19 Oct 2013

Without Government, Who Would Give Us Traffic Jams?

Economics 46 Comments

In a previous post, I noted with irony that Eugene Fama listed roads matter-of-factly as an example of where government spending could be quite productive. In case some readers don’t understand, this is funny to Internet anarcho-capitalists, since a popular meme for us is, “Who Will Build the Roads?” I had asked if anybody knew whether Fama really thought government did a good job with roads (compared to the outcome in the private sector), or if perhaps he was just giving a standard example to show that he wasn’t preaching anarchy in his opposition to the Obama stimulus package.

Anyway, in the comments I was surprised to see some of the regulars here taken aback by most post. I thought the prima facie case against government ownership of roads was obvious, but OK:

One of the obvious flaws in government roads is that they don’t charge enough. That’s what causes traffic jams during rush hour. This isn’t a rare thing that happens on the day before Thanksgiving; it happens twice every workday. There are several hours when you just know “don’t try to drive if you can avoid it” if you live in a crowded area.

Whatever inefficiency Fama sees in the form of collecting payment on privately-produced roads, I have to imagine the lost productivity just due to New York City rush hour would dwarf it. I bet there are easily a million professionals who spend an extra two hours in traffic per workday, than they would have to in a private road system. Let’s say they generate an average of $50 per hour. So that’s easily $100 million per workday just in New York City alone.

Walter Block has written extensively on this topic. He likes to stress the large number of traffic fatalities that would undoubtedly be lower in a market road network. In other words, it’s not a fact of nature that such-and-such people die each year in traffic accidents. There are certain intersections and other areas that are notorious for accidents, and the political rulers aren’t nearly as concerned about fixing the problem as private owners would be.

46 Responses to “Without Government, Who Would Give Us Traffic Jams?”

  1. Robert Fellner says:

    Clifford Winston of the Brookings Institution wrote a fantastic book highlighting the atrocious failure that the government provision of roads, bridges, canals, airports, air traffic control systems, highways, and railroads has been.

    It’s a fantastic book and especially compelling because it is coming from a non-Austrian perspective.

    Some key points from my short blog review:

    Cost overruns on public road projects average at least 8.4% with a very large standard error, indicating some overruns are much greater. It appears this consistent pattern of under estimating the cost of public projects is due to strategic misrepresentation or “lying” as the authors of the study on the matter determined.

    Policymakers continue to ignore improvements in technology that would increase lifespan of roads and reduce long term maintenance costs dramatically. This is because they come with high up-front costs that are politically expensive. Meaning there is little incentive for them to take the short-term heat from voters of raising spending/taxing just because it is the best decision and will be save money – as well as lives – in the long run. In other words, the inherently short run focus that is government is what is responsible for maintaining the long term health of the nation’s infrastructure. I wonder what Apple’s stock price would do under that type of leadership….

    The traffic control system in many cities today were developed by inexperienced public officials for whom the automobile was a new mode of transportation. By refusing to use superior methods road congestion, and thus emissions, is greater. Safety is worse, unemployment increases, and data indicates the health of newborns is adversely affected.

    Book title is Last Exit, I highly recommend it.

  2. Cosmo Kramer says:

    Why argue “who would build us roads” and instead argue what transportation would look like altogether without government? And maybe the private sector would build roads with far fewer stop lights/signs?!?!

  3. Dan says:

    Rush hour is nearly 24/7 where I live. I’m walking down Santa Monica Blvd right now and there is a ton of traffic for people trying to get on the 405. Most of it is created by the constant construction that’s been going on for that past 2+ years with no end in sight.

    • Joseph Fetz says:

      I used to live on Armacost Avenue, and every friday I would have to take the 405 down to Vista to practice with my band. It always amazed me that there were like 8 lanes going each way on the 405 (near the airport), yet traffic was still bumper to bumper. It would take me about 2 hours just to get to Huntington Beach, but after that it was clear roads all the way to North County.

      • Dan says:

        Yeah I’m like .5 miles further up Santa Monica towards the 405. With the construction the past couple years it is awful getting on that thing. I try to avoid heading east of the 405 if I can avoid it.

        • Joseph Fetz says:

          When I used to live in LA I would have to head east every morning for work along the 10. At that time the lane-lines were so faded on the 10 that there simply was no way to know which lane you were actually in, and because it was morning, having the sun shining directly in your eyes wasn’t helpful. It was quite an amusing commute, everybody heading east on a 3 lane highway, with no lane-lines to give you guidance, and the sun shining in your face, yet somehow I never saw an accident.

  4. Jonathan Finegold says:

    The public goods argument for roads that I’ve seen is that the marginal cost of the road is zero (with high fixed costs), so the price of road usage should be zero. I don’t know if I buy the argument, but maybe I haven’t thought about it enough.

    In case anybody who reads the comments knows the public goods argument for roads well, maybe they can make a more complete case. But, it seems to me that the fixed/marginal cost distinction here suffers from some conceptual frictions, because of the physical nature of the good. If it were possible to build the road in the same way as divisible goods, where we would produce an additional unit of output at the margin, then there would be no fixed cost case. But, since roads can’t be built in this way — instead the entire cost has to be paid for when the road is built –, we have the zero marginal cost problem. My point is, are marginal costs really zero? (“Obviously,” they are, but I’m questioning whether economists should really interpret it this way.)

    • Tel says:

      The public goods argument for roads that I’ve seen is that the marginal cost of the road is zero (with high fixed costs), so the price of road usage should be zero. I don’t know if I buy the argument, but maybe I haven’t thought about it enough.

      That’s the same argument used for software, BTW.

      Several different approaches have happened there:
      * GPL or BSD license with support fees.
      * Proprietary license with license fees
      * Software as a service with monthly fees

    • Tel says:

      But, since roads can’t be built in this way — instead the entire cost has to be paid for when the road is built –, we have the zero marginal cost problem.

      That’s not exactly true. Firstly roads do require ongoing maintenance, and secondly it is perfectly possible to design a process of continual infrastructure upgrade where cost is spent over time rather than all upfront.

      I was chatting with someone from Equinix datacenters recently, and he was explaining what they call “Deferred Capex” which is a deliberate strategy for infrastructure expenses where you leave chunks of it planned by unfinished and you design a completion schedule that stays slightly ahead of what you sell.

      Admittedly, a datacenter is not a road, but it still fits the infrastructure point of view as far as economists are concerned. The private sector seems to be handling that type of infrastructure perfectly well.

      • Jonathan Finegold says:

        But, the problemstill persists. Say I want to build a highway between point A and B. I can build one lane at time, but within a certain range I can still allow additional traffic at a marginal cost of zero.

        • Bob Murphy says:

          JF, I’m in the middle of a work project and can’t think this through properly, but something is screwy here. I think maybe you’re not defining the correct good. It’s certainly not true that the owners of the road can provide another unit of service at zero marginal cost.

          • Jonathan Finegold says:

            If there is only one car on the I-5 in California, the cost of allowing an additional car on the highway is effectively 0.

        • Chris P says:

          Couldn’t this same argument be used in all sorts of businesses? If I build a golf course, it’s all fixed cost, my marginal cost is zero, therefore golf courses should be free?

          • Keshav Srinivasan says:

            The public goods argument is about goods that are nonrivalrous and nonexcludable. Nonrivalrous means that there’s no marginal cost (or low marginal cost) in letting an additional person consume the good. Nonexcludable means that it’s hard to stop people from consuming the good. People argue that roads are a public good because there’s a high cost in excluding people by putting toll booths every X miles.

            In contrast, a golf course, even though it’s nonrivalrous, is easily excludable: you just need to put a fence all around it, and then put a gate or a door for people to enter it. That requires relatively little cost compared to continually stopping people at toll booths. (and it’s mostly a fixed cost, the marginal cost is fairly low).

            • Jonathan Finegold says:


              See for example Douglas Allen’s The Institutional Revolution, where he describes the problem as one of non-rivalry. So, it’s not that roads can efficiently collect payment, it’s that economic theory suggests they shouldn’t require payment at all. (Which is probably why sometimes tolls are allowed to exist until the initial cost of building are covered.)

              • Tel says:

                But roads are rivalrous, as the headline of the article points out… traffic jams happen. That’s rivalry right there in physical manifestation.

                … probably why sometimes tolls are allowed to exist until the initial cost of building are covered.

                Check out the Sydney Harbour Bridge, it was a government project and the toll has paid for construction about 8 times over, but we still have a toll because no government would ever toss away a revenue stream.

              • Jonathan Finegold says:


                I didn’t say that tolls are always eliminated when revenue = initial fixed cost of production. I say that this might explain why they sometimes are. I’m aware of public choice problems

                And, yea, that’s why (in another response to you) I suggested a range of non-rivalry. Maybe it makes sense to charge at the point where the road becomes rivalrous and marginal costs > 0.

              • Jonathan Finegold says:

                But, think about this, when traffic on highway 1 accumulates to the point where MC > 0, a competitor may choose to enter the market with highway 2. This might drive down the accumulation of traffic back to where MC = 0 (and this may be one reason why some economists suggest roads may be underprovided for).

              • Tel says:

                Let’s suppose I have a product that costs me $10 to make and I have two potential customers. One is willing to pay $100 for the product, the other is willing to pay $50.

                So if I sell to both at the marginal price they are willing to pay I bring in $150 gross and make a profit of $130. However if those customers ever get talking to one another, the guy who paid $100 will be pissed off and never talk to me again.

                I could sell both at $50, and make a profit of $80, or I could sell one at $100 and make a profit of $90 and tell the other guy to get stuffed.

              • Keshav Srinivasan says:

                As I said, being non-rivalrous is not enough to be a public good. It needs to be non-excludable as well. Look up “club good”, which is the term for a non-rivalrous excludable good. Club goods, like golf courses and satellite TV, are NOT public goods.

              • Jonathan Finegold says:


                Right, that’s my point — it’s not a straightforward public good argument.

            • Major_Freedom says:


              Roads are excludable as well.

          • Jonathan Finegold says:

            It might be because nobody thinks the gov. building golf courses is good expenditure, even if golf courses are underprovided. But, it is often invoked in other cases, such as IP.

    • Bala says:

      The public goods argument for roads that I’ve seen is that the marginal cost of the road is zero (with high fixed costs), so the price of road usage should be zero.

      Isn’t the very concept of price being explained by costs rather problematic? I mean conceptually, I find the whole thing rather silly.

      • Tel says:

        Isn’t that the definition of a competitive market? Where prices are determined by costs, because otherwise there is room for a new business to step in and make money on the gap.

        That’s presuming no one is preventing competition.

        • Bala says:


          I thought that to an Austrian, an explanation of price based on pre-existing prices is a flawed explanation of price. You know… “turtles-all-the-way-down”.

          That’s all I was trying to point out. I am not surprised to see LK jumping at that explanation, but to see JMFC and other Austrians taking it seriously and discussing it as if it makes any sense at all sort of shocks me.

          • Major_Freedom says:


            Pricing based on costs of production does not have an infinite regress. If we go back far enough, eventually we get to raw materials being priced according to direct supply and demand in the short run, with costs of production still being the determinant in the long run.

            Crude oil is priced in the moment according to supply and demand. But over the long run, prices will tend to gravitate around the costs of producing oil.

            But once we get to final manufactured goods, costs of production play an even stronger role in the moment. You see this when you go into a restaurant and the owner will charge a price according to costs of production (plus profit), and will agree to sell zero or maximum number of meals that day, depending on the demand. In these cases, the owner will not change his asking prices with every change in daily demand from his customers.

            • Bala says:

              But over the long run, prices will tend to gravitate around the costs of producing oil.

              This does not mean the costs are the appropriate basis to decide what the price should be. This is just the explanation for the status of profit/loss in the long-run equilibrium.

      • Lord Keynes says:

        “Isn’t the very concept of price being explained by costs rather problematic?”

        Yes, for those who did not live in the real world:

        “For several years a group of economists in Oxford have been studying problems connected with the trade cycle. … Among the methods adopted is that of discussion with business men, …. and … inquiries about the policy adopted in fixing the prices and the output of products.” (Hall and Hitch 1939: 12).

        An overwhelming majority of the entrepreneurs thought that a price based on full average cost (including a conventional allowance for profit) was the ‘right’ price, the one which ‘ought’ to be charged. In some cases this meant computing the full cost of a ‘given’ commodity, and charging a price equal to cost.
        (Hall and Hitch 1939: 18).

        Hall, R. L. and C. J. Hitch. 1939. “Price Theory and Business Behaviour,” Oxford Economic Papers 2 (May): 12–45.

        • Bala says:

          Yes, for those who did not live in the real world:

          So says Lord “turtles-all-the-way-down” Keynes.


      • Jonathan Finegold says:


        It’s not a cost theory of prices. The model of perfect competition is basically an illustration of what prices would be if returns on factors are equalized. In a world with zero profits, the price is going to equal costs (it’s driven down to that point) — beyond that, it’s not in anybody’s interest to produce something that costs more than what you can sell it for.

        • Bala says:

          I do get that part, but does the model of perfect competition make any sense and therefore does a discussion of what price should be based on that help?

          • Jonathan Finegold says:

            Not by itself. It just helps get an idea of what goods/services we may need to look at more carefully. Then you’d do something like a comparative static to get an idea of what the welfare loss would be under different regimes (e.g. public v. private provision).

            There’s a strong case that, even in most markets, government involvement makes the welfare loss worse. I think you can make this case for the various industries that are regulated in some way by the government. But, at the same time we need to remind ourselves that this doesn’t imply an a generalized a priori case against government involvement.

            • Bala says:

              Why are we discussing welfare loss under a private regime? Isn’t welfare supposed to be maximised (with every individual at the highest position on his value scale) when individuals are freely and voluntarily engaging in exchanges? In that case, doesn’t such a comparative static turn out to be devoid of meaning?

              Isn’t the case to be made more like ALL government intervention, violent as it is by nature, diminishes welfare? Why do we need sophisticated models to tell us what we already know by basic economic reasoning?

            • Bala says:

              Sorry about the multiple posts.

              I thought the a priori case against government involvement is that it is a case of violent exchange and hence certainly lowers the well-being of the victim with no clear statement possible on the overall cost-benefit situation on account of the impossibility of interpersonal comparison. Why do we need sophisticated models to explain this to us?

              • Jonathan Finegold says:

                Because not everyone (including me) thinks that’s the right way to look at it.

              • Bala says:

                Why would it not be the right way to look at it? What’s wrong with it? Well-being being a subjective appraisement, how are we to “measure” welfare and engage in meaningful welfare comparisons in situations involving violent exchange?

              • scineram says:

                I subjectively value current public roads over who-knows-how-priced private ones.

  5. Ken B says:

    Near Toronto there is a privately owned highway, the 407. (It was built as a joint venture and the govt interest was sold.) Everyone who uses it grumbles about the cost but loves the road. And yet I hear constant complaints about the very idea of private roads and how awful it is the govt sold its half and how iniquitous the idea of building more such roads really is. It’s sometimes very disheartening: we have a shining example of the idea’s success, everyone benefits, and yet the idea is unpopular in Ontario.

    • Major_Freedom says:

      It’s probably because people who drive on that highway can see the costs of their road usage much more clearly as they get a bill in the mail detailing how much they drove and what they owe.

      Same thing is true where I live. If taxes paid for the roads, it’s much harder to see what people are being charged, and those who make little money will believe public roads are cheaper, because their share of the total cost is a lot less. But for the wealthy, they easily grumble about how much taxes they pay, especially for stuff they don’t even use as much as those taxes would suggest.

      I wouldn’t call the option in Ontario “unpopular”, because we should be looking at people’s feet, or rather their wheels, not their mouths. Many Canadians like to bitch about stuff, despite the fact that they tend to be nice.

  6. Gene Callahan says:

    “This isn’t a rare thing that happens on the day before Thanksgiving; it happens twice every workday.”

    Look at the BQE, and your case gets much stronger: it is one ongoing traffic jam that has lasted since I moved into Brooklyn, at least.

Leave a Reply