Hacker News Reader

No, the new demand is induced.

Among the many things people consider when buying a house or choosing a new place to rent, is the commute time to wherever they might need to go.

A suburb with an hour commute from the job is low on most people's list. Here comes a new highway, or new highway lanes, and now it's only 30 minutes. Great, let's move there... along with 20,000 other people and within a couple years the commute is back to an hour.

Busy roads are busy because of existing demand, new demand can be created by new roads/lanes and cause the expanded roads to become busier, but that's a really good thing not a bad one. New roads should be planned/expanded with that in mind because the moment existing demand starts to exceed what the current roads can handle it becomes time to build new roads or expand existing ones again.

Allowing outdated and inadequate infrastructure to make people miserable and hinder a thriving local economy is a terrible idea. Improving and adapting our roadways to meet the needs of a growing community is exactly what we want cities to do. We shouldn't use induced demand as an excuse to neglect that responsibility, we just have to understand the effect so that we can plan for it.

Let's combine two things:

1. the observation that, AFAIK, there are no instances anywhere in the world where adding roads or lanes to roads resulted in over-capacity for the roads

2. your claim that this is because of existing demand that was not previously being met

This would appear to imply that the existing demand - which I'd like to call "implicit demand", because it was not visible via behavior until the roads were expanded - is in fact so large that to meet it would require building a road system of so vast a scope that we clearly would not want it.

Put differently, if building roads merely exposes existing demands, and the 10-100x fold increase in roads over the last, say, century, has not yet met those demands, then we should conclude that it is either impossible to meet those existing demands or will require basically the total destruction of so much land that it will be untenable.

Mostly I think the idea of induced demand should encourage us to increase the availability of transportation in “novel” ways as opposed to just adding more road. You can add more road forever and just get more traffic. If instead you add rail, say, you satisfy demand and avoid a concrete highway hellscape.

I 100% agree. Too often though it's used as an excuse to not update roadways in places where there are zero efforts being made to introduce new public transportation.

It's also worth noting that induced demand exists even for public transportation systems. A single rail line can solve the bandwidth problem for a longer time than a new lane on a highway, but if growth continues you'll still end up needing to add new trains and rail lines or add additional modes of mass transportation.

Induced demand is never a reason to leave the problem of bad infrastructure unaddressed, it's just something we have to take into account as we try to improve things.

There were suggestions (pre-Covid) that the Elizabeth line, which increased east-west capacity in London by some ridiculous figure (over 50% if I recall correctly) could be saturated by induced demand (or if you prefer the realisation of suppressed demand) within six months of opening.

Those also have the same induced demand effect people are talking about. It's just that they can absorb a lot more usage than roads.

Adding a passenger rail is in principle not very different from adding a 10-lane corridor. Including the problems of what those people do to enter or leave it, and the fact that it too can become overloaded.

That said, yes, of course the rail is a much better option.

It is more complicated than this... There is a trade off while going to popular places. An Opera house is a big attraction that you tend to get tickets for so you're willing to spend some extra time in traffic for it... Now lower the time in traffic and suddenly you're deciding to get a meal a nice restaurant in the area too. But people not going to the Opera will also go to that nice restaurant that the travel cost has lowered. And because this is the US and cars are king, each one of us is taking our own set of 4 wheeled transportation. And that transportation needs to park somewhere. And this is where things tend to fall apart quickly. At the end of the day it would be far cheaper and easier for everyone if we were taking public/mass transportation to these busy spots, but because we don't design our lives around this it becomes impossible.

It's also true that once a city finally stops ignoring the inadequate bandwidth of an extremely busy road and expands it, new businesses and attractions may want to be built along that more pleasant road leading to more people wanting to take the road.

I agree entirely that a great public transportation system is a much better solution, but for places that don't have one and aren't going to create one, insisting that induced demand makes expanding roads inappropriate is just... wrong.

Ultimately, if falls on cities to provide people with a way to get to where they want to be and leaving bottlenecks in place makes people miserable and stifles the local economy.

> It's also true that once a city finally stops ignoring the inadequate bandwidth of an extremely busy road and expands it, new businesses and attractions may want to be built along that more pleasant road leading to more people wanting to take the road.

Once you get out of your car, a street with more lanes of traffic is far far less pleasant to be in than a street with less. To use an example from Philadelphia near me, which of these streets would you rather spend time in? Vine Street, with 3 lanes of traffic in each direction and a freeway below in the center median?

https://maps.app.goo.gl/3KqvmfbxLcN8wQtp9

Or the much narrower 17th street, with only two lanes total?

https://maps.app.goo.gl/ozmEtLGfh2hojyVo7

As someone who lived nearby for about 7 years, I know which I'd choose.

You'll notice that most of things on the two lane road are intended for people (shops, banks, churches) unlike the first road where the things are mostly for cars (gas stations, parking lots).

There's nothing wrong with big roads designed primarily for cars that bring people into or away from quieter areas designed primarily for people where they can walk around and shop.

> There's nothing wrong with big roads designed primarily for cars that bring people into or away from quieter areas designed primarily for people where they can walk around and shop.

There is in cities, because there those two things cannot be so neatly separated.

Some historical context here: Vine Street cuts through Chinatown. Prior to its expansion, there were homes and businesses in the space that's now roadway. They were demolished to make way for the car.

The scars of that are still visible today; if you walk through the neighborhood it's clear that Chinatown north of Vine is struggling compared to its counterpart on the south side, and the two sides feel cut off from one another.

And it makes crossing more dangerous. My commute for years involved crossing Vine, and though I sometimes would bike I had to plan my route very carefully because there were few places where doing so would be at all safe.

People live and work on either side of that street. There are senior homes within a few blocks of it.

> and stifles the local economy.

Oh no, we're going going to stifle the local economy for billions of dollars, and all we have to do is take on 10s to 100s of billion of dollars of debt to avoid it...

This is just turning into a hedonistic treadmill of turning more and more of your space into transportation infrastructure by means of debt.

It's hardly hedonistic to want to leave your house and investing in transportation infrastructure should pay for itself.

The point is that automobile infrastructure does not pay for itself; it is too energy-inefficient, too upkeep-inefficient, and too space-inefficient. Cars do not scale.

Right, and I think what you're getting at is that there is a distinction between the generated demand that the new lanes induce, and the latent demand which was already present before the new lanes were added but went unsatisfied.

Adding new lanes to the road allows that latent, pent-up demand to be satisfied--the induced/generated demand is the mechanism of that satisfaction.

>The "demand" already exists and simply isn't being met by the inadequate infrastructure.

the induced demand argument isn't simply that we should never build transportation infrastrucutre, it's that adding a lane to an existing road is a bad way to meet that demand because the traffic that one lane can carry is insufficient to exceed the demand that one lane will induce. yes, the road expansion will be "successful" in that people will use it. but if it was pitched as a way to improve travel times, reduce congestion, or improve safety, that will probably be a failure.

if you're not going to the opera because the road is too busy, and the added lane releases that unmet demand, the road will still be just as busy as before and you still won't get to the opera. everybody else is doing the same math as you, and deciding that if the road isn't too busy they're going to drive. and so the road is always too busy.

The "induced demand" isn't a bad thing. It means more people are getting some utility out of them. The reason why folks bring up "induced demand" is when someone says that roadways should be expanded to reduce commute times. Commute times will often not be reduced: only that more people will be able to commute.

Usually the solution then is not to optimize for road transportation. After a threshold, it makes sense to switch to a more efficient use of land. Essentially, if you added two more lanes to 101, it's still going to take you as long to get to Palo Alto.

Roads are very good things. But they are very costly land-use-wise. So ideally you have lots of roads as a discovery mechanism and then place high-frequency fast transit along the most well-used. Most of America is in regulatory hell, though, so transit is slow and low-frequency, so the road may actually be net better.

To reduce congestion you need to speed up and expand the slowest alternative, public transport. That is the only proven method to reduce congestion.

The peak demand for car travel is way beyond the infrastructure which is possible to construct and maintain.

More info:

https://youtu.be/RQY6WGOoYis