, without a vote, to every driver of an efficient vehicle, paid for by the road program. Supporters of higher efficiency standards tend to treat this as a problem for the financing system to solve rather than a reason to slow efficiency gains, while defenders of the fuel tax point out, with some justice, that a financing system cannot survive if its base is programmed to shrink by law. Both positions have merit, which is why the efficiency cause belongs in any honest account: it is nobody’s fault and everybody’s problem.

Vehicles that skip the pump entirely

The third cause is the newest and, over a long horizon, the most threatening to the fuel tax model. Electric vehicles pay no federal fuel tax at all. Under 26 U.S.C. Section 4081, the federal excise applies to taxable fuels, which the statute defines to include gasoline, diesel fuel, and kerosene; electricity is not a taxable fuel, so a vehicle that runs on electricity contributes nothing to the Highway Trust Fund no matter how many miles it travels or how much pavement it wears. The Congressional Research Service put the point directly (CRS R48472): the fuel taxes, in the Service’s words, do not account for miles traveled by electric vehicles. Through August 2016, the electric share of the fleet remained small enough that the revenue effect was modest, but the direction was unmistakable, and every projection of vehicle technology pointed the same way. A tax on gallons cannot capture miles traveled by vehicles that use no gallons, and no rate increase on gasoline can fix that, because the vehicles in question buy no gasoline at any price.

The electrification cause differs from the other two in an important respect. Inflation erosion can be addressed by indexing, and efficiency erosion can be debated as a policy trade-off, but electrification severs the proxy relationship itself. The fuel tax worked as a user fee only because fuel consumption tracked road use. Once a meaningful share of travel consumes no taxed fuel, the tax stops being a rough measure of use and becomes a levy on a shrinking subset of drivers, disproportionately those who cannot yet afford newer or electric vehicles. That distributional sting is worth stating plainly, because it shapes the fairness debate around every alternative: as the fleet electrifies, the remaining fuel tax burden falls more heavily on lower-income drivers of older gasoline vehicles, which is the opposite of what a fair user fee should do.

If electric vehicles use the roads, why does the tax code miss them?

The tax code misses them because 26 U.S.C. Section 4081 taxes taxable fuels, defined as gasoline, diesel fuel, and kerosene, and electricity is not among them. The statute was written for liquid fuels removed from terminals and refineries, so electric miles travel outside the taxable event entirely, and no rate change can reach them.

Three causes that multiply

The three causes do not add; they multiply. A tax that loses value to inflation, applied to fewer gallons per mile, collected from a shrinking share of vehicles, declines faster than any single cause would suggest. If inflation cuts the real value of each cent by two-fifths on the consumer-price measure, and efficiency cuts gallons per mile by roughly a fifth for new vehicles, then the real revenue per mile from a new efficient vehicle is down by more than half even before electrification enters the picture, and the construction-cost measure puts the loss near three-quarters. The trust fund did not face one headwind; it faced three, blowing in the same direction, for more than two decades, while Congress held the nominal rate fixed and kept authorizing spending. That is the mechanical explanation for the gap that opened in 2008, and it is the reason the gap, once opened, never closed.

The arithmetic of the gap

It helps to walk through the compounding slowly, because the three causes are often discussed in isolation and their combined effect is larger than the sum suggests. Start with a dollar of 1993 fuel tax revenue, measured in what it could buy in highway construction. By the Congressional Research Service’s construction-cost accounting (CRS R48472), the gasoline tax had lost approximately 73 percent of its purchasing power on that measure, which means each cent collected bought a little more than a quarter of the pavement it once bought. On the Consumer Price Index measure, the same 18.3 cents that reaches the trust fund equated to roughly 29 to 31.7 cents in inflation-adjusted terms, a loss of about 37 to 40 percent. Either way, the first cause cut deeply. Apply the second. Between model year 1993 and model year 2015, new light-duty real-world fuel economy rose from 20.9 to 24.8 miles per gallon, an 18.7 percent improvement, with a 28 percent gain since model year 2004 (EPA-420-S-16-001). For the newer portion of the fleet, that means roughly a fifth fewer gallons, and therefore roughly a fifth less tax, per mile traveled. Combine a two-fifths inflation loss with a one-fifth efficiency loss and the real revenue per mile from an efficient vehicle is down by more than half on the consumer-price measure, and by far more on the construction-cost measure. Then comes the third cause, which does not reduce the tax per mile but eliminates it: under 26 U.S.C. Section 4081, electricity is not a taxable fuel, so every mile traveled electrically contributes nothing. The arithmetic is illustrative rather than exact, because fleet turnover lags, diesel and gasoline splits differ, and the inflation measure is contested, but the direction and the order of magnitude are not in doubt. A financing system that loses two-fifths of its real value per unit, collects fewer units per mile, and watches an expanding share of miles exit the base entirely cannot sustain the spending its authorizing statutes promise. The gap that opened in 2008 was not a surprise in the data; it was the predictable product of this arithmetic, visible years in advance to anyone who ran the numbers.

Why a higher gas tax would not finish the job

This is the complication on which the most common reform proposal fails. Raising the fuel tax, or indexing it to inflation so that it rises automatically, would address only the first of the three causes. Indexing would stop the inflation bleed; it would restore the real value of each cent collected and keep it restored without requiring Congress to vote for increases. What it would not do is restore the gallons. An indexed tax on fuel still collects nothing from electric vehicles, still collects less from every efficient vehicle than from every inefficient one, and still watches its base shrink as the fleet turns over. The proponents of indexing answer that it is the most administratively simple fix available, that it preserves the user-pays link for the large majority of travel that still burns fuel, and that the perfect should not be the enemy of the good. The critics answer that indexing a declining base is a stay of execution rather than a cure, and that it would lock in a financing model whose fairness erodes with every electric vehicle sold. Both claims are substantially true, which is why the indexing debate recurs without resolution: it is a real improvement that does not solve the underlying problem, and its supporters and its detractors are describing different halves of that fact.

The three-cause test can be applied systematically to every proposal, and it is the most useful analytical tool this comparison offers. Take any financing proposal and ask three questions. Does it restore the purchasing power the frozen rate lost? Does it keep pace with fuel economy gains, so that revenue per mile does not fall as vehicles grow more efficient? Does it reach the miles traveled on electricity? A higher fuel tax answers the first question yes, the second only temporarily, and the third no. Indexing the tax answers the first automatically, the second no better than a one-time increase, and the third no. A mileage charge answers all three yes, which is why it is the only proposal that survives the test in full, though it raises the administrative and privacy questions the alternatives section examines. Tolling answers a different question entirely, tying payment to specific facilities rather than to system use. General fund financing declines the test, abandoning the user-pays framework the questions assume. The test does not pick a winner. It clarifies what each option costs and what each leaves unsolved, which is what a reader needs before applying the deciding factor.

A program that grew while its fuel shrank

The five reauthorizations between 1991 and 2015 did not merely extend the 1956 program; they transformed its purposes while leaving its financing untouched, and that widening of ends against a narrowing of means is an independent source of the gap. The Intermodal Surface Transportation Efficiency Act of 1991 reimagined the program around intermodalism, giving metropolitan planning, transit, and non-highway purposes a standing they had never held. The Transportation Equity Act for the 21st Century of 1998 and the Safe, Accountable, Flexible, Efficient Transportation Equity Act of 2005 continued the expansion, adding programs, set-asides, and eligibilities that reflected each era’s coalition. The Moving Ahead for Progress in the 21st Century Act of 2012 consolidated and reformed, and the Fixing America’s Surface Transportation Act of 2015 extended the framework for five more years. Each of these laws was a genuine legislative achievement, negotiated across parties and regions, and each left the fuel tax rate exactly where the Omnibus Budget Reconciliation Act of 1993 had put it. The pattern is worth stating as a general rule of this period: Congress repeatedly found the votes to authorize more spending for more purposes, and never found the votes to raise the revenue to pay for it.

The states did not wait

While the federal rate stood still, the states faced the same arithmetic and many of them acted on it, which creates a natural experiment in what happens when the paralysis lifts. State legislatures, which set their own motor-fuel levies on top of the federal rate, repeatedly adjusted those levies across this period: some raised the per-gallon rate by statute, some indexed it to inflation or to construction costs, some tied it to the price of fuel, and some restructured it entirely. The pattern is instructive. Where the federal Congress faced a concentrated political cost for raising a visible national tax, individual states faced different political calculations, and enough of them moved to demonstrate that the obstacle was political rather than technical. Raising a fuel tax is administratively trivial; the collection machinery already exists. What is hard is assembling the votes, and the states showed that the votes could be assembled when the need was concrete and the alternative was visible to voters. States that explored mileage-based charges contributed practical lessons that informed the federal pilot programs, including Oregon’s OReGO effort, which grew directly out of state-level experimentation.

The forecasts saw it coming

The gap that opened in 2008 surprised no one who followed the numbers. Years before the trust fund’s balances ran out, the Congressional Budget Office, the Congressional Research Service, and the Department of Transportation’s own analysts were projecting the point at which dedicated revenues would fall short of authorized spending, and their projections converged on the same answer: without a rate increase, a spending reduction, or a new revenue source, the fund could not sustain its obligations. The arithmetic was not subtle. A nominal rate fixed since 1993, eroding against inflation, applied to a fleet burning fewer gallons per mile, could not support spending levels that rose with construction costs and programmatic ambition. The forecasters ran the scenarios, published the charts, and testified before the committees, and the committees thanked them and carried on. One explanation for the inaction is electoral: the legislators who would have paid the price for a tax increase were not the legislators who would have reaped the reward of a solvent trust fund, because voters punish visible taxes immediately and reward sound infrastructure financing diffusely, if at all. Another is institutional: the committee split meant that no single set of legislators owned both the warning and the remedy. The forecasts saw it coming; the politics looked away. That is not a failure of analysis. It is a datum about what analysis can and cannot move.

Inflation’s quiet politics

There is a reason the nominal rate survived twenty-three years without a vote, and it is not that Congress forgot. A tax fixed in nominal cents is a tax that cuts itself, and a self-cutting tax is the most politically convenient kind. Every year the rate sat still, inflation reduced the real burden on drivers without any legislator having to cast a vote, take a position, or defend a record. The 73 percent loss of purchasing power against highway construction-cost inflation, and the 37 to 40 percent loss against the Consumer Price Index, were not merely economic facts; they were political accomplishments, achieved without fingerprints. Legislators who would have faced furious opposition for voting to cut the gas tax by two-fifths achieved exactly that cut by doing nothing, and drivers who would have protested a rate increase experienced the erosion as stable pump prices rather than as a policy choice. The quiet has a cost, and the cost is borne by the program the tax was meant to finance. A voted tax cut is debated, scored, and weighed against its consequences; an inflation-eroded tax is none of those things, so the revenue loss arrives without any corresponding decision about what the program should stop doing. That is how Congress came to authorize spending the tax could not support: the spending decisions were made loudly, in reauthorizations with press releases and signing ceremonies, while the revenue decisions were made silently, by inaction, in the space between votes.

The 1993 increase in retrospect

The last increase of the federal fuel tax repays close attention, because the circumstances of its passage explain why nothing like it happened again. The Omnibus Budget Reconciliation Act of 1993 (P.L. 103-66) added 4.3 cents per gallon effective October 1, 1993, bringing the gasoline rate to 18.4 cents and the diesel rate to 24.4 cents. It was enacted as part of a deficit-reduction package, passed without a single vote to spare in a closely divided Congress, and signed by a president who spent the rest of his tenure insisting he regretted the tax increases it contained. The fuel tax increase was not, in other words, a transportation decision. It was a budget decision that happened to touch transportation revenue, carried along by a reconciliation vehicle whose purpose was closing the deficit. A portion of the 1993 increase was initially directed to deficit reduction rather than to the trust fund, a detail that underscores how the fuel tax had become a general revenue instrument in the eyes of the budget process even before the transfers made that character explicit. The political aftermath taught every legislator who watched it the same lesson. The 1993 package became a weapon in the elections that followed, its tax increases invoked as evidence of government overreach, and the Congress elected in 1994 arrived with a deep institutional aversion to voted tax increases of any kind. The fuel tax, visible on every pump and paid by nearly every voter, was the worst possible candidate for an increase in that environment: its costs were concentrated on an identifiable act by identifiable legislators, while its benefits were spread across a road network whose improvements voters experienced as background conditions rather than as political achievements. After 1993, the political system demonstrated, repeatedly and across party control, that it would not raise the fuel tax again, and every subsequent decision was made in the shadow of that demonstrated fact.

Why did Congress leave the gas tax unchanged after 1993?

Raising a visible per-gallon levy concentrates political cost on the legislators who carry it, while the benefits are diffuse. Across five reauthorizations from 1991 through 2015, Congress found it easier to authorize spending against a frozen rate and let inflation quietly reduce the real burden than to vote for an increase.

What Replaced User Financing in Practice

Erosion is a slow process until it is not. For years the trust fund absorbed the three causes through balances accumulated in better times, drawing down the cushion that earlier surpluses had built. Then the cushion ran out, and the system crossed a line that changed its nature. The Congressional Research Service describes the crossing in plain language (CRS R45350.1): in 2008 Congress began providing Treasury general fund transfers to keep the Highway Trust Fund solvent, and every year since 2008 there has been a gap between the dedicated tax revenues flowing into the fund and the cost of the surface transportation spending Congress has authorized. That sentence is the hinge of this entire comparison. Since 2008, the Highway Trust Fund has been unable to support the spending Congress authorizes from its own dedicated revenues. The user-pays principle did not fade; it stopped, and general revenue took its place.

The gap that opened in 2008

The year 2008 matters for two reasons, one mechanical and one symbolic. Mechanically, it was the year the arithmetic finally failed: dedicated receipts fell short of authorized outlays, and the shortfall could no longer be covered from accumulated balances. Symbolically, it was the year the trust fund became, in practice, a general fund program wearing a trust fund’s name. Congress responded the way legislatures respond to a financing gap they do not wish to close with taxes or with cuts: it moved money from the general fund. The first such transfer came on September 15, 2008, under P.L. 110-318, in the amount of 8.017 billion dollars. That transfer was presented at the time as an extraordinary measure, a bridge across a temporary shortfall. It was not temporary. The gap recurred every year afterward, and the transfers recurred with it, growing from an emergency patch into the structural financing of the program.

It is worth pausing on what the Byrd rule was supposed to do at this moment, because the contrast is instructive. The 1956 design said that when authorizations outran receipts, apportionments to the states would be cut pro rata. That was the discipline mechanism: spend what the tax produces, or raise the tax. In 2008, faced with exactly the circumstance the Byrd amendment was written for, Congress did neither. It did not raise the fuel tax, and it did not cut apportionments. It transferred general revenue instead. That decision, renewed in every surface transportation law since, effectively repealed the pay-as-you-go discipline without repealing the statute that announced it. The Byrd rule remains on the books at 26 U.S.C. Section 9503(d); it simply no longer governs, because Congress keeps the fund solvent by other means before the rule’s trigger can bite.

What changed the day the first general fund transfer arrived?

The September 15, 2008 transfer of 8.017 billion dollars under Public Law 110-318 changed the trust fund from a user-financed account into a hybrid. Dedicated receipts no longer set the ceiling on authorized spending, the Byrd rule’s discipline lost its practical force, and every later reauthorization could be written against revenues the fuel taxes did not produce.

One hundred forty-three point six billion dollars of general revenue

The cumulative scale of what followed is the single most important number in this comparison, and it should be reported with its source and period attached. Through the FAST Act, signed December 4, 2015, Congress had authorized a cumulative total of 143.6 billion dollars in transfers to the Highway Trust Fund, a figure the Congressional Research Service (CRS R45350.1) describes as roughly 22 percent of the fund’s outlays over the period. Of that 143.6 billion, 139.9 billion came from the general fund of the Treasury. The FAST Act itself authorized the last 70.0 billion dollars of the total, which means that nearly half of all the transfers ever authorized were authorized in the single most recent reauthorization, a measure of how the practice accelerated as the underlying gap widened. These are authorized transfers through December 2015, reported by CRS; they do not include anything after the article date, and the comparison that follows uses no post-2016 totals.

Consider what 22 percent means. More than one dollar in five spent from the Highway Trust Fund in this period did not come from highway users at all. It came from income taxpayers, from corporate taxpayers, from the same general revenue pool that finances everything else the federal government does. The trust fund’s defenders sometimes describe the transfers as loans or as temporary backstops, but the authorizations carry no repayment terms and the pattern shows no sign of reversing; they are general fund financing in substance, whatever the legislative language calls them. The program that was designed in 1956 so that drivers would pay for roads had become, by 2015, a program in which all taxpayers paid for more than a fifth of the roads, with the share rising in each successive law.

User-pays in rhetoric, general revenue in practice

Perhaps the most striking feature of this period is how little the language changed while the financing did. Authorizing statutes, committee reports, press releases, and floor speeches continued to describe surface transportation spending as trust fund spending, paid for by users, disciplined by the fuel tax. The appropriations still flowed through the trust fund’s accounts. The apportionment formulas still ran. A reader following only the rhetoric could be forgiven for believing the 1956 bargain intact. The numbers tell a different story, and the dissonance between the story and the numbers is itself a policy choice with consequences.

The defenders of the rhetorical continuity argue that it preserves something valuable: the political understanding that transportation is a user-financed enterprise, which in turn protects the program from being treated as just another discretionary claimant. If the public and the Congress believe drivers pay for roads, the argument runs, then road spending keeps a legitimacy that pure appropriations would lack, and the transfers are merely a technical bridge until the revenue system is repaired. The critics answer that the rhetoric has become a fiction that blocks repair. As long as everyone pretends the trust fund works, no one musters the political will to fix it, whether by raising the fuel tax, indexing it, or replacing it with a charge that actually tracks use. The fiction, in this view, is not harmless: it lets Congress authorize spending it will not pay for with dedicated revenue, while telling voters the users are paying. Both positions deserve their weight. The rhetoric does protect the program’s political standing, and it does obscure the financing reality.

The transit account’s weaker claim

Within the trust fund sits a complication that weakens the user-pays claim even on its own terms. The Mass Transit Account, created by the 1982 Great Compromise, receives 2.86 cents of every gallon’s federal tax, gasoline and diesel alike, while the Highway Account receives 15.44 cents of the gasoline tax and 21.44 cents of the diesel tax. The transit account is financed by fuel taxes paid by motorists, yet its beneficiaries are transit riders, who pay no fuel tax for the rides the account subsidizes. The Congressional Research Service (CRS R45350.1) reports that 85 to 90 percent of Highway Trust Fund receipts come from highway users, which means the transit account is, in economic substance, a transfer from drivers to transit riders administered through the trust fund’s books.

The history explains the arrangement without justifying it in user-pays terms. In 1982, the five-cent tax increase needed urban votes to pass, and urban legislators conditioned their support on dedicating a penny of the increase to mass transit. It was a legislative bargain, struck openly and defensibly: the coalition for better transportation included transit advocates, and the price of their support was a share of the revenue. The defenders of the arrangement argue that transit reduces highway congestion, which benefits drivers directly, so drivers are not really subsidizing strangers but buying themselves faster commutes; they add that the metropolitan economies transit serves generate the fuel tax revenue in the first place. The critics answer that these are spillover benefits of the kind every public program claims, and that a user fee which taxes one group to serve another has ceased to be a user fee in any meaningful sense. The truth sits between the positions: the transit account is a policy choice Congress made and has reaffirmed in every reauthorization since, and it is best understood as a political settlement rather than as an application of the benefit principle.

What the transfers bought

The 143.6 billion dollars in authorized transfers through the FAST Act did not vanish into abstraction; it bought something concrete, and any fair accounting must credit it. It bought continuity. State transportation departments plan in multi-year horizons, letting contracts, hiring engineers, and sequencing construction across seasons, and the apportionments that flow from the trust fund are the foundation of those plans. Had Congress allowed the Byrd rule to operate as written in 2008, apportionments would have been cut pro rata, projects would have stalled midstream, and the disruption would have cascaded through state budgets and construction employment in the middle of a severe recession. The transfers kept the money flowing, the projects moving, and the planning horizon intact. That is a genuine achievement, and the legislators who voted for the transfers understood themselves to be protecting a vital national program from a financing failure, not raiding the Treasury.

But the transfers also bought something less admirable: the indefinite postponement of a decision. Every transfer relieved the pressure that might otherwise have forced Congress to choose among raising the fuel tax, cutting spending, or designing a replacement revenue system. The Byrd rule was supposed to make that choice unavoidable; the transfers made it avoidable, year after year, and each year’s avoidance made the next year’s harder, because the gap kept widening as the three causes of erosion continued their work. The FAST Act’s authorization of the final 70.0 billion dollars of the cumulative total illustrates the dynamic: the largest single tranche of general fund support in the program’s history was attached to a five-year reauthorization that, like its predecessors, left the fuel tax rate untouched. The transfers were good stewardship of the construction program and poor stewardship of the financing system, and the comparison between trust fund and general fund models must hold both halves together: general revenue kept the roads funded, and in doing so, it finished the work the erosion had started, converting the trust fund from a user-financed enterprise into a general fund program in all but name.

Why the fiction persists

If the user-pays principle ended in practice in 2008, why does the rhetoric survive? The persistence of the fiction is not an accident; it serves interests across the political spectrum, which is why it endures. For the authorizing committees that write surface transportation law, the trust fund framing preserves jurisdiction and purpose: a dedicated program with dedicated revenue is their program, while a general fund appropriation belongs to the appropriators. For legislators who favor transportation spending, the fiction supplies legitimacy, because spending described as paid for by users is easier to defend than spending described as adding to the deficit. For legislators who oppose tax increases, the fiction supplies cover, because transfers described as keeping the trust fund solvent sound less like spending and more like stewardship. For the public, the fiction supplies simplicity: drivers believe their fuel taxes pay for roads, and that belief sustains a rough sense of fairness that the reality, more than a fifth of the money coming from general revenue, would complicate. The fiction persists because it is useful, and it will persist until the cost of maintaining it, in foregone reform and compounding gaps, exceeds its political value to those who maintain it. Naming the fiction is not cynicism; it is the precondition for an honest choice between rebuilding user financing and formalizing general funding.

The politics of the transfer votes

The transfers that kept the trust fund solvent were never anyone’s preferred policy, which is part of why they kept happening. No legislator campaigned on transferring general revenue to the Highway Trust Fund; no advocacy group demanded it as a first choice; no administration proposed it as a vision of transportation finance. The transfers were what remained when every preferred option failed: the tax increase nobody would vote for, the spending cut nobody would defend, the mileage charge nobody had built yet. That residual character shaped their politics. Because the transfers were attached to larger legislative vehicles, most often the reauthorization bills themselves, they were voted on as part of packages whose main attraction was the spending they preserved. A legislator who opposed general fund financing in principle could still vote for a reauthorization that contained it, because the alternative was allowing apportionments to collapse. The bipartisan character of the transfers deserves emphasis, since transportation financing is sometimes misdescribed as a partisan fight. The first transfer, under P.L. 110-318 in September 2008, was signed by a Republican president; the largest, under the FAST Act in December 2015, was signed by a Democratic president; and the Congresses in between divided control of the chambers without interrupting the pattern.

Deferred maintenance is debt by another name

Beneath the financing debate lies a physical reality that neither user fees nor general revenue can wish away: the Interstate System built under the 1956 Act was aging, and aging infrastructure does not negotiate. Pavement deteriorates on a schedule set by traffic and weather, bridges corrode on a schedule set by chemistry, and the cost of reconstruction rises the longer it is postponed, because a road that could have been resurfaced must instead be rebuilt. This physical compounding mirrors the fiscal compounding described earlier, and it sets a floor under the spending side of the comparison. Whatever financing model Congress chooses, the network requires a certain level of investment to remain safe and functional; spending less than that level does not save money but borrows against the future at punitive interest, in the form of higher reconstruction costs, detours, and the economic drag of unreliable transportation.

The financing models differ in how they respond to this reality, and the difference cuts both ways. The trust fund model, at its best, tied investment to a revenue stream that grew with use: more driving meant more fuel tax, which meant more money for the wear the driving caused. That feedback loop was the model’s great virtue, aligning the financing with the physical need automatically. But the loop depended on the proxy holding, and once efficiency and electrification broke it, the model began to underfund precisely as needs were rising, because the revenue fell while the pavement aged. General fund financing has no such feedback loop, which its critics cite as a fatal flaw: nothing in the appropriations process measures road wear or translates it into dollars. Its defenders answer that the appropriations process can consider needs directly, funding reconstruction because engineers say it is needed rather than because a tax formula happens to produce the money, and that this is a more rational basis for investment than the accidents of fuel consumption. Both claims have merit, and both point to the same underlying truth: the physical needs of the network are independent of the financing debate, and any financing system will be judged, in the end, by whether the bridges stand. The user-pays principle is a theory of fairness and discipline; it is not a theory of concrete. When the two conflict, the concrete wins, which is why the transfers, for all their damage to the principle, kept the program alive.

The balance sheet as political theater

The trust fund’s accounts have continued to present an orderly appearance throughout the transfer era, and that appearance is itself part of the story. Each year, dedicated fuel tax receipts are credited to the fund, authorized spending is charged against it, and the general fund transfers arrive to cover the difference, so the balance never goes negative and the apportionments never stop. A reader who looked only at the fund’s solvency, the fact that it meets its obligations, would conclude the system works. A reader who looked at the sources of the money would conclude something else: that the solvency is purchased annually with general revenue, and that the fund’s balance sheet records a political decision rather than an economic equilibrium. The distinction matters because balance sheets persuade. The orderly accounts reassure state planners, calm bond markets for transportation debt, and give legislators a document to point to when they claim the program is sound. The theater is not dishonest in the narrow sense; every transfer is authorized in law and recorded in the accounts. But it is theater in the broader sense, because the performance of solvency substitutes for the substance of it, and the audience is encouraged not to look behind the curtain.

The alternative presentation would be rawer and more informative: a trust fund account showing dedicated receipts covering roughly four-fifths of outlays, with the remainder labeled plainly as general fund support. Nothing in law prevents Congress from presenting the accounts that way, and some analysts have urged exactly that candor, arguing that honest labeling would force the financing debate the transfers keep deferring. The resistance to candor comes from the same coalition that sustains the fiction described earlier: authorizers who want their program to look self-sufficient, legislators who want their votes to look costless, and a public that prefers the simplicity of drivers paying for roads. The balance sheet will change only when the politics change, which means the accounting follows the power rather than disciplining it. That inversion, the servant becoming the master, is the final consequence of the transfer era: the trust fund’s books, designed in 1956 to constrain spending to revenue, function in the transfer era to conceal the gap between them.

What twenty-two percent means for democratic accountability

The figure that 143.6 billion dollars in authorized transfers represented roughly 22 percent of Highway Trust Fund outlays through the FAST Act (CRS R45350.1) is usually quoted as a measure of fiscal imbalance. It is also a measure of something less discussed: the gap between what voters are told and what is true, and what that gap does to democratic accountability. A voter who believes drivers pay for roads will evaluate transportation policy on that understanding, supporting or opposing tax and spending proposals against a factual premise that has not held since 2008. When more than a fifth of the money comes from general revenue, the voter’s mental model of the program is wrong in a way that systematically distorts their judgment: they will underestimate what the program costs the Treasury, overestimate the discipline the fuel tax provides, and misjudge who bears the burden of every proposal to change it. Accountability requires a visible connection between decisions and consequences, and the transfer era severs that connection twice over. First, it severs the connection between the decision to authorize spending and the decision to finance it, because the spending is debated in reauthorizations while the financing arrives through transfers that attract little attention. A legislator can vote for generous authorizations, vote against tax increases, and never face the contradiction, because the transfers reconcile the two votes out of public view. Second, it severs the connection between the taxes voters pay and the services they receive, because income taxpayers financing a fifth of the road network have no way of knowing they are doing so. The fuel tax is visible at the pump; the general fund transfer is invisible in the withholding. The critics of the presentation argue that this particular fiction is load-bearing: it sustains the political coalition for transportation spending, and removing it might endanger the program. A coalition sustained by a misunderstanding is fragile in a different way, because it cannot survive the moment the misunderstanding is corrected. Whether the true story leads the reader toward rebuilding user fees or toward formalizing general funding is, once more, the deciding factor.

What keeps the Highway Trust Fund solvent after 2008?

General fund transfers keep it solvent. Beginning with an 8.017 billion dollar transfer on September 15, 2008, under P.L. 110-318, Congress has authorized transfers every year since, reaching a cumulative 143.6 billion dollars through the FAST Act of December 2015, of which 139.9 billion came from general Treasury revenue (CRS R45350.1).

Four Ways to Pay for the Road Network

The comparison so far has described a design, its erosion, and its quiet replacement. This section turns to the alternatives, because the point of diagnosing the trust fund’s condition is to ask what should take its place, or what could repair it. Four broad approaches present themselves, each with a genuine case and a genuine cost, and each best understood as an answer to a different version of the financing question. They are presented here with equal care, because transportation finance is contested terrain: tax policy and the allocation of spending between modes are matters on which reasonable people, and reasonable legislators, disagree. The positions below are attributed to their holders rather than endorsed, and the reader is invited to weigh them against the deciding factor named at the outset.

Index the fuel tax

The simplest reform on offer is to index the per-gallon rate to inflation, so that the tax holds its real value without requiring Congress to vote for increases. The mechanism is familiar from other parts of the tax code, where brackets and thresholds adjust automatically, and its appeal is administrative simplicity: no new collection system, no new bureaucracy, no change in what drivers experience beyond the price at the pump keeping pace with the general price level. Supporters argue that indexing preserves the user-pays link for the great majority of travel that still burns taxed fuel, that it restores the 1956 bargain’s self-discipline at minimal disruption, and that it is the only reform that can be implemented without building an entirely new revenue apparatus. They point to the construction-cost figures as the measure of what has been lost and argue that indexing to a construction-cost index, rather than the general Consumer Price Index, would tie the tax to the actual cost of the program it finances.

The critics of indexing make three objections, and each has force. First, as the earlier section established, indexing addresses only the inflation cause; it does nothing about efficiency or electrification, so it stabilizes a declining base rather than restoring a growing one. Second, the fuel tax is regressive in its incidence: lower-income drivers spend a larger share of their income on fuel, drive older and less efficient vehicles, and have the least ability to switch to electric alternatives, so an indexed tax that rises automatically would steadily increase a burden that already falls hardest on those least able to bear it. Supporters answer that the regressivity can be offset through the income tax or through rebates, but the offset has to be enacted and sustained, which returns the politics to Congress. Third, indexing by automatic formula removes the periodic moment of democratic accountability that a voted increase provides; critics of automatic government growth see that as a vice, while supporters see it as the virtue that makes the reform work. The indexing debate, in short, is a debate about whether the fuel tax deserves to be saved, and the positions divide along the same line as the deciding factor itself.

The fuel tax road, examined closely

The fuel tax road has administrative virtues that none of the alternatives match. Collection occurs at roughly a thousand terminal racks rather than at hundreds of millions of vehicles, which is why the compliance cost per dollar of revenue is low and why the tax reaches essentially every gallon of taxable fuel sold. The states, which administer their own fuel taxes through parallel systems, have a century of combined experience with the mechanism, and the federal-state coordination is well established. The rate structure even contains a rough weight-distance logic: diesel, burned disproportionately by heavy trucks, is taxed at 24.4 cents against gasoline’s 18.4, which partially compensates for the greater pavement damage heavy vehicles cause. The defenders’ case, made by transportation economists and user-fee advocates, is that these virtues compound: a tax that is cheap to collect, hard to evade, and already embedded in every related administrative system should not be discarded because its base is eroding, but repaired. The critics’ case, made by equity advocates and by analysts focused on electrification, is that the repair cannot hold: the base erosion is structural, the burden falls hardest on drivers of older vehicles, and the political system has demonstrated, through twenty-three years of inaction, that it will not maintain the rate even if it raises it once.

Charge by the mile

The second alternative abandons the gallon as the measure of use and replaces it with the mile. A vehicle-miles-traveled fee, usually described as a road usage charge, would assess drivers a fixed amount per mile traveled, restoring the link between use and payment that efficiency and electrification broke. Because the charge is levied on distance rather than fuel, it treats a gasoline sedan, a hybrid, and an electric vehicle identically for the miles each travels, which answers the fairness objection that electrification raises against the fuel tax. Supporters present it as the true heir of the user-pays principle: if the principle is that users should pay in proportion to their use, then measuring use directly is more faithful to the principle than measuring it through the increasingly distorted proxy of fuel consumption.

The idea had moved from theory to field testing well before the article date. Oregon launched its OReGO program on July 1, 2015, under S.B. 810 of 2013, enrolling up to 5,000 volunteers who paid 1.5 cents per mile and received a credit for fuel taxes paid, so that the pilot tested the charge as a replacement rather than an addition. California authorized its own road charge pilot in 2014 under S.B. 1077 and launched it on July 1, 2016, with more than 5,000 volunteers; the pilot was underway at the article date. At the federal level, the FAST Act created the Surface Transportation System Funding Alternatives program in Section 6020, authorizing 95 million dollars over five years for state pilot grants to test alternatives to the fuel tax; in August 2016, the program was authorized and its funds available, but it was not yet operating. These pilots were designed to answer practical questions: how to measure miles without invasive tracking, how to collect the charge at reasonable cost, how to handle interstate travel, and how to protect privacy.

The objections are substantial and must be given their full weight. Privacy advocates argue that any system capable of measuring miles is capable of tracking movement, and that the assurances of anonymization and data minimization, however sincerely offered, ask drivers to trust institutions with a detailed record of where they go. Collection costs are a second concern: the fuel tax is collected cheaply from a small number of refiners and importers high in the distribution chain, while a mileage charge must be collected from hundreds of millions of individual vehicles, which implies administrative costs an order of magnitude higher, at least in the early years. Rural drivers raise a third objection, arguing that they drive longer distances by necessity rather than choice, so a per-mile charge would fall more heavily on them than a fuel tax does; supporters answer that rural drivers also tend to drive less efficient vehicles, so the fuel tax already burdens them, and that the comparison is closer than it first appears. Finally, there is the transition problem: running two systems during a long phase-in, crediting fuel taxes against mileage charges, and preventing double payment all add complexity that the pilots were only beginning to explore. The mileage charge is the most intellectually coherent answer to the erosion problem, and it is also the farthest from being ready.

The debate over mileage charging turns on three clusters of objections, each of which the pilots were designed to inform. The privacy objection holds that reporting miles requires tracking movement, which the fuel tax avoids entirely, and that any system capable of varying charges by time or place is a system capable of surveillance. The pilot designs answered with options: less precise reporting methods that reveal total miles without location, more precise methods for those who want credit for off-road or out-of-state miles, and a choice left to the driver. Whether those answers satisfy the objection is contested, and the contest is ultimately about trust in the administering institutions rather than about technology. The administrative-cost objection holds that metering hundreds of millions of vehicles costs far more per dollar of revenue than taxing fuel at a thousand terminals, and that the transition costs, devices, accounts, enforcement, would consume years of revenue. The pilot data on costs were preliminary at the article date, and the honest statement is that full-scale costs remain uncertain. The acceptance objection holds that drivers understand and tolerate the fuel tax, invisible as it is within the pump price, while a separate per-mile bill would make the cost of road use salient in a way that invites resistance. The defenders answer that salience is a virtue: a visible price for road use is precisely what the user-pays principle requires. Each objection is serious, and the article presents them as the participants state them because the mileage charge stands or falls on whether these three concerns can be answered at scale.

Open the toll gates

The third alternative is the oldest form of road user fee in existence: the toll. Tolling charges drivers directly for the use of specific facilities, matching payment to benefit with a precision no fuel tax can achieve, and it has the political advantage of being voluntary in a way taxes are not, since drivers can choose alternate routes. Federal law, however, restricts tolling on the facilities the trust fund built. The general prohibition on tolling federally funded highways is found at 23 U.S.C. Section 301; the exceptions are collected nearby, principally in Section 129. The exceptions that matter are narrow. Bridges and tunnels on the Interstate System may be tolled under 23 U.S.C. Section 129(a). High-occupancy-vehicle lanes may be converted to high-occupancy toll lanes, known as HOT lanes, under 23 U.S.C. Section 166, which allows solo drivers to buy into carpool lanes at a market price. And the Interstate System Reconstruction and Rehabilitation Pilot Program, created by Section 1216(b) of the Transportation Equity Act for the 21st Century (P.L. 105-178), permits tolling on existing Interstate facilities for reconstruction purposes, but it is limited to exactly three facilities in three different states, was amended by Section 1411(c) of the FAST Act, and in 2012 all three slots were conditionally reserved, for Interstate 70 in Missouri, Interstate 95 in Virginia, and Interstate 95 in North Carolina, with no state having completed a conversion through the program. The older construction toll pilot authority expired on August 9, 2015, and is omitted from this account.

Supporters of broader tolling authority argue that the restriction is an anachronism: the Interstate System the 1956 Act built had aged and needs reconstruction, tolls would let the states that need the revenue raise it from the drivers who benefit, and modern electronic tolling has eliminated the booths and delays that once made tolls unpopular. They add that tolling is the purest available form of user financing, since it charges for actual facility use rather than for the rough proxy of fuel. The opponents answer with the double-payment objection: drivers already paid for these roads through decades of fuel taxes, and tolling them on top charges twice for the same pavement. They add concerns about diversion, as tolled traffic spills onto local streets not built for it, and about equity, since tolls on commuter corridors fall on workers with the least flexibility in when and where they drive. Supporters respond that the double-payment objection proves too much, since every generation pays for infrastructure its predecessors financed, and that congestion pricing on HOT lanes has shown tolls can manage demand as well as raise revenue. The tolling debate is, at bottom, a debate about whether the Interstate System should remain a free good financed broadly or become a set of priced facilities financed directly, and it divides, once again, along the line of the deciding factor.

Fund roads from general revenue

The fourth alternative is to stop pretending and formalize what has already happened: finance surface transportation from general revenue, openly and permanently. The case for this approach begins with honesty. Since 2008, general revenue has covered the gap every year; cumulative authorized transfers reached 143.6 billion dollars through the FAST Act. Formalizing the arrangement would end the fiction, allow Congress to debate transportation spending on its merits alongside other priorities, and free the program from dependence on a revenue instrument that technology is rendering obsolete. Supporters add two substantive arguments. First, the benefits of the road network are genuinely broad: freight, emergency services, and the commerce the network enables touch every household, so financing through broad-based taxes matches payment to benefit better than a levy on drivers alone. Second, general fund financing is stable in a way the fuel tax no longer is; income tax receipts grow with the economy, while fuel tax receipts shrink with efficiency, so the general fund ties transportation to a growing rather than a declining base.

The case against is the case for discipline, and it is the strongest argument the trust fund’s defenders have. Once transportation spending competes in the annual appropriations process without a dedicated revenue stream, it becomes one claimant among many, subject to deficit politics, to the priorities of appropriators rather than authorizers, and to the risk that a future Congress undervalues maintenance and construction in favor of more immediately visible spending. The history of programs that lost dedicated funding is not encouraging to those who value stable infrastructure investment, and the defenders of user financing argue that the fuel tax, for all its flaws, created a constituency for transportation spending, the highway users themselves, whose payments legitimized the program. Remove the payment, they warn, and the legitimacy follows, leaving the road network to compete with every other claim on the Treasury. Supporters of general funding answer that legitimacy should come from the program’s merits rather than from an accounting fiction, and that a Congress which wants good roads will fund them. Which of these predictions the reader finds more plausible is, once more, the deciding factor wearing different clothes.

The general fund road as budget-process reality

The general fund road should be evaluated as the budget-process reality it is rather than as an abstraction. Under permanent general fund financing, transportation spending would be set through the same process that funds defense, health, and other domestic programs: authorizations would establish programs, appropriations would provide budget authority, and the spending would compete each year with every other claim on the Treasury. The defenders’ case has three strands. First, the benefit strand: roads serve the whole economy, including freight, emergency services, and commerce that depends on connectivity, so broad-based funding matches the breadth of the benefit better than a tax on motorists alone. Second, the honesty strand: the transfers since 2008 have already made the trust fund a hybrid, so ratifying general fund support ends the pretense and subjects transportation to straightforward budget tradeoffs. Third, the equity strand: replacing the regressive fuel tax with progressive broad-based revenue shifts the burden toward ability to pay. The critics’ case also has three strands. First, the discipline strand: without the Byrd rule or an equivalent, nothing constrains transportation spending except the general budget caps, and transportation would compete poorly against programs with stronger constituencies. Second, the distribution strand: appropriations politics would allocate spending by political weight rather than by system need, a concern the formula apportionments of the trust fund era were designed to mitigate. Third, the volatility strand: transportation would become hostage to deficit politics and to the annual cycle, making the multi-year project planning the program requires more difficult. The record since 2008 offers evidence for both sides: the program continued to function under transfer support, which the defenders cite, but the financing was repeatedly uncertain until each transfer was enacted, which the critics cite.

Who bears the burden: the fairness ledger

Every financing alternative distributes its costs differently, and the fairness of each distribution is contested, so this comparison presents the competing claims without adjudicating them. Start with the fuel tax, indexed or not. Its burden falls on drivers in proportion to the fuel they burn, which means it falls most heavily, as a share of income, on lower-income households: they spend more of their budgets on fuel, drive older and less efficient vehicles, and have the least capacity to switch to efficient or electric alternatives. Defenders of the tax answer that it is a payment for a service used, like a fare, and that fares are not usually judged by their progressivity; they add that the regressivity can be offset through the income tax, though the offset must be enacted and sustained. Critics answer that a fare for an essential network is different from a fare for an optional service, because workers who must drive to reach their jobs cannot opt out, and that the offset is a promise, not a policy.

The mileage charge reshuffles the burden in ways that please some critics and alarm others. By charging per mile rather than per gallon, it removes the penalty on drivers of inefficient vehicles, which helps lower-income drivers of older cars, but it introduces a new penalty on drivers who travel long distances, which hurts rural households that drive farther by necessity. Supporters argue the rural objection is overstated, because rural drivers already pay more under the fuel tax for the same reason, and because the charge can be varied by vehicle weight or location to reflect actual costs. Critics answer that variation by location requires precisely the tracking that privacy advocates fear, which trades one fairness problem for another. The charge also raises a question the fuel tax never faced: whether miles driven off the public network, on private roads or out of state, should count, and who decides.

Tolling concentrates its burden most narrowly of all, on the users of specific priced facilities. That precision is its fairness argument: those who benefit pay, and those who do not use the facility pay nothing. But the concentration is also its fairness objection, because tolled corridors are often the only practical routes for commuters who cannot shift their travel times, which means the toll functions as a levy on workers with the least flexibility. The diversion effect compounds the concern: when priced drivers spill onto unpriced local streets, the burden shifts to neighborhoods that never consented to carry the traffic. Supporters answer that congestion pricing gives drivers a choice the fuel tax never offered, the choice to travel at a cheaper time or by another mode, and that choice is itself a form of fairness. Critics answer that the choice is real only for those with flexible schedules, which is another way of saying it is real only for the affluent.

General fund financing spreads the burden most broadly, across all income taxpayers, which its supporters present as the fairest distribution available: the progressive rate structure of the income tax means higher earners pay more, and the broad base means no single group bears a disproportionate share. Its critics make two fairness objections of their own. First, broad financing means non-drivers pay for roads they may rarely use, which the user-pays advocates consider unjust regardless of progressivity. Second, general fund financing subjects transportation to the distributional politics of the whole budget, where the well-organized routinely outmaneuver the diffuse, so the theoretical fairness of the tax base may not survive contact with the appropriations process. The fairness ledger, in the end, does not crown a winner. Each alternative is fair by one measure and unfair by another, and the reader’s judgment will depend on which measure of fairness they take to be decisive, which is the deciding factor once more, wearing the clothes of equity rather than efficiency.

Would indexing the gas tax restore the trust fund?

Indexing would stop the inflation bleed by holding the rate’s real value steady, but it would not restore the gallons lost to efficiency or capture miles traveled by electric vehicles. It is a real improvement that addresses one of three causes, which is why its supporters call it necessary and its critics call it insufficient.

The hybrid reality

The four alternatives are presented above as distinct choices, but the actual legislation of this period was already a hybrid, and the most plausible futures are hybrids too. The FAST Act of 2015 illustrates the pattern: it authorized the largest single tranche of general fund transfers in the program’s history, 70.0 billion dollars of the 143.6 billion cumulative total, while simultaneously creating the Surface Transportation System Funding Alternatives pilot program, with 95 million dollars over five years for state experiments in replacing the fuel tax, and amending the Interstate tolling pilot program through Section 1411(c). In a single statute, Congress funded the program from general revenue, invested in the research that might one day restore user financing through a mileage charge, and adjusted the legal framework for tolling. That is not incoherence; it is a legislature buying time and options simultaneously, keeping the program alive with the money available while probing the alternatives that might eventually replace the fuel tax. The honest version of the hybrid is one in which each element is named, measured, and defended on its merits. The dishonest version is the status quo, in which general revenue does more than a fifth of the work while the rhetoric claims the users pay for all of it.

What rebuilding would actually require

The verdict of this comparison offers two honest paths, rebuild user financing or formalize general funding, and it is worth spelling out what each would actually require, because the difficulty of each explains why Congress has chosen neither. Rebuilding a user-pays system would have to begin with candor about the fuel tax: indexing the rate to hold its real value, whether against construction costs or the general price level, would stop the inflation bleed, but as established earlier, it would not restore the gallons lost to efficiency or capture electric miles. So indexing would need a companion, and the only companion capable of restoring the use-to-payment link is a mileage-based charge. Building that charge into a national system would require resolving the questions the pilots were only beginning to answer in August 2016: how to measure miles without invasive tracking, how to collect from hundreds of millions of vehicles at acceptable cost, how to credit fuel taxes during a long transition, how to treat interstate travel, and how to protect rural drivers from disproportionate burden. Oregon’s OReGO program and California’s pilot, each with around five thousand volunteers, were proofs of concept, not templates for national administration, and the federal pilot program authorized by the FAST Act at 95 million dollars over five years was funded but not yet operating. A realistic rebuilding sequence would run a decade or more: index the fuel tax to stabilize the base, expand tolling flexibility where states want it, run mileage-charge pilots at growing scale, and transition the base from gallons to miles as the technology and the public acceptance mature. Each step is feasible; the sequence as a whole requires a political commitment across multiple Congresses, which is precisely what the transfer era showed to be scarce.

Formalizing general funding would be legislatively simpler and politically harder in a different way. The legislative mechanics are straightforward: authorize transportation spending from general revenue, repeal or repurpose the fuel taxes, and let the appropriations process allocate the money. The difficulty is not mechanical but constitutional in the small-c sense: it would require Congress to take explicit, recurring responsibility for the level of transportation investment, defending it each year against competing priorities without the shelter of a dedicated fund. The defenders of this path argue that explicit responsibility is healthier than the current arrangement, in which Congress authorizes spending it will not finance and calls the result a trust fund. The critics answer that the appropriations process has a long record of undervaluing maintenance relative to new construction, of favoring the visible over the essential, and that a transportation program stripped of its dedicated constituency would fare worse, not better, in that environment. Formalizing would also force an honest reckoning with the Mass Transit Account’s 2.86 cents, since a general fund program would have to decide openly how much transit to support rather than inheriting the 1982 compromise by default. Either path demands something Congress has avoided since 1993: a deliberate, public choice about what transportation is worth and who should pay. The transfers are what happens when that choice is postponed. They will continue until the choice is made, and the gap they cover will keep growing until then.

The Deciding Factor

This comparison began with a verdict and a named deciding factor, and it ends by returning to both with the evidence in hand. The verdict is that the user-pays principle already ended. Since 2008, the Highway Trust Fund has been unable to support the surface transportation spending Congress authorizes from its own dedicated revenues; every year since, general fund transfers have covered the gap; and the cumulative authorized total reached 143.6 billion dollars through the FAST Act of December 2015, roughly 22 percent of outlays, of which 139.9 billion came from general Treasury revenue (CRS R45350.1). Those are not projections or warnings. They are the recorded financing of the program across eight years. A trust fund that cannot pay for its program from its dedicated taxes, and is kept solvent by transfers from general revenue authorized in every successive law, is a general fund program that has kept the old name. The comparison between the Highway Trust Fund and general fund financing is therefore not a choice between two futures. It is a choice about whether to acknowledge the present.

Applied to 2016, the three-part test for the user-pays principle gives a precise answer rather than a slogan. The first condition, that revenue comes from users, still largely holds on the revenue side: 85 to 90 percent of trust fund receipts come from highway users, per the Congressional Research Service in R45350.1. The second condition, that revenue stays dedicated, has been weakened twice: first by the 1982 bargain that assigned a share of motorist-paid taxes to transit riders, and then by the general fund transfers that made broad-based revenue a structural support of the account. The third condition, that revenue disciplines spending, has failed outright: authorizations have exceeded dedicated receipts every year since 2008, and the automatic cut the Byrd rule contemplates has been preempted by transfers rather than allowed to operate. A principle that requires all three conditions, and that retains only the first in weakened form, has already ended. What remains is the decision about what replaces it, and that decision belongs to the reader.

The deciding factor, named at the outset, is where the reader believes discipline should come from: whether transportation spending should be constrained by what users pay, or determined by what the network needs. The reader’s preference follows from which risk they consider greater. If the greater risk is fiscal indiscipline, the slow conversion of transportation into one more unanchored claim on the Treasury, then the reader will favor rebuilding a user-pays system, whether through indexing, a mileage charge, expanded tolling, or some acknowledged combination, because each of those ties spending to a visible payment by users and restores the constraint the 1956 design intended. If the greater risk is a starved network, the slow decay of pavement and bridges under a financing system that technology has outgrown, then the reader will favor formalizing general fund financing, because it ties the program to the broadest and most durable revenue base and frees investment decisions from the accidents of fuel consumption. There is no neutral position between these risks, because every financing choice insures against one of them at the expense of the other. The honest debate is not about whether roads matter; everyone in this debate agrees they do. It is about which failure mode the country should organize itself to avoid.

Three cautions belong with the verdict, because a verdict stated too simply becomes its own kind of fiction. First, rebuilding user financing is harder than its advocates sometimes suggest. The fuel tax cannot be restored to its 1956 function by a rate increase alone, because efficiency and electrification have broken the proxy on which it depended; a mileage charge could restore the function, but the pilots in Oregon and California were experiments, not systems, and the federal pilot program authorized by the FAST Act was not yet operating in August 2016. Rebuilding means building something new, not repairing something old. Second, formalizing general funding is riskier than its advocates sometimes suggest. The appropriations process has its own disciplines, but they are the disciplines of coalition politics rather than of measured use, and a transportation program that must compete annually against every other national priority may discover that its broad benefits are less persuasive in a markup session than a dedicated revenue stream was. Formalizing means trusting the political process with the network’s future, without the guardrails the trust fund was meant to provide. Third, the status quo is not a stable middle ground but a slow decision by default. Each year of transfers without reform deepens the gap, entrenches the fiction, and makes the eventual choice harder, because the constituency for pretending grows while the revenue base shrinks. The one option this comparison cannot recommend is continuing to describe a general fund program in the language of user financing while authorizing spending the dedicated taxes cannot support.

The thread that runs through this comparison is the one the brief named: fiscal design decisions made in 1956 were still constraining policy decades later. The dedication of fuel taxes to highways, the Byrd rule’s pay-as-you-go discipline, the choice of the gallon as the measure of use, the creation of a Mass Transit Account as the price of a tax increase, all of these were decisions taken for the world of 1956 and 1982, and all of them shaped what was possible in 2016. The fuel tax worked brilliantly while gallons tracked miles; it failed when they diverged. The Byrd rule worked brilliantly while Congress was willing to choose between taxes and spending; it failed when Congress chose transfers instead. The lesson is not that the designers were wrong. It is that financing mechanisms are not neutral plumbing. They allocate power: between users and taxpayers, between authorizers and appropriators, between one Congress and its successors. The trust fund allocated power to users and to the discipline of the pump. General funding allocates power to the appropriations process and to the politics of the moment. Choosing between them is choosing who decides what the country’s transportation network is worth, and that choice, made in 1956 and unmade gradually ever since, was due to be made again, this time with open eyes. The next surface transportation reauthorization will face the same arithmetic this article has traced: a nominal rate frozen since 1993, three compounding causes of erosion, and a gap that general revenue has covered every year since 2008. The legislators who write that law will have the same four alternatives, the same fairness disputes, and the same deciding factor. What they will not have is the excuse of surprise. The numbers have been public for years, the pilots have reported their early lessons, and the cumulative transfers are a matter of record. A choice made with open eyes is the most this comparison can ask for, and it is also the least the country should accept. For readers tracing how this financing debate fits into the broader development of federal transportation law, the US legislation study guide places the highway program alongside the other great legislative enterprises of the modern era. Those who want to review the statutes and financing mechanics studied here can use the companion study notebook, and readers seeking structured practice on the constitutional and procedural questions can use the civics study aid.

The Revenue Gap Table

Period	Nominal federal gasoline tax	Real value of the tax	Sources of erosion active	Annual gap vs authorized spending	Cumulative general-fund transfers authorized	Source
1956 to 1982	3 cents per gallon from June 29, 1956	Full initial purchasing power (baseline)	None; gallons tracked miles closely	None; pay-as-you-go held	0 dollars	Highway Revenue Act of 1956, P.L. 84-627; 70 Stat. 387
1983 to 1993	9 cents per gallon from April 1, 1983 (raised from 4 cents)	Eroding; nominal rate fixed while prices rose	Inflation (rate not indexed)	None sustained; balances covered shortfalls	0 dollars	STAA, P.L. 97-424, Title V, Sec. 511; 1 cent of the 5-cent increase to the new Mass Transit Account
1993 to 2007	18.4 cents per gallon from October 1, 1993 (raised 4.3 cents)	Approximately 73 percent of purchasing power lost vs highway construction-cost inflation (CRS R48472); about 37 to 40 percent lost vs CPI (18.3 cents equated to roughly 29 to 31.7 cents inflation-adjusted)	Inflation plus rising fuel economy (new light-duty real-world economy 20.9 mpg in MY1993 toward 24.8 mpg in MY2015, up 18.7 percent; up 28 percent since MY2004, EPA-420-S-16-001)	Emerging; dedicated receipts fell short of authorized levels as balances drew down	0 dollars	OBRA 1993, P.L. 103-66; CRS R48472; EPA Light-Duty Trends
2008	18.4 cents gasoline; 24.4 cents diesel	As above; erosion continuing	Inflation, efficiency, plus electrification (electricity not a taxable fuel under 26 U.S.C. Sec. 4081; fuel taxes do not account for miles traveled by EVs, CRS R48472)	Gap opens; dedicated revenues below authorized spending every year from this point	8.017 billion dollars (first transfer, September 15, 2008)	P.L. 110-318; CRS R45350.1
2009 through December 2015	18.4 cents gasoline; 24.4 cents diesel (trust fund share 18.3 and 24.3 cents; 0.1 cent to LUST; gasoline split 15.44 cents Highway Account and 2.86 cents Mass Transit Account; diesel split 21.44 and 2.86 cents)	Continued erosion on both inflation measures	All three causes compounding	Gap every year; Congress authorizes transfers rather than raising the rate or cutting apportionments	143.6 billion dollars total through the FAST Act (December 4, 2015), roughly 22 percent of outlays; 139.9 billion of it from the general fund; the FAST Act authorized the last 70.0 billion	CRS R45350.1; FAST Act, P.L. 114-94
In August 2016	18.4 cents gasoline; 24.4 cents diesel; rate unchanged since October 1, 1993	See 1993 to 2007 row; no recovery	All three causes compounding; Oregon OReGO pilot operating since July 1, 2015 (up to 5,000 volunteers, 1.5 cents per mile with fuel-tax credit, S.B. 810 of 2013); California pilot launched July 1, 2016 (5,000 plus volunteers, S.B. 1077 of 2014); federal STSFA pilot authorized (FAST Act Sec. 6020, 95 million dollars over five years) but not yet operating	Gap continues every year	143.6 billion dollars authorized through the FAST Act; no post-2016 figures used	CRS R45350.1; CRS R48472; EPA-420-S-16-001

Frequently Asked Questions

Q: Why is the Highway Trust Fund running out of money?

Three compounding causes eroded the fuel tax revenue that finances the fund. First, the federal gasoline tax has been fixed at 18.4 cents per gallon since October 1, 1993, and inflation cut its real value: the Congressional Research Service found it lost approximately 73 percent of its purchasing power against highway construction-cost inflation (CRS R48472), and about 37 to 40 percent against the Consumer Price Index. Second, vehicles became substantially more fuel efficient, with new light-duty real-world economy rising from 20.9 miles per gallon in model year 1993 to 24.8 in model year 2015, so fewer gallons, and less tax, are collected per mile. Third, electric vehicles pay no federal fuel tax at all, because electricity is not a taxable fuel under 26 U.S.C. Section 4081. Meanwhile Congress kept authorizing spending through five reauthorizations. Since 2008, dedicated revenues have fallen short of authorized spending every year.

Q: When was the federal gas tax last raised for the Highway Trust Fund?

The federal gasoline tax was last raised on October 1, 1993, by the Omnibus Budget Reconciliation Act of 1993 (P.L. 103-66), which added 4.3 cents per gallon and brought the rate to 18.4 cents, where it still stood in August 2016. The diesel rate set by the same law was 24.4 cents per gallon. The previous increase had come a decade earlier: the Surface Transportation Assistance Act of 1982 (P.L. 97-424), signed January 6, 1983, raised the rate from 4 cents to 9 cents per gallon effective April 1, 1983, dedicating one penny of the five-cent increase to the newly created Mass Transit Account in the arrangement remembered as the Great Compromise. No increase has been enacted in the twenty-three years since 1993, across five surface transportation reauthorizations, which is why inflation had so long to erode the rate’s real value.

Q: Do electric vehicles pay into the Highway Trust Fund?

No. The federal fuel taxes that finance the Highway Trust Fund apply to taxable fuels, which under 26 U.S.C. Section 4081 means gasoline, diesel fuel, and kerosene. Electricity is not a taxable fuel, so a vehicle that runs entirely on electricity contributes nothing to the fund no matter how many miles it travels. The Congressional Research Service states the point directly (CRS R48472): the fuel taxes do not account for miles traveled by electric vehicles. Through August 2016 the electric share of the vehicle fleet was small enough that the revenue effect was modest, but the structural problem is permanent under current law: a tax on gallons cannot capture travel by vehicles that use no gallons. This is one reason analysts consider the fuel tax a declining revenue base over the long term, and it is the problem that mileage-based charges are designed to solve, since a per-mile fee treats electric and gasoline vehicles identically.

Q: Should roads be paid for by users or by everyone through the Highway Trust Fund?

That is the central contested question of this comparison, and it divides along the deciding factor named in the article: whether transportation spending should be disciplined by what users pay or determined by what the network needs. The user-pays position holds that drivers benefit most directly from roads, that the fuel tax (or a mileage charge) links payment to use, and that dedicated user revenue protects transportation spending from the politics of annual appropriations. The general-funding position holds that the road network’s benefits spill far beyond drivers, to freight, emergency services, and the commerce the network enables, so broad-based taxes match payment to benefit more fairly, and that income taxes grow with the economy while fuel taxes shrink with efficiency. In practice, the system has been a hybrid since 2008, with general revenue covering the gap every year. Readers must weigh which risk they consider greater: a network starved by a frozen tax, or a budget freed from the discipline of the pump.

Q: How much has been transferred to the Highway Trust Fund from general revenue?

Through the FAST Act, signed December 4, 2015, Congress had authorized a cumulative total of 143.6 billion dollars in transfers to the Highway Trust Fund, a figure the Congressional Research Service (CRS R45350.1) describes as roughly 22 percent of the fund’s outlays over the period. Of that total, 139.9 billion dollars came from the general fund of the Treasury. The FAST Act itself authorized the last 70.0 billion dollars, meaning nearly half of all transfers ever authorized came in the single most recent reauthorization. The first transfer was 8.017 billion dollars on September 15, 2008, under P.L. 110-318, presented at the time as an extraordinary bridge across a temporary shortfall. The shortfall proved permanent: every year since 2008 has seen a gap between dedicated revenues and authorized spending. These figures are authorized transfers through December 2015 as reported by CRS; this article uses no post-2016 totals.

Q: What is a vehicle miles traveled fee instead of the Highway Trust Fund gas tax?

A vehicle miles traveled fee, usually called a road usage charge, would replace the per-gallon fuel tax with a fixed charge per mile driven. Instead of paying tax on the fuel purchased, drivers would pay for the distance traveled, measured by odometer readings, onboard devices, or other reporting methods. The core argument for it is that it restores the link between road use and payment that fuel efficiency and electrification broke: a gasoline car, a hybrid, and an electric vehicle traveling the same miles would pay the same amount. Oregon’s OReGO program, launched July 1, 2015, tested 1.5 cents per mile with up to 5,000 volunteers who received credit for fuel taxes paid. California launched its own pilot on July 1, 2016, with more than 5,000 volunteers. The FAST Act authorized 95 million dollars over five years for state pilot grants, though the federal program was not yet operating in August 2016. Privacy, collection costs, and rural equity are the main objections.

Q: Can states toll existing interstate lanes under Highway Trust Fund rules?

Generally no, with narrow exceptions. The general prohibition on tolling federally funded highways is at 23 U.S.C. Section 301. Congress has carved out limited exceptions: bridges and tunnels on the Interstate System may be tolled under 23 U.S.C. Section 129(a); high-occupancy-vehicle lanes may be converted to high-occupancy toll (HOT) lanes under 23 U.S.C. Section 166, letting solo drivers pay to use carpool lanes; and the Interstate System Reconstruction and Rehabilitation Pilot Program, created by Section 1216(b) of the Transportation Equity Act for the 21st Century and amended by Section 1411(c) of the FAST Act, allows tolling on existing Interstate facilities for reconstruction, but it is capped at exactly three facilities in three different states. In 2012, all three slots were conditionally reserved, for Interstate 70 in Missouri, Interstate 95 in Virginia, and Interstate 95 in North Carolina, and no state had completed a conversion through the program. Broader tolling authority remains one of the most debated alternatives.

Q: Does the Highway Trust Fund pay for transit too?

Yes, through the Mass Transit Account, which receives 2.86 cents of every gallon’s federal fuel tax, for both gasoline and diesel. The account was created by the Surface Transportation Assistance Act of 1982 as the Great Compromise: one penny of the five-cent tax increase that took effect April 1, 1983, was dedicated to transit to secure urban legislators’ votes for the increase. This gives the transit account a weaker user-pays claim than the Highway Account, because it is financed by fuel taxes paid by motorists while its beneficiaries are transit riders who pay no fuel tax. The Congressional Research Service reports that 85 to 90 percent of Highway Trust Fund receipts come from highway users (CRS R45350.1). Defenders argue transit reduces congestion for drivers and serves the metropolitan economies that generate the fuel tax revenue; the counterargument is that a fee that taxes one group to serve another has ceased to be a user fee. Either way, the arrangement is best understood as a political settlement reaffirmed in every reauthorization.

Q: What is the Byrd rule in Highway Trust Fund law?

The Byrd rule is the pay-as-you-go enforcement mechanism codified at 26 U.S.C. Section 9503(d), originating as an amendment offered by Senator Harry Byrd to the 1956 legislation. It provides that if unfunded highway authorizations exceed projected receipts for the next four years, the Secretary of Transportation must reduce state apportionments pro rata. The design was meant to operate automatically, without a further vote, so that authorizations could never permanently outrun the dedicated taxes. In practice its force has been spent by the general fund transfers that began in 2008: each infusion allowed authorizations to continue at levels the fuel taxes could not support, preempting the automatic cut the statute contemplates. The rule was never repealed, but the transfer mechanism gave Congress a fourth option the 1956 design had not anticipated, which is why the trust fund’s discipline eroded without any formal change to the law.

Q: Why does diesel fuel carry a higher federal tax rate than gasoline?

The federal diesel tax stood at 24.4 cents per gallon in August 2016, six cents above the 18.4-cent gasoline rate, and the difference reflects the heavier burden diesel vehicles place on the road network. Diesel powers nearly all heavy trucks, and pavement damage rises steeply with axle weight, so trucks impose far higher maintenance and reconstruction costs per mile than passenger cars. The higher rate is a rough attempt to align what heavy vehicles pay with what they cost the system, consistent with the user-pays principle. Within the trust fund, the diesel tax is split 21.44 cents to the Highway Account and 2.86 cents to the Mass Transit Account, with a tenth of a cent to the Leaking Underground Storage Tank fund, mirroring the gasoline structure. Whether the six-cent differential accurately captures the true cost difference between heavy trucks and passenger cars is itself debated among transportation economists, but the rationale for charging diesel more is the weight-damage relationship, not any difference in how the revenue is used.

Q: Where does the Leaking Underground Storage Tank fee fit in?

A tenth of a cent of every gallon’s federal fuel tax, 0.1 cents of the 18.4-cent gasoline rate and 0.1 cents of the 24.4-cent diesel rate, is dedicated to the Leaking Underground Storage Tank trust fund rather than to the Highway Trust Fund. That is why the Highway Trust Fund itself receives 18.3 cents per gallon of gasoline and 24.3 cents per gallon of diesel, not the headline 18.4 and 24.4 cents. The LUST fund finances the cleanup of petroleum releases from underground storage tanks, a purpose related to fuel but entirely separate from transportation. Its inclusion in the per-gallon rate is a reminder that the figure drivers see as the federal gas tax is a composite of distinct levies with distinct destinations. For most purposes in the highway financing debate, analysts work with the 18.3 and 24.3 cent figures that actually reach the trust fund, since those are the revenues available to cover authorized surface transportation spending.

Q: How did the 1982 tax increase change the purpose of the trust fund?

The Surface Transportation Assistance Act of 1982 (P.L. 97-424, Title V), signed January 6, 1983, did more than raise revenue; it redefined what the trust fund was for. Section 511 raised the gasoline and diesel tax from 4 cents to 9 cents per gallon effective April 1, 1983, the first increase since the fund’s creation, and dedicated one penny of the five-cent increase to a newly created Mass Transit Account. Until that moment, the fund had been exclusively a highway financing mechanism, true to the 1956 bargain in which drivers paid for roads. After 1982, it became a surface transportation financing mechanism, in which drivers paid for roads and for transit. The change was the Great Compromise: urban legislators whose votes were needed for the increase conditioned their support on a transit share. Every reauthorization since has preserved the arrangement, and the 2.86-cent transit share of the current rate is its direct descendant. The 1982 Act thus marks the first major departure from a pure user-pays design.

Q: What happens when Highway Trust Fund spending exceeds its dedicated revenue?

Under the system as designed, the Byrd amendment at 26 U.S.C. Section 9503(d) would trigger pro rata reductions in state apportionments when unfunded authorizations exceed projected four-year receipts. Under the system as practiced since 2008, Congress transfers general Treasury revenue to cover the difference instead. The Congressional Research Service (CRS R45350.1) reports that every year since 2008 has seen a gap between dedicated tax revenues and authorized spending, and Congress has filled it with transfers, beginning with 8.017 billion dollars on September 15, 2008, under P.L. 110-318, and reaching a cumulative authorized 143.6 billion dollars through the FAST Act of December 2015. The practical effect is that state transportation departments continue receiving apportionments on schedule, construction programs proceed without interruption, and the financing shortfall is absorbed by the general fund. The cost is the erosion of the pay-as-you-go discipline the 1956 Act established: a rule that is waived every time it binds no longer constrains anyone.

Q: What did the 2015 FAST Act do about the trust fund’s shortfall?

The Fixing America’s Surface Transportation Act (P.L. 114-94), signed December 4, 2015, addressed the shortfall the way every reauthorization since 2008 had: with general fund transfers rather than a tax increase or spending cuts. The Act authorized the last 70.0 billion dollars of the cumulative 143.6 billion in transfers reported by the Congressional Research Service (CRS R45350.1), meaning nearly half of all transfers ever authorized came in this single law. At the same time, the Act looked toward longer-term alternatives: Section 6020 created the Surface Transportation System Funding Alternatives pilot program, authorizing 95 million dollars over five years for state experiments with replacements for the fuel tax, and Section 1411(c) amended the Interstate System Reconstruction and Rehabilitation Pilot Program governing tolling. The FAST Act was thus characteristic of the era’s hybrid approach, keeping the program solvent with general revenue while funding research into the mileage-based and other alternatives that might one day restore a user-pays base. It did not raise the fuel tax rate.

Q: How would a mileage-based fee address the electric vehicle problem?

A mileage-based fee solves the electric vehicle problem by changing what is measured. The fuel tax fails against electric vehicles because it taxes gallons, and electric vehicles use no taxed gallons; under 26 U.S.C. Section 4081, electricity is not a taxable fuel, so their road use goes entirely unmeasured by the revenue system. A per-mile charge measures distance instead, so a gasoline sedan, a hybrid, and a battery-electric vehicle traveling the same miles pay the same amount. Oregon’s OReGO pilot, operating since July 1, 2015, demonstrated the mechanics: up to 5,000 volunteers paid 1.5 cents per mile and received credit for fuel taxes paid, testing the charge as a replacement rather than an add-on. California’s pilot, launched July 1, 2016, with more than 5,000 volunteers, tested similar approaches. The remaining objections concern privacy in mileage reporting, the administrative cost of collecting from individual vehicles rather than from refiners, and fairness to rural drivers who travel longer distances by necessity.

Q: Why is tolling restricted on roads built with federal aid?

The restriction reflects the 1956 bargain itself. The Interstate Highway System was sold to the public and to Congress as a free system financed by user fuel taxes: drivers would pay at the pump, not at the tollbooth, and the fuel tax was the consideration for keeping the roads free. The general prohibition at 23 U.S.C. Section 301 encodes that promise, barring tolls on highways built with federal aid. The exceptions Congress later carved out, for Interstate bridges and tunnels under Section 129(a), for HOT lane conversions under Section 166, and for the three-facility reconstruction pilot program, are narrow precisely because the underlying promise remains politically potent. The double-payment objection draws its force from the same history: drivers paid fuel taxes for decades on the understanding that the roads would be free, and tolling them feels like being charged twice for the same pavement. Supporters of broader tolling authority argue the promise belonged to the construction era, that the aging system needs reconstruction revenue, and that electronic tolling has removed the delays that once made tolls burdensome. The restriction endures because the promise is still believed.

Q: How do the Oregon and California road-use pilots differ from the fuel tax?

Both pilots replace the fuel tax’s proxy with a direct measure of road use. Oregon’s OReGO program, launched July 1, 2015, under S.B. 810 of 2013, enrolled up to 5,000 volunteers who paid 1.5 cents per mile traveled and received a credit for state fuel taxes paid, so the charge functioned as a substitute for the gas tax rather than an additional levy. California’s road charge pilot, authorized in 2014 by S.B. 1077 and launched July 1, 2016, enrolled more than 5,000 volunteers and was underway at the article date. In both programs, what the driver pays depends on miles driven, not gallons burned, which means efficient and electric vehicles contribute in proportion to their road use for the first time. The pilots were designed as experiments, not as revenue systems: they tested mileage-reporting options, from odometer readings to in-vehicle devices, examined privacy protections, and measured administrative costs and public acceptance. Their findings were meant to inform the federal Surface Transportation System Funding Alternatives program’s state grant experiments authorized by the FAST Act.

Q: What share of trust fund receipts comes from highway users?

The Congressional Research Service reports in R45350.1 that 85 to 90 percent of Highway Trust Fund receipts come from highway users, meaning the motorists who pay the federal fuel taxes on gasoline and diesel. The remainder reflects the account’s other dedicated receipts and the structural mismatch at the heart of the transit question: the Mass Transit Account’s 2.86 cents per gallon is paid by motorists through the fuel tax, while the transit riders who benefit from the spending pay no fuel tax for their travel. That mismatch dates to the Great Compromise of 1982 and was a deliberate political bargain rather than an accident. The figure matters because it quantifies how close the trust fund still comes to the user-pays ideal on the revenue side, even as the general fund transfers since 2008 have broken the link on the spending side. The distinction between the revenue side and the spending side is central to the verdict: the taxes still fall mostly on users, but the spending they authorize is no longer constrained by what the users paid.

Q: Would indexing the fuel tax to inflation solve the trust fund’s problems?

Indexing would solve one of the three problems and leave the other two untouched. An indexed fuel tax would hold its real value automatically, ending the slow bleed that cost the tax approximately 73 percent of its purchasing power against highway construction-cost inflation (CRS R48472), and it would do so without requiring Congress to cast visible votes for increases. But indexing changes the rate, not the base. It would not restore the gallons lost to fuel efficiency, which rose 18.7 percent for new light-duty vehicles between model year 1993 and model year 2015, and it would not capture a single mile traveled by electric vehicles, which pay no federal fuel tax because electricity is not a taxable fuel under 26 U.S.C. Section 4081. Supporters call indexing the most administratively simple reform available and argue the perfect should not be the enemy of the good. Critics call it a stay of execution for a declining base and warn it would lock in a levy whose burden falls increasingly on lower-income drivers of older gasoline vehicles. Both descriptions are substantially accurate.

Q: Why does the construction-cost measure of the tax’s decline exceed the CPI measure?

Because the prices that matter for buying roads rose faster than the prices consumers pay generally. The Congressional Research Service reports in R48472 that the gasoline tax lost approximately 73 percent of its purchasing power measured against highway construction-cost inflation, while on a Consumer Price Index basis the 18.3 cents the trust fund receives would need to be roughly 29 to 31.7 cents to match its 1993 purchasing power, a loss of about 37 to 40 percent. Road building is intensive in asphalt, steel, labor, and right-of-way, inputs whose prices outpaced the general price level over the period. The statement that the rate freeze cut the tax’s real value by more than half is therefore valid only when tied to construction costs, which is the correct basis for a tax whose purpose is to purchase construction, and misleading when presented as a general inflation claim. A reader who hears that the tax lost half its value should ask which measure is being used, because the answer determines whether the proposed remedy is adequate.