Matt Yglesias has been making the point consistently for a long time, and he should keep drilling it in until its obviousness becomes apparent: traffic is caused by improperly priced roads.
As he’s said on a few occasions (can’t find the exact posts right now): when we see video of long bread lines in the former Soviet Union, we know why they exist. They exist because the price of bread has been set artificially low. Since the price is artificially low, more people go hunting for bread than would go if the price were allowed to find its level. Since the price is artificially low, companies produce less bread than they would if the price were allowed to find its level. There’s a mismatch between the number of people seeking bread and the number of loaves of bread available for sale.
But we never think of that when we see traffic, for some reason. Clearly, though, the same mechanism is at work: there’s a scarce resource (slots for cars on the road), there’s a fixed supply, and there’s a certain demand. If people had to pay more to ride on the roads, they’d presumably drive less. What you want is that the last person to get on the road is just willing to do so — if the price were just a tiny bit higher, he’d find other means of getting to work.
So the roads are clogged, and traffic is unbearable, because prices aren’t being set by the market. They’re being set artificially — at zero, in fact — by the government. As Jamie Galbraith put it in another context, “this process is so simple that the mind recoils from it.”
If libertarians should be shrieking about anything, it should be the artificially low price of driving. The environmental consequences of this are stark: more people get on the road than would otherwise, so more smoke goes in the air; more people drive than probably would otherwise, so development patterns change to accommodate cars; so the landscape gets scarred with new, wider highways, new subdivisions, new Wal-Marts, etc.
I can’t let my speculation run too rampant there. Suppose roads were privately run. Maybe people would have no problem paying more; as it is, they seem to tolerate longer and longer commutes, which impose a less-obvious but still great price on drivers (less time with their families, more frustration, road rage). Maybe private companies would build extremely wide roads, thereby incurring a large fixed cost but basically negligible marginal cost for each driver on the road. The cost of driving might then look like the fixed cost amortized over the expected number of drivers, plus some profit margin for the corporation. And maybe that cost wouldn’t be so high.
But the point is that we don’t know, and the continually worsening state of traffic gives us reason to believe that the price hasn’t been set appropriately. If it were any other commodity being allocated, the presence of long lines would indicate a failure of central planning. Somehow Americans don’t jump to that conclusion with their roads, and they should.
Yglesias seems to have taken up a [foreign: Carthago delenda est] approach with traffic pricing. I’ll join him in it.
But how can you “market price” roads? Roads are natural monopolies. There’s only so many roads that it makes sense to build between places, usually only one or two. I don’t have a choice of competing roads to take in front of my house.
So, without some kind of government interference, the roads would be subject to most of the problems of monopolies.
Why not just let the government control the roads like they do now, but charge more appropriate fees for their use?
They do charge for the roads, though. They tax the citizenry to pay for the roads. There is zero apparent cost, though, because they don’t tie the charges to usage. A use-tax for roads would be interesting, though. Say if all car owners had to report mileage every January and pay a marginal tax based on use-levels?
Gas tax is by far the simplest way to do this. I really don’t get why people might support a use/mileage tax but not a gas tax.
Another good option is congestion pricing. My vaguely-informed impression is that it has been successful in London.
I don’t have permission to edit my last comment, and WordPress did something weird with the URL, converting the underscores to italics. Anyone know how to fix that?
Here, I made my own personalized shorl:
Hell, you wouldn’t even have to add the real cost of road use to the gas tax to alleviate our traffic woes. Just adding in the negative externalities associated with burning the gasoline would probably be enough! Then you could rebate that back to people by income to counteract regressive effects and there would be no net increase in taxes.
Yeah, talking about a fully free-market solution was wrong. Setting the pricing more appropriately would be better. Higher prices when traffic is higher, lower when it’s lower. Maybe outsource the job of setting up payment systems to a private company, and give them a fraction of the proceeds. This will encourage them to set the revenue-maximizing price. (I’ve not thought this through. It’s just off the top of my head.)
WordPress converts underscores to italics because each comment uses Markdown syntax, which does exactly that.
Measuring negative externalities seems hard, no? Whereas empirically measuring, say, the price-elasticity of demand seems easier.
A use tax or a gas tax sound like fine ideas. Just taxing people isn’t nearly good enough, though: right now I pay the same for the roads as everyone else, but I use them way less than most people. So the current system isn’t proportional to use. (An obvious observation, but I wanted to make sure it was clear.)
In Seattle gas taxes only pay for about 5% of roads. There’s no local vehicle tax and state vehicle taxes are low. The rest are property and sales taxes, federal grants, etc. I suspect most other American cities are about the same because of the usual tax competition/race to the bottom reasons.
I agree with Adam on this. In Houston, we considered a variable pricing policy (aka congestion pricing). This seems to make good sense. On the potential monopolies of roads, the only way to counteract this would be to support alternative transportation methods. Of course, these require substantial subsidizing (consider, rail) to make them possible. I also agree that the the tax/mile is a confused position: it is prohibitively expensive (using gps recorders in cars or some such) and it provides a disincentive for fuel economy.
Oh yeah, and here’s a libertarian who makes your argument, Steve: http://www.freedompolitics.com/articles/traffic-898-private-california.html
Not true — you don’t pay the already substantial gas taxes, right? (I suppose you do pay them to the extent they are built into the prices of things you buy that were transported on the roads, but you certainly pay far less than someone who drives.)
Yeah, I don’t pay the gas taxes. I was thinking that I pay for roads through regular income taxes. I’m curious what fraction of state and federal roads are funded through gas taxes. This website says we pay 23.5 cents per gallon towards gas tax in MA:
As it happens, it looks like average fuel economy is just about 22.5 miles per gallon:
So we pay about a cent per mile in gas taxes.
I’m a month late,but I have to post my disagreement with this. Firstly, I’m not sure I follow this:
“What you want is that the last person to get on the road is just willing to do so if the price were just a tiny bit higher, he’d find other means of getting to work.”
But surely what you actually want is for the price that drivers experience to drive on a particular road to be the cost to the person selling road use to one extra driver. But that cost is so negligible as to be zero. Think about it – it doesn’t cast the owner of the road anything meaningful to let one more car on his road (unlike a shop selling an extra widget which requires running a machine/funding labour to produce one more widget). Therefore, the price of the road should be zero.
Then later, you and others commenting are talking about negative externalities like burning fossil fuels. And yes, these should be dealt with, either through cap-and-trade, or a carbon tax. But these aren’t inherent in driving (indeed, the point of implementing one of these systems is to allow the market to move us from fossil fuels to non-climate changing fuels).
On the other hand, there are gigantic positive externalities inherent in driving itself. Someone driving down the road doesn’t just benefit that someone. It benefits all of us economically too. Maybe they’re going to work – good for the economy. Maybe they’re taking their kids to school – allowing the workers of the future to get educated. Maybe they’re delivering something for some company – powering the economy. And on and on.
And with a positive externality, the right response is to subsidize it. In this case massively. Which means taxes not associated with driving paying for investment in roads (or, perhaps more appropriately for these economically tough times – stimulus money). The government ought to be doing the same thing with public transport too.
I also disagree with this:
“right now I pay the same for the roads as everyone else, but I use them way less than most people. So the current system isn’t proportional to use.”
You seem to be making a taxes paid per capita argument here, but that’s only an average. Depending on where people are in the income scale and so on, people pay different amounts of tax. In fact, if everyone was contributing the same for roads, I would say that was unfair as it would be what we in the UK call a poll tax (of Margaret Thatcher and John of Gaunt fame – I understand that the term “poll tax” in the US is used more in relation to the Jim Crow laws and so the word has a different symbolism there).
It’s not entirely clear why the fact that “the current system isn’t proportional to use” is necessarily a negative. Lots of government services are free at the point of use. For instance, the police. In fact, in certain circles, it can be seen as the hallmark of a civilized society for certain services to be free at the point of use – like health care here in the UK.
Also, I don’t agree with this:
“Maybe outsource the job of setting up payment systems to a private company, and give them a fraction of the proceeds. This will encourage them to set the revenue-maximizing price.”
There are a number of problems with this:
As pointed out above, roads are a natural monopoly. If you outsource, and incentivize the private company to maximize revenue, then it will abuse that monopoly position in just the ways ordinary monopolies do.
Corporatism. Unless there are significant efficiency gains from outsourcing, I see no reason why a specific business should leech off of government.
The negative effect of outsourcing on people’s relationship with government. I seem to remember Tony Judt wrote something on this before he died.
There’s nothing special about the private sector efficiency wise. This is, I feel, the number one lesson that I wish neoliberals would learn. Yes, if you have a number of private clothes shops competing with each other, then society will get the best result. But have just a single private company, and why exactly would the private sector set the most efficient price? A firm is just as centrally planned as a government department. Ultimately, the private sector is best at what its best at, and the public sector is best at what its best at.
Just like there are market failures and government failures (public choice theory), there are “quasi-market” failures. See this paper for some detail:
Finally, I’ll point out that road pricing has legitimate civil liberties problems.
Oh, and obviously this has nothing to do with merits or anything, but road pricing is politically toxic over here:
(note that the plans described were eventually dropped). I suspect that in America, there would be plenty of people jumping up to attack the idea as “socialist”.
Sorry for the long comment, and for being so negative. I’ve been away and trying to catch up on blogs I’ve missed, and I had to comment.
“But surely what you actually want is for the price that drivers experience to drive on a particular road to be the cost to the person selling road use to one extra driver. But that cost is so negligible as to be zero.”
That’s not so clear to me. I think that’s sort of assuming the equilibrium of a perfect market, no? In such an equilibrium, price = marginal cost. I’d think the reasonable standard for what people should pay to get on a road is what it’s worth it to them to get on the road, no?
Secondly, though, I think your estimate of the costs understates the negative externalities from traffic. I believe this is the Wired piece where they discussed one guy’s model of NYC traffic, wherein he finds “that every car entering the [Central Business District] causes an average of 3.23 person-hours of delays. Multiply that by $39.53a weighted average of vehicles time value within and outside the CBDand it turns out that the average weekday vehicle journey costs other New Yorkers $128 in lost time.”
The argument that there are significant positive externalities from driving doesn’t make sense to me. The argument is that people driving cars — or taking trains, or whatever — are often on their way to engaging in productive economic activity, which is a net social gain? I mean, eating food helps people engage in productive economic activity; should we subsidize food on that basis? Seems like an odd argument for externalities.
The argument that taxes for services ought to be, in certain conditions, spread evenly across the society, even across those who don’t use the services, makes sense to me in general. That’s the argument for health insurance, and I certainly believe in what they call “community rating” (same premium charged to everyone). I’d need to think about why I don’t think that’s a good idea for traffic. I’m pretty sure one can find something that differentiates the two cases.
“As pointed out above, roads are a natural monopoly. If you outsource, and incentivize the private company to maximize revenue, then it will abuse that monopoly position in just the ways ordinary monopolies do.”
But the government already maintains a monopoly here, no? So wouldn’t you expect that the government is already abusing the monopoly position? What would be worse about allowing a private company to abuse the monopoly position?
“Theres nothing special about the private sector efficiency wise.”
Yes, true, but: the private sector’s incentives here would be different than the government’s. The private sector’s incentive is to maximize profit. I could see an argument that the government wants to maximize tax revenue, but that’s not entirely clear to me.
There are also quasi-private schemes, in which private companies carry out some of the functions of road maintenance and toll collection and so forth, subject to regulations on what they can charge, how much they can raise fees in a year, etc.
“Finally, Ill point out that road pricing has legitimate civil liberties problems.”
Interesting. Like what?
Thanks for your excellent comments!
Hm. Looks like the link I posted got dropped. It’s this one: http://www.wired.com/magazine/2010/05/ff_komanoff_traffic/all/1
I wonder if the comment got bunged in other ways.