Rad Geek People's Daily

official state media for a secessionist republic of one

Yes, Virginia, government roads really are government subsidized, and no, they don’t approximate freed-market outcomes

Here's a pretty old post from the blog archives of Geekery Today; it was written about 16 years ago, in 2008, on the World Wide Web.

When left-libertarians argue with more conventionally pro-capitalist libertarians about economics, one of the issues that often comes up is government control over roads, and the ways in which state and federal government’s control over roads has acted as a large subsidy for economic centralization and national-scale production and distribution networks (and thus, to large-scale big box retailers, like Wal-Mart or Best Buy, dependent on the crafty arrangement of large-scale cross-country shipping as a basic part of their business model). People who have a problem with this analysis sometimes try to dispute it by arguing that government roads aren’t actually subsidized — that heavy users of government roads are actually getting something that roughly approximates a freed-market outcome, because users of government roads pay for the roads they get, in proportion to how heavily they use them, because government roads are funded by gasoline taxes, tire taxes, and government-imposed licensing fees, which all go up in cost more or less proportionally to increases in use of government roads. Thus (the argument goes), funding for government roads is more like a fee-for-service transaction on a freed market than it’s like a classic case of government subsidies. But in fact, this argument is completely bogus, for at least three reasons.

The first reason is that, contrary to popular misconception, government-imposed gasoline taxes and user fees on road users do not actually fully fund the costs of government road-building and maintenance; government funding of roads actually includes a substantial subsidy extracted from taxpayers independently of their usage of the roads. Government budgets for road building and maintenance in the US draw from general funds as well as from earmarked gas taxes and user fees, and those budgets are subsidized by state, local, and federal government to the tune of about 20–70 cents per gallon of gasoline expended.

The second reason, which ought to be obvious to libertarians given how much we have talked about the use of eminent domain over the past few years, is that government road-building is substantially subsidized by the fact that government can — and routinely does — use the power of eminent domain to seize large, contiguous stretches of land for road building at arbitrarily fixed rates below what the land-owners could have demanded in a free market land sale. Even if it were the case (as it is not) that usage-based levies like gasoline taxes and government licensing fees were enough to cover the budget for government road building and maintenance, that budget has already had a massive, unmentioned government subsidy factored into it due to the use of eminent domain.

The third reason is that a freed market is able to match the supply for roads to the demand at something like the appropriate cost not only because people pay for the roads in proportion to their use of the roads, but also because the prices for road use are set by negotiations between road users and road builders in a competitive market, and because the ownership and management patterns of roads are determined by patterns of free economic decisions to buy, sell, lease, develop, abandon, reclaim, and subdivide land. Freed markets aren’t just a matter of paying for what you get (as important as that is); they also have to do with the freedom to get what you get by alternative means, and with patterns of ownership and control based on consensual negotiation rather than on force. No matter how roads are funded, there is no way to approximate freed-market results with government monopoly on sales or politically-determined allocation of ownership. (Again, this is something that ought to be obvious; it is just the socialist calculation problem applied to the market for road transportation.)

And roads funded by government-imposed gasoline taxes will always be either noncompetitive or subsidized: if there were any significant private roads competing with roads funded by government gasoline taxes, the taxes on the gasoline that drivers burn on those roads become a subsidy to the government-controlled roads. The more users use the non-government roads, the more they would be subsidizing the government roads.

Further, the ownership and management patterns of government roads are determined by electoral horse-trading and arbitrary political jurisdictions, not by free economic actors. As a result, decisions about what roads to build, how to direct funds to those roads, how to price the use of those roads, etc. are typically made by state or federal legislatures, or state or federal executive bureaus. Governments are far more responsive to political than to economic pressure; governments generally will not, or cannot, sell off roads or spin off control over local roads to the people who use them most and can best manage them; state and federal governments exercise centralized control over far larger fiefs than it would ever be possible or profitable to amass on a free market. Thus, for example, because the building and maintenance of roads in Las Vegas is controlled, not by free market actors in Las Vegas, but rather by the Nevada state government, we have Las Vegas drivers paying in 70% of the state’s gas taxes and getting back only 61% of the state’s spending on roads (which is an increase over the 2003–07 average of 53%) — meaning that we are forced to turn tens of millions of dollars over to subsidizing highway building and maintenance in the rest of Nevada. Here’s NDOT’s reasoning as to why we should get stuck with the bill:

If NDOT based its road building program strictly on usage, [NDOT assistant director of engineering Kent] Cooper said, then no new highways would be built outside of Clark County.

He noted that freeways in Las Vegas attract 150,000 to more than 200,000 vehicles a day. No other area in the state has such high use.

— Ed Vogel, Las Vegas Review-Journal (2008-11-26): Southern Nevadans get less bang for their road tax buck

Now, maybe Kent Cooper thinks that it is just and wise to force Las Vegas drivers to pay tens of millions of dollars in subsidies so that NDOT can build expensive roads that nobody wants to use.

Maybe he’s right about that, and maybe he’s wrong. But whatever the case may be, the only way to get freed market results in roads is by freeing the market. Under government ownership, government funding, and government control, roads are subsidized by taxes that are levied independently of road usage, built using a subsidy created by forced seizure of land, and users of high-volume local roads are typically forced to subsidize expensive, long-distance cross-country roads that they aren’t using. This kind of allocation of resources for long-distance, non-local highways — which further distorts an already subsidy-distorted system by distorting the flow of money within that system away from the heavily-used local roads and into the high-cost, high maintenance long-distance roads, can certainly not be called any kind of approximation of a freed market in roads.

17 replies to Yes, Virginia, government roads really are government subsidized, and no, they don’t approximate freed-market outcomes Use a feed to Follow replies to this article · TrackBack URI

  1. Black Bloke

    I’m glad you addressed this point. I’ve been unclear about this myself.

    I had used the calculation problem when it came to roads before as well as the problem of eminent domain, but the other objection regarding the specifics of the user fees reply is new to me.

    Thanks for this.

  2. anonymouse

    And you probably haven’t even counted all the externalized costs of roads that don’t normally get counted under road spending, such as the police who patrol the highway, the pollution caused by cars driving on the highway, and so on. I doubt there would be very many private roads, but private rail lines did and do exist in many places, including the entirely private US freight rail system.

    Eminent domain, abused though it sometimes (often?) is, seems like a practical solution to the tough problem of getting long contiguous rights of way. I’m really curious though to hear any ideas you may have as to how this can be done without government force, given that every landowner will attempt to charge for his land as if he were the last thing standing in the way of the project.

  3. David Z

    I agree with the Bloke – this is as brilliant a deconstruction of the “roads” myth as I’ve ever encountered.

  4. David Z

    anonymouse –

    one method of dealing with the ‘infinite holdouts’ you describe, would be for the road company to propose two or more possible routes, which would force the landowners into a sort of competition, the deal would only go through if the collective could come to terms, internally, with the offering for the bulk of the land.

    Also, you’re neglecting the fact that people do want roads, more-or-less, and so granting rights-of-way/easements for such purposes would also not be out of the question.

  5. quasibill

    Excellent work, as usual Charles.

    The eminent domain point is one that many people overlook – our straight as an arrow 8 lane highways just are highly improbable in a free market. It takes coercion to achieve that sort of result. The most prominent anecdote for me is when I worked for PennDOT as an intern and they were involved in a “reverse takings” case that had its genesis, IIRC, in the 1950s. At the time this state highway (2 lane) was made, PennDOT (or its predecessor) had seized a 10 foot lane from the front yard of this couple. After the seizure, the highway itself came within 10 feet of the front door of their house, which sat on a lot that was below road grade. Every winter, windows in the house would be broken by passing snowplows as they pushed the snow and slush forcefully into the front of the house. Every 18 wheeler shook the house violently, to the point of breakage inside (the couple invested in specialty cabinets to protect their dining ware). The right of way that they seized extended to the top step of the stairs to the front door.

    And yet, PennDOT successfully argued that they hadn’t rendered the premises unusable, and therefore only paid a pittance for the “nearly worthless” tract of front yard. 50 years later, after the pioneering work of several attornies, the widow filed a reverse takings suit, and PennDOT was furiously spending money to make sure that they didn’t pay her an extra dime. This money came from general funds.

    And this goes to the question anon asks – straight arrow roads are far less likely in a free market. I personally think that this is indisputable. People develop all sorts of “irrational” attachments to land, and they won’t sell for any reason. That’s why old roads meander with all sorts of S curves and other strange configurations. You also probably wouldn’t have tunnels, and quite so many large bridges. But small roads would exist, and that would be sufficient to serve the regional economy that would exist. The straight arrow highways (and their predecessor straight arrow rail lines, also formed through “internal improvement” projects by various government entities) are a necessary precondition to the current national and global economies. The economy would be very different in their absence, staying more regional and local.

  6. anonymouse

    And thinking more about this, one of the problems that eminent domain encounters is that people tend to assume that the government has infinite money at its disposal, whereas some private road company would not. Also, a private road company would have to consider the actual costs of property acquisition when planning the route (a lot of the weirdness of the Chicago El can be explained that way), while the government can just assume that it will get sufficient money to pay each landowner some “appropriate” amount and bulldoze with impunity.

  7. Alex Koumparos

    I don’t disagree with any of the observations about how unfree (and unjust) the government transit system is as outlined in the excellent analysis above, but is it not premature to assume that government road subsidies are a disproportionate benefit to big box stores? Suppose WalMart carries 20,000 gallons of milk to serve a given region, and the alternative is, say, 200 small retailers each carrying 100 gallons of milk. Is it not likely (or at least possible) that the distribution cost per gallon of milk is lower (under any sort of transport network) for WalMart than for the 200 small retailers? If it is the case that whatever the road system, WalMart’s distribution costs are lower per lb than a small competitor, wouldn’t government transport subsidies disproportionately benefit the smaller companies.

    Note, I’m not shilling for WalMart – I don’t really care what the shape or size of firms ends up being under a just system of property rights, but I am curious whether these subsidies are really as (on balance) beneficial to WalMart as is popularly supposed.

  8. Araglin


    This was indeed another great post. For all the critiques of eminent domain I’ve seen (mainly taking the position of “it’s not necessary” against Richard Epstein and the like), I don’t believe it’s ever occurred to me how the use of eminent domain (in order to assemple enough contiguous tracts to build a highway or Interstate) give an additional artificial advantage to persons and businesses dependent upon long-distance travel and trade, at the expense of local producers producing for local use (over and above that conferred by the financing of road construction and maintenance).

    Another issue I haven’t given sufficient attention to is that of “hooking” suburban houses up to municipal utility grids without charging the new “user” the full marginal cost of the “hook-up.” I imagine this is just a way of redistributing wealth from incumbant landowners (or those otherwise bearing the incidence of property taxes or utility rates) within a municipality to politically-connected real estate developers.

    Does anyone know of good sources dealing with this question?

  9. Araglin


    I think I saw somewhere that because of the point you brought up, governments ultimately intending to take full title to a parcel often do so using a two-step maneuver whereby they first impose an onerous regulation that has the intented effect of reduces the parcel’s value say from $200,000 to $20,000 (which would not rise to the level of a “regulatory taking”* under current 5th amendment jurisprudence – because it’s not worth nothing and still has some permitted use); and, then, once the property’s value has been slashed, they then come back and take title to the parcel itself, “saving” them $180,000. Have you heard of this?

    Also, in the event of a constitutional “taking,” isn’t one only legally entitled to compensation for the loss of value to one’s own property from the taking of one’s own property taken in isolation (and not the diminution resulting from the takings from neighboring landowners)? If this is right, couldn’t those wielding the power of eminent domain “game the system” by simply taking from numerous continguous landowners one-by-one, with each such taking reducing the “pre-taking” value of the next one (and thus reducing the amount required to be paid in compensation)?

    Thanks, Araglin

    *Could you difference among the terms “regulatory taking,” “reverse taking,” and “inverse condemnation”? I have seen all of them used a decent amount recently but only happen to have a distinct sense for the meaning of “regulatory taking.”

  10. quasibill


    To my memory, a taking can be found even when not all economic value has been stripped from the property. Lucas certainly involved a (politically motivated?) concession that ALL economic value was lost, but the SCOTUS has since ruled (with Scalia doing his best footnote kung fu, IIRC) that such is not a precondition. A government need only significantly interfere with “investment backed expectations” (a phrase I found suspicious even before I became a libertarian) to be found to be taking a property through regulation. Your contiguous landowner scenario may exist, though, because of standing requirements and simple ignorance or inertia (most people won’t realize how much their property value has been decreased, or won’t care because they plan on staying there anyway). Which is also the reason why, in less affluent areas, I think your regulatory two-step could be effective as well, though I can’t say I’ve witnessed it.

    In PA, we don’t (‘didn’t’ is probably more appropriate for me now since I’ve been out of that sector for a while) use the term “inverse” but “reverse”, so most likely they are the same thing in your jurisdiction. All it is is an eminent domain action instituted by the allegedly aggrieved landowner, who claims that her property was taken even though the state did not institute an eminent domain proceeding, or at the very least did not institute one against part of a piece of property. Most often in my experience, these cases arise from either partial condemnations like the one I described, or from regulatory takings, an example that I can provide is where someone owns a plot of land on which is found an endangered species (here in PA, the bog turtle seems to be able to travel many miles from natural water sources to create problems for land owners who have land that receives run-off from other developed properties; either that, or the environmental groups have gotten very pro-active and are “seeding” them wherever they can). Since developing on endangered habitat requires onerous allowances to maintain the habitat, absentee owners, who have held the land in the hope of future development, often bring reverse takings claims against the state DEP after the DEP imposes the onerous burdens.

  11. Rationalitate

    Two things:

    1. Roads actually do not receive much of an accounting subsidy at all (that is, a subsidy when you exclude eminent domain), mostly because there’s a huuuuge “user fee” when it comes to roads that you’re not taking into account: the cost of the car, of the gasoline, of the car’s maintenance, and of the insurance. When you factor those in, the user fees paid by individuals to use the road/car system surely cover well over 80% of the total costs.
    2. In order for the outcome to approximate free market conditions, the government would have to price the roads high enough that the profits would be similar to the profits of renting out land that roads are situated on in the most productive other use. Imagining how this would work in the center of a city, it’s obvious that the vast majority of the roads would be auctioned off and developed as luxury condos, retail, and office properties, and the cost of driving would increase as the supply of roads dwindled, until critical mass is achieved for profitable private mass transit, which has the potential to be far more lucrative an enterprise at certain densities as compared to roads/cars.
    3. The biggest help that roads receive isn’t a true subsidy at all: it’s zoning and parking regulations, or land use regulations in general. They generally force low density where higher density is demanded by market participants, making it easier and cheaper for the govenrment to provide roads to all those who want to live somewhere, and making it more difficult for private transportation companies (like the private mass transit this country had before the progressive reformers got rid of those nasty companies starting around the 1910s) to compete.
  12. Rationalitate

    …I guess those were three things.

  13. TGGP

    This site claims that zoning laws are more important than subsidies in favoring autos over mass-transportation and denser living: http://marketurbanism.com/2008/11/12/matt-yglesias-fails-to-make-the-right-case-against-highways/

  14. Discussed at marketurbanism.com

    Yes, Virginia, government roads really are government subsidized, and no, they don't approximate freed-market outcomes | Market Urbanism:

    […] I came accross an article by Charles Johnson, who blogs at Rad Geek. Â The article had linked to a Market Urbanism post […]

— 2009 —

  1. Discussed at radgeek.com

    Rad Geek People’s Daily 2009-04-25 – Three notes for the critics of the critics of apologists for Wal-Mart:

    […] GT 2008-12-01: Yes, Virginia, government roads really are government subsidized, and no, they don’… […]

— 2011 —

  1. Discussed at radgeek.com

    Rad Geek People’s Daily 2011-04-28 – Subsidized Order:

    […] would be more accurate to say that people like their cars when they come with a massively subsidized network of government roads and a government-imposed glut of subsidized parking […]

  2. Sheldon Richman

    Rad, right again! Great piece!

Post a reply

Your e-mail address will not be published.
You can register for an account and sign in to verify your identity and avoid spam traps.

Use Markdown syntax for formatting. *emphasis* = emphasis, **strong** = strong, [link](http://xyz.com) = link,
> block quote to quote blocks of text.

This form is for public comments. Consult About: Comments for policies and copyright details.

Anticopyright. This was written 2008–2011 by Rad Geek. Feel free to reprint if you like it. This machine kills intellectual monopolists.