In 1938, in Santa Fe, New Mexico, the photographer Dorothea Lange turned her camera to a mundane feature of the American landscape: a sign advertising gasoline prices. Unlike most such signs, however, this one broke down its 20½ cents-per-gallon price: gasoline, 5½ cents; state, 5 cents; Uncle Sam, 1 cent; city, 1 cent; railroad, 2¾ cents; agent, 1¼ cents; and “Me,” 4 cents. (See figure 1.) The joke lay partially in what many saw as the modern equivalent of highway robbery. Yet the humor also depended on another, equally important characteristic of gasoline prices, which always took the form of a flat, per-gallon rate: the rate of taxation was neither listed nor even necessarily known by the seller. Unlike goods with itemized sales taxes, gasoline appeared to be tax free, even if buyers and sellers both typically knew this was not the case.
This photograph, entitled “Santa Fe, New Mexico. Gas station price analysis,” was taken by Dorthea Lange in 1938. Courtesy Library of Congress, Prints and Photographs Division, Farm Security Administration/Office of War Information Collection, LC-USF34-019287-E.
This photograph, entitled “Santa Fe, New Mexico. Gas station price analysis,” was taken by Dorthea Lange in 1938. Courtesy Library of Congress, Prints and Photographs Division, Farm Security Administration/Office of War Information Collection, LC-USF34-019287-E.
The invisibility of gas taxes is partly a historical accident—the tax has always been an excise tax collected from wholesalers rather than a sales tax collected at the pump—but it also symbolizes the significant, yet opaque role that gas-tax revenues have played in underwriting the expansion of the nation’s vast automotive infrastructure. When automobile ownership began to surge in the late 1910s and early 1920s, gas taxes provided a new, almost magically large source of revenue that allowed states to embark on aggressive road-construction campaigns. But when strapped state governments in the early 1930s attempted to allocate a portion of gas-tax income to nonhighway purposes, opponents launched a successful campaign that prompted a growing majority of state legislatures to earmark gas taxes for highway expenditures. This system, coupled with new highway-planning methods that prioritized “self-funding” highways, resulted in a powerful, self-replicating system in which new highways generated new traffic, which in turn generated new revenues that were legally reserved for more new highways.1
When the U.S. Congress created an inviolate Highway Trust Fund in 1956—into which federal gas-tax revenue flowed, obligated entirely for interstate highway bills—it greatly expanded the system by elevating it to the federal level. Yet if the interstates themselves, like the car-oriented landscapes that they spawned, were quite visible, the underlying system of legally dedicating gas taxes to building an ever-larger, traffic-inducing automotive infrastructure was not. Indeed, beginning in the interwar period, state and federal gas-tax policies, coupled with new highway-planning techniques, reconfigured the American environment by funding the growth of a vast automotive infrastructure that was designed explicitly to stimulate near-constant growth in American demand for gasoline.
A Fiscal Deus ex Machina
By the end of World War I, traditional revenue sources could no longer keep up with the escalating costs of road construction. Before the war, a little over half of all rural highway funds came from state-issued bonds. Property taxes, another major source of highway revenue, funded most local government programs and projects—not just roads—and were already strained to produce sufficient revenue to meet growing highway demands. Older systems of road finance that allowed citizens to “work off” road taxes by supplying labor, horses, and equipment gradually disappeared as road construction machinery became more costly and sophisticated and road work slowly professionalized. In addition, World War I caused construction costs to skyrocket at precisely the same time traffic demands became overwhelming. As late as 1921 road revenues still came from a mix of property taxes, general fund allocations, bonds, federal aid, and motor vehicle license and registration fees. Even with the substantial new income from license and registration fees (which, almost overnight, had come to account for roughly one-fifth of state highway revenues), road-related revenues lagged significantly behind demand.2
Gasoline taxes provided a relatively uncontroversial solution to an increasingly contentious fiscal problem. Justified as a highway “user tax” along the lines of license and registration fees, the gas tax had a number of obvious advantages: it targeted the primary users of improved highways, it taxed motorists roughly in proportion to their use of roads, and it generated substantial revenue that states could use for road work or as security for bond issues. Most importantly, it had very low administrative costs, since states assessed a small number of wholesalers rather than a large and dispersed group of gasoline retailers.3
As state after state became aware of its strengths—in part because the American Association of State Highway Officials (AASHO) aggressively championed the idea—the new tax spread quickly. Five years after the first appearance of the tax in 1919, thirty-six states had adopted a gas tax; by 1929 all forty-eight had followed suit. Moreover, because the tax was invisible at the pump and because new oil-field discoveries and booming production in the second half of the 1920s caused oil prices to fall more or less steadily, states were able to raise gasoline taxes to a national average of four cents per gallon by 1931 without appearing to affect the falling price of gasoline.4 The prices that motorists paid at the pump actually fell, even as gas taxes inched significantly upward.
As a source of new revenue, the gas tax was substantial enough to underwrite mushrooming state road-building budgets, which, from 1902 to 1927, skyrocketed from 2 percent to 25 percent of total state expenditures. Together, gas taxes and motor vehicle fees generated an expanding proportion of state highway funding, rising from 20.4 percent in 1921 to 59.5 percent in 1930. In addition, because politicians justified both gas taxes and vehicle fees as road-user taxes, most states devoted gas-tax income exclusively to highways. The result was a de facto segregation of road-user revenues from general funds. Income from gas taxes exploded from $5.3 million in 1921 to $431.4 million in 1929, pushing gas taxes past vehicle fees as the primary source of highway revenue. By the end of the decade twenty-one states had eliminated property taxes as a highway-funding source.5
Motorists had good reasons to support gas taxes during the 1920s. The principle of user taxes, for example, seemed fair enough, given, in this case, the strains that escalating automotive traffic put on the nation’s highways and the widespread desire among motorists for better roads. The tax was neither onerous nor obvious—paid a few cents at a time, with the exact amount unadvertised—and it funded conspicuous, large-scale road construction. Moreover, gas taxes had strong support among state legislators, and initially even garnered the support of oil industry executives, who noted a direct link between better highways and growing gasoline sales. This widespread support for a substantial new tax took many by surprise—“Who ever heard, before, of a popular tax?” asked Tennessee’s chief tax collector in 1926. Yet in the context of the general prosperity, falling oil prices, and rising automobile use of the 1920s, many regarded the gas tax as a deus ex machina that saved state treasuries from bankruptcy and state legislators from having to say no to a growing motorized constituency clamoring for more, better, and faster road improvements.6
The Great Depression and the Politics of “Linkage”
When the Great Depression struck, most government receipts declined precipitously, but the gas tax proved remarkably resilient and stable as a source of revenue. Personal and corporate income tax revenues, for example, both fell in 1932 to just half of 1927 levels; property tax revenues also plummeted. By comparison, state gas-tax income rose in the first years of the depression to a record high of $537.4 million in 1931, dipped a negligible 4.3 percent in 1932 to $514.1 million, and then began to climb again, reaching $948 million in 1941. By that time, gas taxes produced half of all state highway revenues. Gas taxes proved such a reliable generator of revenue that the federal government introduced its own one-cent-per-gallon levy as part of the Revenue Act of 1932, which for the next two decades consistently produced more revenue than Congress appropriated as federal aid for highways.7
The combination of gas taxes, vehicle taxes, and federal spending kept large state highway budgets intact through the worst years of the Great Depression, despite massive revenue shortfalls that caused total state and local contributions to highway expenses to plummet. Federal aid offset part of the difference, and substantial federal work-relief grants earmarked for road projects offset the rest. Between 1933, when Franklin D. Roosevelt took office, and 1942, the first full year of U.S. participation in World War II, federal work-relief agencies devoted $4 billion to roads and streets and in the process expanded the parameters of federal aid for roads beyond the traditional focus on rural highways. Buoyed by stable gas-tax revenue and greatly expanded federal spending, highway budgets fell only slightly in 1932 and 1933; by 1934, and then through the rest of the decade, both expenditures and highway income exceeded 1929 levels.8
In that context, many state legislatures began to allocate portions of the large, steady income from gas taxes toward nonhighway expenses, provoking the first widespread backlash against the tax. Before the depression, this practice, which came to be known as “diversion,” had been insignificant, claiming between 2.0 and 3.3 percent of total collections between 1927 and 1931. As the depression deepened, however, that range became even higher (between 15.9 to 18 percent between 1934 and 1937), prompting intense opposition from the automobile and road-building industries. The National Highway Users Conference (NHUC), for example, which formed in 1932, became the biggest opponent of diversion. Chaired by Alfred P. Sloan, the president of General Motors, and led by representatives from automobile, oil, bus, truck, road, and farm organizations, the NHUC aggressively lobbied Congress to prohibit the practice. It achieved at least part of its goal with the passage of the Hayden-Cartwright Act of 1934, which reiterated NHUC arguments almost verbatim, proclaiming it “unfair and unjust to tax motor vehicle transportation unless the proceeds of such taxation are applied to the construction, improvement, or maintenance of highways.” In a huge loophole, however, the law would penalize a state for diversion only if it exceeded its current rate—grandfathering in the large-scale diversion occurring in predominantly large, heavily urbanized northeastern states. In addition, the law had an ironic do-as-I-say-not-as-I-do element: Congress at the time allocated just 60 percent of federal gas-tax income to highway purposes and diverted the remainder to the general fund.9
A more significant and longer-lasting legacy of NHUC’s antidiversion campaign can be seen at the state level. Increasing numbers of state legislatures enshrined the principle of “linkage”—that highway user–tax receipts should be used exclusively to fund highway spending. A more extreme form of linkage (and one nearly impossible to reverse), first seen in Minnesota in 1920, came in the form of constitutional amendments earmarking gas taxes for highway expenses. Sixteen states adopted such amendments between 1934 and 1945, and another nine did so by 1956, thus making linkage a constitutional requirement in a majority of states. By 1974, forty-six of fifty states required linkage either by statute or by constitutional amendment.10
What began as opposition to diversion eventually became opposition to higher gas taxes, dampening further increases in the second half of the 1930s. Between 1918 and the passage of the Hayden-Cartwright Act in 1934, for example, the gas tax rose more or less steadily from nothing to a national average of 5.21 cents per gallon. From 1934 until 1941, by contrast, the average increased less than half a cent. Gas-tax income continued to grow on the strength of expanding automobile ownership and rising per-vehicle gasoline consumption, but steady increases in the tax rate effectively ceased by the mid-1930s.11
Most important, as state after state legally linked the large, stable income from gas taxes to highway-improvement budgets, highway administrators forged a significant new relationship between gas-tax revenues and highway planning. As it had with the principle of linkage, the Hayden-Cartwright Act gave highway planning a big boost, allowing states to use up to 1.5 percent of federal-aid funds on planning activities. The Bureau of Public Roads (BPR) pushed states to spend this money on detailed, standardized highway planning surveys, which began in 1936. The surveys compiled detailed inventories of the physical characteristics of every mile of road in state highway systems—ranging from isolated country roads to heavily traveled highways such as the famed Route 66—and conducted extensive statewide traffic counts. Together, the surveys gave state and federal highway planners an unprecedented amount of detailed information on traffic patterns and road conditions across the nation. Just as significantly, the planning surveys included detailed financial studies designed to identify and classify every dollar of revenue collected for (and expended upon) roads and streets. These studies, in conjunction with new systems of data analysis, allowed planners for the first time to develop an accurate picture of the exact relationships among road expenses, traffic volumes, and highway-user taxes. The goals, at least initially, were to account precisely for how all of the many levels of government spent user taxes and thereby to identify the extent to which user revenues actually paid to improve and maintain individual highways. How much of each highway, in other words, did gas taxes (and other user fees) really fund?12
The answers that the studies produced—and the method they used to produce them—marked a watershed change. The method was startlingly simple: for any given stretch of road, administrators determined how much “income” it generated by multiplying its traffic volume by the average per-mile user-tax revenue that each motorist paid. By comparing the resulting figure with the improvement and maintenance costs for the same stretch of road, administrators could determine, with compelling accuracy, which highways carried enough traffic to “pay for themselves”—and which did not. This simple calculation allowed planners to create a cost-effectiveness index and classification system for different types of roads and to justify traffic-service priorities in clear-cut economic terms. “Highway builders are proceeding on the principle,” explained Thomas MacDonald, the chief of the BPR, in 1938, “that the utilization of the highways must produce directly the revenues with which to finance their construction.” As long as states did not make “large diversions of this income to other purposes” or spread them “over an ever-increasing mileage,” MacDonald reassured the nation’s politicians, highway planners could provide self-liquidating highways that would not make unpredictable demands on state budgets.13
When that formula combined with the new state-level budgetary principle of linkage, a powerful, self-replicating system emerged: gas taxes funded more and better roads, more and better roads generated new traffic and longer trips, new traffic and longer trips consumed more gas, higher gas consumption created more tax revenue, and more tax revenue funded more and better roads. The new plans rested on a majestic vision of self-fueling growth and promised ever-improved highway facilities. And the entire system rested on the premise of the steadily growing consumption of a seemingly limitless supply of cheap, minimally taxed gasoline.
Gas Taxes, Federal Linkage, and the Interstate System
Beginning in the mid-1930s, as the BPR worked to cement its vision of a self-replicating, self-fueling system of highway expansion, it encountered disagreement over what a “modernized” national highway system ought to look like. Most highway administrators had become convinced that it should not look like the existing “primary” highway system, epitomized by Route 66, the self-proclaimed Main Street of America, whose colorful roadside commerce and diagonal path across the center of the country gave it a larger-than-life presence in American popular culture. The endless string of restaurants, tourist cabins, gas stations, and assorted tourist traps that gave highways such as Route 66 their popular appeal was also responsible for turning huge stretches of the primary system into dangerous, traffic-choked ribbons of congestion. Long strings of roadside businesses also made it inordinately difficult (and expensive) to resolve traffic and safety issues by widening highways, since businesses tended to locate directly on the edge of the right of way, hemming in expansion. To deal with those problems, some highway officials wanted to emulate Germany’s new high-speed, limited-access highways, the autobahnen; others found inspiration in the Pennsylvania Turnpike, a 160-mile, limited-access, toll-financed superhighway that opened in 1940; still others were motivated less by particular examples than by the desire to reduce the nation’s appalling accident rate, which in 1935 translated into roughly 827,000 reported accidents that killed 37,000 people and seriously harmed 105,000 more. Whatever their disagreements, all proponents of better highways faced a common constraint: limited resources to implement their vision. Because “modernized” highways almost invariably meant wider highways and frequently meant introducing limited access, their land requirements—and thus their expense (particularly in urban areas)—were lavish by prevailing standards. The short-lived popularity of toll roads grew from their promise of generating project-specific revenue streams and not needing to compete with other projects for limited funding. In response to growing support for toll roads in the second half of the 1930s, the BPR issued Toll Roads and Free Roads (1939), a scathing assessment of the financial viability of toll roads and a counterproposal for a new system of “modernized” interregional highways to be funded by existing gas-tax revenues.14
Highway planners spent the war years expanding this blueprint, which they unveiled in Interregional Highways (1944). The report recapitulated the main points of Toll Roads and Free Roads, emphasizing the plan’s traffic-service priorities, its rejection of tolls, and its call for an updated interregional highway system. New in the report was the proposed size of the system (which grew from 26,000 to 39,000 miles), the recommendation that the entire system have restricted access, and a proposal to build parking facilities near traffic interchanges on the edge of big-city downtowns. In Congress, protracted wrangling produced the Federal-Aid Highway Act of 1944, which, at $500 million per year for three years, was at the time the largest-ever federal highway bill. Moreover, and of enormous significance, it authorized—but did not dedicate any specific funding to—a new 40,000-mile system based on the proposals in Interregional Highways, which it named the National System of Interstate Highways.15
For the better part of a decade after its creation in 1944, the new interstate system plan limped along underfunded and largely unexecuted, but in the early 1950s two key developments transformed it into the most wide-ranging, landscape-altering public works project in U.S. history. The first was the Federal-Aid Highway Act of 1954, which earmarked the first money specifically for the interstate system ($175 million) and raised total federal aid to highways by roughly 50 percent for 1956 and 1957 so that for the first time highway expenditures matched federal gas-tax income. The second development began when President Dwight D. Eisenhower became actively involved in highway politics, declaring his intention to create a new transcontinental system that would solve the nation’s increasingly well-publicized highway woes. The resulting maneuvers, countermaneuvers, plans, and counterplans are among the best studied in American highway history and ultimately culminated in the passage of the momentous Federal-Aid Highway Act of 1956 and the Highway Revenue Act of 1956.16
In addition to expanding funding for the federal-aid program, these two acts upgraded, slightly expanded, and fully funded the interstate system that Congress originally approved in 1944, officially renaming it—in a sop to Cold War politics—the National System of Interstate and Defense Highways. The legislation differed notably from previous federal-aid acts in three areas. First, the “new” interstate system was the same transcontinental highway network established in 1944, with the exception of adding one thousand miles to the system and formally reviewing the urban routes designated in 1947.17 However, by stipulating that the entire system be built to uniformly high, multilane superhighway standards—even in locations where traffic volume did not justify them—Congress greatly expanded the plan’s scope and stepped beyond the longstanding engineering principle that traffic demands alone should shape the nation’s highways.
Second, the legislation funded the construction of the entire interstate system, treating it as a single project to be finished in a designated period. This conceptualization broke with a basic understanding that had guided both federal aid and the various federal-aid highway systems since at least 1921. Disbursing federal aid for highways had always been an ongoing process of dividing available money among states, not a program designed to build a fixed, predetermined highway system. In contrast to the 1956 legislation, earlier federal-aid systems had existed primarily to ensure that limited expenditures would not be dispersed in an uncoordinated way.18
Third, and crucial to its political success in Congress, the legislation formally enshrined linkage as federal policy. To meet the 90 percent federal share of the interstate system’s cost, Congress raised the federal gasoline tax by one cent and created new excise taxes on tires and vehicles. That revenue, along with other existing highway-user income, flowed into a new Highway Trust Fund, the sole purpose of which was to pay interstate construction bills. The Highway Trust Fund effectively insulated its revenue from political debates over appropriations, even after cost estimates for completing the system began to balloon—rising from $27 billion to $39.5 billion between 1954 and 1958 alone. Meanwhile, rising gasoline consumption, coupled with another one-cent hike in the federal gas tax in late 1959, kept Highway Trust Fund coffers full. With the contents of the trust fund “linked” to interstate construction, the system’s completion was guaranteed even after the winds of antihighway sentiment began to blow with growing force in future decades.19
The implications of these interrelated developments for the history of oil in the United States have been profound. The linkage of federal gas-tax revenues to interstate construction cemented a version of the self-fueling system that was on an even grander scale than the one highway planners first conceived of during the Great Depression, which coupled the nation’s expanding automotive infrastructure to its expanding gasoline consumption. As a result, the principle of dedicating ever-growing gas-tax revenue to highways became a cornerstone of the American political economy for the next half century. Even after opponents finally “busted” the Highway Trust Fund, securing some of its revenue for other purposes such as public transit, the revenues reallocated have been small enough to have the effect of reinforcing, rather than challenging, the dominance of highways in American transportation policy.20
Equally important, and central to how the self-fueling system operates, has been the concept of invisibility. It has played a role in the form of invisible taxes, duties paid by gas wholesalers and therefore never seen by consumers. Invisibility has also helped the self-fueling system in the political world: the legal linkage of gasoline taxes to road budgets has shielded car- and highway-centered transportation policy from the regular discussion and debate that typifies the allocation of general funds. And finally, from an environmental standpoint, the repercussions (aside from smog) of burning gasoline in staggering quantities have been difficult to see, despite the ubiquity of gasoline-burning vehicles in the United States. Even the ecological effects of fundamentally reconfiguring the landscape to make travel by automobile easier have tended to recede into the background, since car-dependent environments have become the norm, dominating the places where most Americans live their lives. Visible or not, however, in the United States oil has been and remains pervasive: fueling vehicles, spurring economic growth and, not least, underwriting the massive costs of building and maintaining America’s vast automotive infrastructure.
He wishes to thank Ed Linenthal, Brian Black, and the anonymous JAH referees for their insightful comments
