An article on a highway project in the Pacific Northwest caught my eye:
However, the shiny new document leaves out an essential consideration when it comes to projecting the future effects of I-5 expansion in this long-constrained corridor, an omission that would have been much less noticed in a decade ago but which sticks out like a sore thumb now. It almost completely sidesteps the concept of induced demand, which posits that additional roadway capacity will prompt more trips as road users seek to take advantage of faster trips, ultimately cancelling out many of the promised benefits that come from adding that new capacity, especially congestion reduction.
Opponents of a new and bigger bridge connecting Vancouver, Washington and Portland, Oregon claim that it would cause more people to use the bridge. Supporters of the project assume that there would be no increase in the number of cars crossing the bridge. That strikes me as kind of odd.
Consider the following analogy. A movie theatre is so popular that it often completely sells out. The management committee is considering an expansion of the movie theatre. One group claims that an enlarged movie theatre would attract more patrons. The other group claims that enlargement of the theatre would not result in any increase in movie attendance. Which of those groups would you expect to support expansion, and which would you expect to be opposed? Do you see the problem?
Of course there are many differences between movie theaters and bridges, and I promise we’ll look at those differences. But I first wanted people to consider how odd it is that the opponents of highway expansion projects are typically the same people that believe it would induce more demand for its service.
Supporters of bridge expansion are typically political leaders who wish to cater to their electorate. There are two kinds of voters, those who pay attention to the bridge expansion issue, and those who do not. I suspect that there is a strong correlation between voters who support bridge expansion and those who already use the bridge, if only because they are probably better informed about the situation than other voters. When supporters of bridge expansion deny that there would be induced demand, they are implicitly suggesting that all of the benefits would go to existing users in terms of less traffic congestion. But that outcome seems extremely unlikely, as it violates the law of demand. When an increase in supply makes something cheaper (in terms of the opportunity cost of time), it leads to greater quantity demanded. There would be induced demand.
Opponents of bridge expansion also have an incentive to cater to voters with the most intense interest in the issue. They may wish to argue that the bridge expansion won’t do any good at all, as it would induce so much extra demand that traffic congestion would become just as bad as before. But that argument also violates the law of demand! If there were no reduction in traffic congestion, then what would induce any new drivers to start using the bridge? (In fairness, the author of this article does not claim that induced demand would prevent any reduction in congestion, but I’ve seen others make that claim.)
One side is essentially arguing that demand curves are perfectly vertical, and the other is implicitly arguing that demand curves are perfectly horizontal. In fact, demand curves slope downward.
So what’s the answer? Should the bridge be built?
Elsewhere in the article, the author makes it clear that his opposition to bridge expansion is linked to environmental concerns. Ideally, you want to have a Pigovian toll to reflect any sort of traffic externalities, including congestion, pollution, global warming, suburban sprawl, etc. If that toll were in place, then it would be easier to evaluate the project on a cost/benefit basis.
(Although even in that case there might be other complications, such as indirect effects on the usage of other roads that do not have Pigovian tolls. So I don’t mean to suggest that a Pigovian toll on the bridge completely solves the problem, rather that it makes it easier to evaluate the pros and cons of a new bridge.)
PS. In previous posts I suggested that Vancouver, Washington was an attractive place for libertarians. You can work in a state with no state income tax (except capital gains), and shop in a state with no sales tax. And the Pacific Northwest tends to be pretty liberal on social issues like drugs, abortion and right to die. So perhaps we also need to consider whether this bridge would allow for the expansion of the little libertarian paradise in southwest Washington.
Here’s a picture of Vancouver, with beautiful Mt. Adams in the background.
READER COMMENTS
David Henderson
Nov 11 2024 at 3:42pm
Great post.
Dale Doback
Nov 11 2024 at 6:33pm
I’m not following. To me the proponents of the movie theater expansion claim that existing movie goers will be able to see more movies per week. The opposition claims they will see a similar number of movies per week. I don’t see how they’re making different claims about the shape of the demand curves when the claims are rates per unit time, not quantities.
Scott Sumner
Nov 11 2024 at 7:20pm
I don’t follow your question. What’s the difference between quantities and rates per week?
Matthias
Nov 11 2024 at 8:29pm
Related: it’s almost impossible to explain that if there’s congestion, then tolls and congestion charges are free for the average user.
More generally, if there’s a service with a queue, usage of the service is rationed by who’s people’s willingness to put up with queues and congestion etc. It’s like an implicit auction.
Unless there’s collusion, this auction will tend to soak up the customer surplus people get from eg using the bridge.
If you increase prices or charge a toll, the bids in the implicit auction are now baskets of money plus queuing hassle. The auction will still soak up the customer surplus, but at least the bids aren’t completely made up of wasted effort.
For the average user, the increase in monetary fees is pretty much offset by the decrease in waiting around. (For someone who values their time a lot more or a lot less than the average user, the total value of a winning bid would change.)
The same applies not just to toll roads and bridges, but also to eg long waiting lists for medical procedures.
john
Nov 13 2024 at 5:51am
It’s almost impossible to explain what if there’s congestion?
Jose Pablo
Nov 12 2024 at 7:29pm
I don’t know about “induced demand”. You don’t cross bridges just because there is no congestion on them. You still need a reason to cross the bridge. I don’t think that people from Vancouver and Portland spend their time looking for non-congested bridges to cross them just for the fun of it (although you never know).
“Induced demand” can only came from already existing reasons to cross the bridge that a) provide less utility than the opportunity cost of the congestion and b) the utility they provide is time dependent, since congestion also varies greatly with time.
This “time dependent latent demand” should be there if the new bridge is going to “induce any demand”. In the absence of this latent demand, the new bridge “per se” won’t generate any “induced demand” (the demand curve would be, in fact, perfectly vertical).
A most likely bigger effect of the expansion will be a shift in time of demand. From the now less congested hours to the now more congested ones, undoing the relocation that the actual scarcity had very likely caused. The effect will then be “inducing demand” on the now most congested hours at the expense of the existing demand in the now less congested hours.
Scott Sumner
Nov 13 2024 at 3:56pm
“You don’t cross bridges just because there is no congestion on them.”
Of course you do. It’s Economics 101. Supply and demand. Have no idea what you are talking about.
Jose Pablo
Nov 13 2024 at 7:54pm
You have no idea what I am talking about.
If you build a bridge between the Kalahari Desert and Point Nemo in the middle of the Pacific nobody will cross it. Not even people well versed on Economics 101. The demand is not for bridge capacity. Is for the activities that require to cross the bridge. And the new capacity could only marginally change the demand for this kind of activities (for instance if the main effect of the actual constraints is just to shift demand for this “crossing-the-bridge-activities” from the rush hours to the less crowed hours).
In fact your example of the movie theater is an interesting one. The main effect of the movie theater expansion could very easily be to relocate movie goers from the most inconvenient showing times they were previously forced to attend) to the more convenient ones (the ones that were over crowed before the expansion). The total increase in movie goers can very easily be small despite the increased capacity. What really increase the demand is the quality of the movies. Compare with that the “induce demand” of the movie theater capacity expansion could very easily be irrelevant.
Grand Rapids Mike
Nov 12 2024 at 10:08pm
Regarding induced demand. The Chicago beltway highway I-294, seems to be a classic case of induced demand. The beltway, which starts in Indiana from I-94, goes thru Illinois and then to Wisconsin, has increased its supply of available lanes in the last 30 years or so from 2 to 3 t0 4 lanes and now a 5th lane is under construction. Each increase in lanes/supply has resulted in corresponding traffic increases, especially during rush hours, the supply increase has created more demand.
robc
Nov 13 2024 at 10:10am
No, there is no new demand, there was a shift along the demand curve. Quantity increased with the shifting of the supply curve, demand remained the same.
Jose Pablo
Nov 13 2024 at 8:33pm
Interesting way of looking at it.
But the relevant supply curve is not the “capacity of the bridge supply curve” the relevant supply curve is the “activities at the other side of the bridge supply curve“. The effect of the bridge expansion is to reduce the cost of these activities at the other side of the bridge by reducing the “hours sitting in traffic” part of their cost. But if the “hours sitting in traffic” cost is very low compare with the total cost of the activities at the other side of the bridge, the increase in quantity demanded due to the shift of the supply curve will be negligible.
Scott Sumner
Nov 13 2024 at 3:58pm
Sounds like a highly successful project.
Luc Mennet
Nov 12 2024 at 10:39pm
The term “Induced Demand” irritates me to no end. There’s no demand being induced, the demand curve is exactly where it was! The only thing moving is the supply curve being shifted. I know that Libertarian adjacent types have a bit of a bad habit of smugly saying their opponents “don’t know basic econ,” but come on, but I don’t think you could possibly find an econ 101 textbook that wouldn’t clear up this misunderstanding in a single graph. The only thing that really complicates this is that price is denominated in “hours sitting in traffic” instead of a currency.
I think so much of the discussion around this is people kind of smugly bringing up the concept of induced demand not necessarily as something they feel they have a good intellectual grasp over, but more just a way for them to feel smarter than all those dumb highway planners who OBVIOUSLY just don’t know what induced demand is. /s
I think that there are some interesting debates to be had in this conversation, both when it comes to environmental concerns as well as more locally practical ones (ie for some projects more lanes might not meaningfully increase throughput as the actual bottlenecks are on the ramps) but that kind of discussion is so rare by comparison.
Jose Pablo
Nov 13 2024 at 8:20pm
Very interesting!
The only thing moving is the supply curve being shifted.
And maybe not even this since the “hours sitting in traffic” “price” could be but a very small part of the total “utility” of the activities demanded at the other side of the bridge, if this is the case, and it very well could be, the shift of the supply curve will be minimal.
And, in fact, there is a “substitution effect” between the different times for crossing the bridge. So the “price” would be the “change in utility for changing the time of the activities at the other side of the bridge” which should be, by definition of “substitution”, even a smaller part of that total utility than the price of “hours sitting in traffic (at rush time)“
Mactoul
Nov 12 2024 at 11:02pm
Both proponents and opponents of the bridge suffer from anti-growth bias. In reality, a road or a bridge should be built if the building leads to more usage of it. That would be efficient investment
Juan
Nov 13 2024 at 5:58am
There’s sometimes an approximation by physicist where if a number is very large relative to another number, that very large number is effectively infinite. 🙂
jj
Nov 13 2024 at 7:43pm
Luc above makes a good point, but setting aside the sloppy terminology of ‘induced demand’, the reality is that there is additional induced driving. However what the road construction opponents are ignoring is that all of this extra driving enables benefits which the drivers are enjoying. The same number of people living further out means more driving, but also bigger lot sizes. The same number of people driving further for work means a bigger employment area, and better job matching. The same number of people driving further for entertainment means more entertainment options for them. So yes, maybe constructing roads does not reduce congestion — it just improves everything else.
Grand Rapids Mike
Nov 15 2024 at 4:41pm
Will add another wrinkle to this subject. Going way back in time one of my economic papers was on the impact of the on and off ramps of I-94 from Kalamazoo up to the first exit to Ann Arbor. At each point of the off ramps there was at least 1 restaurant, gas station at each ramp and more where the ramps lead to a city. Another example of the induced demand impact of freeways, an obvious point. When I started the paper was nor that sure of the impact as became obvious.
John
Nov 17 2024 at 3:36am
Wikipedia says…
The University of California at Berkeley published a study of traffic in 30 California counties between 1973 and 1990 which showed that every 10 percent increase in roadway capacity, traffic increased by 9 percent within four years time. A 2004 meta-analysis, which took in dozens of previously published studies, confirmed this. It found that:
…sounds like one of them is right.
Comments are closed.