Delucchi Study Finds That U.S. Motorists Do Not Pay Their Way


A dozen or so years ago, back when congestion pricing was a distant dream and New York City’s number one transportation priority was to squeeze more transit funding from government, the Tri-State Transportation Campaign commissioned me to determine which was greater: the dollars that New York State governments took in from drivers, or the dollars spent on drivers’ behalf. I spent months immersed in bookkeeping arcana, parsing revenue pots like the statewide Petroleum Business Tax and expenses like fire department equipment for prying crash victims from mangled vehicles, before I emerged with an answer.

My report, Subsidies for Traffic (PDF), established that all levels of government combined spent considerably more building and managing roads in New York State than they reaped from gas taxes, road tolls and traffic tickets. For every dollar expended on the road network by government, drivers kicked in just 65 cents. The other 35 cents — a cool $2 billion a year — was paid for out of general revenues, primarily property taxes collected by cities, towns and counties.


A year later, in 1995, I rolled out a similar analysis covering New Jersey. The implied annual subsidy for Garden State drivers was a smaller but still sizable $700 million, equivalent to 23 cents of each dollar of governmental road-spending.

Now a new study by Mark Delucchi, research scientist at the U.C. Davis Institute for Transportation Studies and the nation’s leading taxonomist of motor vehicle-related revenues and costs, has found that the New York and New Jersey pattern of taxpayers subsidizing motorists holds true across the entire United States. In Do Motor Vehicle Users in the US Pay Their Way? (PDF), a forthcoming article for the journal Transportation Research A, Delucchi writes:

To pay for [road] infrastructure and services, governments collect revenue from a variety of [motor-vehicle user] taxes and fees. The basic objective of this paper is to compare these government expenditures with the corresponding user tax and fee payments in the U.S.

The analysis indicates that in the U.S. current tax and fee payments to the government by motor-vehicle users fall short of government expenditures related to motor-vehicle use by approximately 20-70 cents per gallon of all motor fuel. (Note that in this accounting we include only government expenditures; we do not include any "external" costs of motor-vehicle use.)

That implied subsidy of 20 to 70 cents a gallon — which excludes social and environmental costs such as climate damage and uncompensated crash costs, which Delucchi has tallied elsewhere — equates to 7 to 25 percent of the current price of gasoline. On a dollar basis, U.S. drivers are underpaying local, state and national governments by $40 to $105 billion a year

Delucchi’s conclusion, "motor-vehicle users in the U.S. — unlike users in most European countries — do not ‘pay their way’," will come as no surprise to many of us. Still, putting the Delucchi seal of approval on the "subsidies for traffic" thesis is a watershed event. Dismantling those subsidies may have just gotten a little easier.

Photos: Jeknee and Leggnet on Flickr.

23 thoughts on Delucchi Study Finds That U.S. Motorists Do Not Pay Their Way

  1. Usually I agree with your points, but this one just doesn’t make a lot of sense to me. Taxes in this country have no relationship to the services an individual receives – your income taxes pay for schools whether or not you have children, emergency services whether or not you call 911, and parks whether or not you use them. Just because I don’t drive, and I am generally anti-automobile, doesn’t mean I don’t feel that I should pay for road maintenance. The food I eat and products I buy come in on vehicles. Having well-maintained roads is important for the circulation of surface mass-transit, emergency vehicles, deliveries, and many other benefits to non-drivers (even biking). Maintaining the road network also has important military implications in the (unlikely) event of an attack. I don’t think we should expect the maintenance costs to be paid for exclusively by drivers.

  2. Ed, I’d sort of agree – except for the folks who don’t think peds and bikes belong on the roads because we’re not paying our fair share. (I’m also confounded by people who will gladly pay the interest payments and gas taxes on their SUV – but complain about school levies and social services. (but this is another post))

    I think the bulk of the fees should be paid by the people who can do the most harm to the infrastructure – heavy trucks, then light trucks, then cars, and finally bikes and peds.

    Cue the “it will just get passed on to the consumer” music, and the argument that “America stops when trucks stop”. 😉

    Automobiling is cheap in this country. I look forward to reading the report.


  3. NYC subway riders, who create far,far fewer negative external costs like air pollution, traffic congestion, crashes , climate change etc pay about the same portion of their ride: 65%. Makes sense?

    CK I’d like to see you update these two excellent studies, and do an NYC only version. Be good support for congestion pricing. (Nice piece. Write more for Streetsblog!)

  4. True Mike, but, and I can’t believe I’m saying this, shouldn’t everyone share the costs of municipal transportation costs because of the range of region-wide benefits? I think we have decided as a society that it is so vastly important for business and safety to have a well-maintained road network that we are all willing to provide it at our mutual expense. The problem as I see it has more to do with reducing our dependency on combustion vehicles, and limiting the cost of road maintenance (among other impacts) by reducing overall traffic. But I don’t think that having the public share the costs of maintenance is unfair. Unless, of course, you might like to privatize the transportation system – howabout if Ford and GM owned the roads and had to pay for the maintenance?

  5. This report is significant for two reasons. Mike identifies the first: other transport modes (Amtrak comes to mind) are expected to “pay their way,” but here is yet more evidence that we do not apply the same standard to autos. The second is that it recasts the choice of the 90%+ Americans who drive alone to work everyday as an economic decision, not a “love affair with the automobile/open road/freedom/etc” or some other oft-repeated cultural phenomenon. If gas is so damn cheap (cheaper than water) and as a motorist in New York you get $2 billion in annual subsidies, doesn’t it make sense to drive?

  6. Ed — the services you cite (schools, 911, parks) are true “public goods” insofar as they’re not consumed or diminished with use; essentially just as much is available for me whether you use any or not. Not so with roads for much of the time. Moreover, the staggering social and environmental externalities of driving behoove us to employ any and all reasonable means of charging for all additional miles driven.

    JK — thanks for kind words (Ed too). Even if the bucks are offered, someone else needs to do the updates, the work is so tedious.

  7. I think that a comparison between Amtrak and automobile drivers is somewhat unfounded because rail features a private operator with exclusive use of the network, while roads are more of a public service. Certainly the MTA is largely financed by out-of-system revenue sources like state-wide taxes, but I guess that is something the MTA is criticized for and this report is trying to defend that position by showing that the cost of auto road maintenance is similarly shared? If you want to start talking about creative financing options for roadway maintenance, I’m all for it.

    As for the issue of driving being too cheap in the U.S., that the state subsidies, oil wars, and negative environmental impacts are too great a cost for us to bear, that I certainly agree with. I just don’t think the fact that car drivers don’t ‘pay their way’ is really noteworthy I’m sorry to say.

  8. CK- I suppose what I’m arguing is that roads are a true “public good”. And I would also argue that schools, 911, and parks do diminish with use, at least from a capital maintenance perspective. Anyway, I don’t want to take away from the value of Delucchi’s study, I just think the trajectory of the discussion should be more towards how to relieve the general (driving and non-driving) public of these costs rather than just shifting the burden to drivers, even as a driving-deterrent measure. How can we all reduce the public costs of roadway maintenance and harmful impacts of auto use? Because roads are good for both the driving and non-driving public. Maybe I’ll have time to write more later…

  9. Ed, people regularly refer to Amtrak and other transit systems as a “money-losing service,” and urge voters and politicians to “end subsidies.” Charlie’s point is that the automobile infrastructure is also a “money-losing service,” at which point the question is where the subsidies will bring most benefits to the public.

    Amtrak is not a private operator; it’s a government-chartered nonprofit corporation. If it ever got to a point where Amtrak’s revenues exceeded its expenditures there would be no investors getting rich; they would just use the “profits” to expand the network or reduce fares, save them for future use, or give them back to the government.

  10. I’ve hung around enough economic conservatives to know that the knock against public transit and Amtrak is this:

    How can government interfere with the private market by getting into the business of running a railroad?

    Swirl those words “government” and “business” around with one another and see what happens:

    Government . . . business . . . government . . . business . . . Communism!

    In fact, government interference with the private market is no less present in the enormous hidden automobile subsidy that Komanoff and Delucchi have drawn our attention to.

    A rail subsidy wouldn’t be needed if there wasn’t a powerful government subsidy drawing people away from rail and into their cars.

    The problem is that people across the entire political spectrum, myself included, can easily imagine a private railroad, but have difficulty seeing how the accomodations for the automobile could be anything but publicly funded (although the consistent conservatives in one of the reddest states in the nation are trying.)

    The reason for this dichotomy I suspect is that the public involvement in roads has been ancient and incremental (the dirt post roads that carried Paul Revere’s horse are the direct ancestors of the Interstates), while the public involvement in railroads and buses has been abrupt and recent (caused when the rail systems collapsed after massive public investments in road paving etc. began after World War II).

    Ed, in New York, the food you eat arrives by truck for two reasons 1) the George Washington Bridge, and 2) the Verrazano Narrows Bridge. If it arrived by rail, you wouldn’t have to pay for its shipment as a taxpayer. Energy sipping freight railroads are generally expected to pay for maintenance of their right-of-way, while energy guzzling 18-wheelers benefit from your tax dollars maintaining our asphalt infrastructure.

    And before we start getting into the idea that rail infrastructure benefits only “a private operator with exclusive use of the system,” let’s recall that anybody can board a train or a bus, but that the automobile is out of bounds for those too poor, too old, too young or too disabled physically or mentally.

    Finally, this is relevant because as long as we are talking about creating congestion pricing as a way of reducing traffic, we should also be talking about all the ways that public policy is encouraging traffic by making driving cheaper than it otherwise would be.

  11. I haven’t read any of these studies yet, but, from bicyclists’ point of view I suspect they understate the taxpayers’ subsidy of motorized traffic. The taxes and fees that motorists pay go mostly for Interstates and other highways that bikes usually do not use. Instead, bikes ride mostly on local streets, which are paid for, mostly by general revenues and local assessments. When cars and trucks use these streets, they are “freeloaders” in the same way that bikes are.

    However, the funding for bike amenities and mass transit subsidies tends, in most states, to come out of “highway” trust funds, which are funded by gas taxes. Bike facilities cost a piddling amount of money — a bike activist I know in Wisconsin likes to say that all of the state’s bike projects could be funded by the rounding error in a single major highway project — and these “diversions” are more than offset by the money that goes from general revenues to automotive subsities.

    But these “diversions” do create the illusion that motorists are subsidizing other forms of transportation, which gives “motorists’ rights” activists something to whine about. It would be nice to find a more straightforward way to finance transportation in this country, but I suppose it’s unlikely that this will happen.

  12. Komanoff: To what extent does the Delucci analysis take into account general-revenue military expenditures used to enforce our inalienable right to other peoples’ oil?

  13. Aaron Donovan,

    That’s a great post and I agree with you on just about everything. I do think we shouldn’t be careful to overstate the case though. When you say “A rail subsidy wouldn’t be needed if there wasn’t a powerful government subsidy drawing people away from rail and into their cars” I think we may be just doing that.

    Gasoline in Europe easily costs 20-70 cents per gallon more in Europe (actually it’s roughly twice the American price in many Western European countries) so assuming that the costs for the road network are roughly the same in Europe, this price at least equals the true market price. Yet rail in Europe still requires government subsidies.

    There are, I imagine many reasons for this, including policy reasons (e.g. even this higher price fails to capture all negative externalities involved in car travel or the train price does not capture the positive externalities of the resulting pleasant cities etc.). There are probably other reasons that people could supplement.

    The only country that has a supposedly unsubsidized and functional rail network is Japan — and even then, some people claim that it’s not really unsubsidized. (I don’t really know the details of this debate).

  14. (Taxes in this country have no relationship to the services an individual receives)

    A substantial share of the money collected by local governments is fee revenue, not taxes. It is a fee if you are only charged if you use a public service or facility, and only to the extent that you use it.

    Fees can be regressive, unless fees for services consumed by the better off are used to cross-subdsidize less extravagant facilities for the worse off.

    As I wrote on Room 8, fees are appropriate if by using a public facility or service, the user is doing some kind of public harm — by taking a share of something that is scarce, for example. That’s why we are charged for the most basic of goods, water — because when it was free people wasted both the water and sewage treatment capacity.

  15. Ed, I don’t really buy your “common good” argument. The price of the food and goods you buy already includes transportation expenses. You shouldn’t have to pay twice. Yes, the cost of the apple would go up if gas prices rose, but so should it.

    The inevitable response by some is “the poor”, who will no longer be able to afford the apple. Whatever… all I know is the amount of money out of my pocket probably won’t change either way. I just feel that a simpler tax system would make our true costs easier to understand — although government doesn’t want that.

  16. One point that often gets missed in the discussion about the necessity of roads for deliveries, emergency services, etc. (which I agree with — those functions have to have roads) is this: If the road network were built only to support those functions, what would it look like? And the answer is that it would almost certainly be considerably smaller (in terms of capacity) than it is today. The cost of adding to that (theoretical) network in order to make mass automobile use possible, and of maintaining that extra capacity, should be borne entirely by private automobile users.

  17. A few observations about Delucchi’s study:

    1. Delucchi includes the category “Interest earnings on payments invested to cover highway and other capital” as a payment made by motor vehicle users for the use of highways. He is referring to the excess monies collected for the Highway Trust Fund, which are invested in Treasury securities. This category represents 37-74 percent of Delucchi’s most stringent calculation of payments (WOC #1). It represents 28-66 percent of Delucchi’s second-most stringent calculation of payments (WOC #2).

    Delucchi reasons that motor vehicle users contributed the original capital in the form of gas taxes. Therefore, the interest that is earned on that capital also qualifies as a payment made by motor vehicle users.

    This is logical from an accounting point of view. But I’m not sure I buy it from a fairness point of view. That interest on Treasury securities is being paid by all U.S. taxpayers, regardless of whether or how much they use motor vehicles.

    2. Delucchi’s study only addresses highways and highway expenditures. The National Highway System has 162,000 miles of roadways, which is 4% of the 4 million miles of roadways in the U.S.

    How are the other 96% paid for? Ultimately, most of the money comes from property, sales and income taxes. Local roads are more subsidized than highways, and get little or no financing from motor vehicle user fees.

    IMHO, motor vehicles are more subsidized than Delucchi implies.

  18. Tim (#16) —

    Good point. I said something similar in “Crossroads,” a companion report I did in 1995, also for TSTC, on taxpayer subsidies for roads in New Jersey:

    “While it is true that non-drivers benefit from freight movement and municipal services that require roads, New Jersey’s roadway infrastructure has grown far beyond the level of a “common carrier” offering a modicum of access to trucks, buses and public services.”

    I hope to put up “Crossroads” on the Web soon.

    Laurence (#17) —

    Delucchi’s study addresses all roads, not just major highways. Your point on interest payments is beyond my scope; I’m fwd’ing your comment to Mark for possible response.

  19. Delucchi’s study addresses all roads, not just major highways.

    Charles — I am looking at Delucchi’s Table 2, “Motor-vehicle infrastructure and services provided by the public sector, 2002.” I see in WOC #1:

    Annualized cost of highways
    Highway law enforcement and safety

    I am not seeing public expenditures on non-highway road construction or maintenance. I am happy to be corrected; can you point out what I am missing?

  20. Laurence — “Highways” are another term for “roadways” in Delucchi’s study (ditto in the literature, generally). Pls direct further Q’s to Mark. Thanks. — CK

  21. Hi All,

    Great discussion!

    1. The interest charge that I estimate is not the relatively minor “real” payment on invested HTF monies, but rather is interest I impute to both the capital outlay and user payment accounts. This imputed interest does not show up in the accounts of transportation agencies, which report annual capital outlays and receipts. (By contrast, I estimate annualized capital costs, which include an interest charge.)See the paper for more discussion.

    2. I have submitted a paper to the journal Energy Policy on the military costs of oil use in the U. S. They’ve reviewed it, and now I am revising it. The paper is based on Research Report #15 in my Social Cost Series, on my faculty web page (

    3. There has been a lot of research on the social return on investments in highway infrastructure; as I recall, the general finding is that the rate of return was relatively high 40 years ago, when the system was new and expanding, but now is no higher than the return on other public investments. This may (or may not) be relevant to the discussions about who should pay for the highways..

  22. The reason this research is important is that current public policy decisions about transportation, particularly at the federal level, are made on the fallacious assumption that auto-user subsidies do not exist or at least are not significant. Defenders of automobile over-use forever complain about subsidies– real or imagined — to all other foms of transportation and also demand that every dime of auto user fees and taxes be spent solely and directly on more roads. This latter argument is about as absurd as arguing that all alcoholic beverage tax revenues be reserved for construction of bars, breweries and distillaries! As long as the cost of auto use remains heavily subsidized, auto use will remain destructively high. And we really must understand that we cannot sustain auto use at its current rate. U.S. residents drive more miles per vehicle, more miles per person and more miles per roadway than almost any other people on earth. In the process we consume more petroleum, both in total volume and on a per capita basis, than any other nation. We generate more air, water and ground pollution than does any other nation — altho China and India may soon surpass us simply because their populations are greater than this country’s population. We will still be the “world pollution kings” on a per person basis. One of the most straight fowrad ways to change this is to better align direct user costs with direct use of automobiles and all other pollution generating activities.

  23. I wonder if increased health costs should be considered subsidies to automobiles? By making our streets unsafe for bicyclists and walkers we don’t get daily exercise. And big parking lots make business too far apart to walk. Add the global warming, obesity, diabetes,noise pollution, water pollution…

Leave a Reply

Your email address will not be published. Required fields are marked *


Wisconsin’s Highway Spending Mania Makes Less Sense Every Day

Wisconsin isn’t known as a state that makes smart use of transportation dollars, whether it’s Scott Walker rejecting federal funds for high-speed rail service, denying funds for what would have been Milwaukee’s first suburban commuter rail service, or cutting millions in state aid for transit. Now a new report from the Wisconsin Public Interest Research […]

Drivers Cover Just 51 Percent of U.S. Road Spending

There’s a persistent misconception in American culture that transit is a big drain on public coffers while roads conveniently and totally pay for themselves through the magic of gas taxes. And that used to be true — at least for interstate highways, a fraction of the total road network. But that was many, many failed […]