Hacker News Reader

> I’m also not sure why we need to create a false dichotomy between roads and bike lanes.

Because there is finite width for existing roads between existing buildings, and so quite literally, adding a bike lane often requires removing a car lane. It's not a false dichotomy at all, it's literally measured in meters.

> Have cities considered elevated bike lane networks on pillars...

Do you have any idea how astronomically expensive that kind of stuff would be? It's just not even on the table, realistically speaking, for 99.9% of places.

Not disagreeing with you, but elevated bike lanes do exist and they are pretty cool.

Oh yeah. No, I've seen what that "0.1%" can look like, and they're so fun to ride too... I wish the kind of money was available to put those all over!

We have some of that in my city, alongside a big freeway (see [1]) and yes, it’s ridiculously expensive. It is great though, and one of the only options given the number of very busy intersections this popular route crosses. But in the inner city, it wouldn’t make any sense.

Inside the CBD, they’ve been taking out on-street parking to put in separated bike lanes, and that‘s just as good. They completely removed car traffic from a bridge in the middle of the city also to add an extra bus land and a bike lane, and while some drivers were annoyed, it used to drop you in a extremely congested bit of the city so wasn’t very convenient in the car anyway.

1. https://www.tmr.qld.gov.au/projects/Veloway-1-Pacific-Motorw...

If you expand a road, there will be greater usage because demand soaks up supply.

However, cars are big and takes up a lot of space, especially since people regularly drive sedans that seat only one individual with max capacity for four to five people. This requires a lot of space to support cars.

Of course, you can continue to expand roads. Problem with that idea is that the roads displace other uses of land such as homes and businesses.

Roads increase utility of cars, for sure, but that comes at the expense of everything else.

It’s easy for people to forget just how spatially inefficient cars are because we’re so used to 30-40% of city space being dedicated to roads and storage.

We saw a dramatic example of this next to the bike path I ride to daycare/work: hundreds of new apartments, restaurants & other retail shops, outdoor public spaces, etc. next to a subway station. This was just part of a retail parking lot which broke 50% usage _maybe_ a couple of times a year. The number of people using the space when up by at least an order of magnitude and it’s a lot more consistent economic activity.

> If you expand a road, there will be greater usage because demand soaks up supply.

This can only occur when there is sufficient latent demand, or phrased differently, the supply of roads (and sufficiently fungible transit modes) is so insufficient that that moderate increases in supply do not create a surplus. This lack of surplus (traffic) representing an economic inefficiency which it can only be desirable to reduce[0].

What is stupifying is how 'induced demand' (read: building more roads will cause more people to perform economic activities facilitated by transit) is contorted into an argument against building roads.

[0]Until such point that traffic inefficiency is in equilibrium with others. e.g. I can't see Manhattan widening 5th ave anytime soon, but most of America isn't pressed for real estate.

The article points out that more supply of roads also leads to less density, which requires more driving, which requires more roads. This is not the short-term induced latency but for sure a part of it and part of why katy highway exists.

Lets look at the null position and agree that induced demand is bad and we shouldn't build more roads, which means the question becomes:

How many roads should we be tearing up[0], and how do we know when we've torn up the correct amount? Since we a. clearly aren't at the optimal amount of roads[1], and b. induced demand argues that we shouldn't be building more.

[0]A clearly nonsensical position.

> A clearly nonsensical position.

I don't find it nonsensical at all. My city is actively planning to reduce the amount of normal roads in the city center and introducing more & better bicycle lanes. This is very much needed, since while the central crossing does have (some) gutter bike lanes leading to it, it comes close to suicide to use them. Accordingly, no one does and many use the sidewalks that are already too small for comfort without bikes (the city is in a bad spot there, because it's very space constrained in that place. The roads are badly overloaded in the center and no lanes can be added due to the lack of space).

A nearby larger city would be well advised to remove traffic & lanes from central locations and allow more bike paths and crucially allow expanding the badly overloaded subways with trams (and dedicated tram ways) near the center. Due to various factors, there's no real option to use more tunnels. They have been routinely over the limits for particulate matter and were successfully sued over it.

Now, roads outside of cities are complicated. The larger ones usually can't be easily reduced, but eat a lot of maintenance costs that could go towards higher capacity options (rail). Looking at https://en.wikipedia.org/wiki/Braess%27s_paradox and carefully analyzing existing roads may also make sense. In general, car based travel costs society the most per mile by far. Introducing measures like tolls & car-free cities as well as making alternatives more attractive with shifted spending may reduce demand and thus enable tearing down roads.

The question hinges on a false dichotomy that the only solution to demand is more roads, which is part of what this article is arguing against.

Edit: An obvious but politically hard solution to overloaded roads is encouraging higher density.

Sedans? Try 5000lb, 7-8 passenger SUVs.

> If you expand a road, there will be greater usage because demand soaks up supply.

Only up to a limit, as I mentioned. They won’t keep “soaking up” supply. People are not looking to drive 24/7. They are looking to reach destinations and spend time there.

> Of course, you can continue to expand roads. Problem with that idea is that the roads displace other uses of land such as homes and businesses.

This seems like a vague argument. Road space displaces other uses marginally. If your population were held equal, then you can build more roads, displace some amount of other uses, and the increased traffic from those displaced wouldn’t be nearly as much as the extra traffic capacity the road provides. If your population increases or usage patterns change, then it is simply “latent” demand that is being met, since there’s an appetite for more travel throughout and increased speeds and all the utility that brings. All of this applies to anything that we can model with supply and demand, including bike lanes.

> Induced demand as a concept just doesn’t make sense. If demand is “induced” it just means that the supply wasn’t adequate to meet demand. That’s it. It isn’t a large mystery, and there isn’t any logical argument or evidence that greater supply will keep “inducing” demand infinitely. There are a finite number of people and finite minutes in a day, and if they are able to derive value from driving more, then they will do so, but only up to a limit. In return for that driving, they will also gain something for it (the value that motivated them to drive). The phrase “induced demand” is simply a part of the rhetorical war waged by anti car activists.

Right, it's not surprising but highway expansions are often sold as "reducing congestion" rather than the reality of "causing more people to drive more miles with the same resulting level of congestion."

If the actual goal of highway expansions is to make it efficient for people to drive more miles/have longer commutes then there isn't an issue with there being "induced demand," but there may be various reasons why this isn't actually a very good goal.

Highway expansions do have benefits to people who end up moving further out (more affordable housing), but there may be better ways to achieve the goals of providing housing to more people around urban areas by increasing density, or to help people get around urban areas by transit, etc.

> There are a finite number of people and finite minutes in a day, and if they are able to derive value from driving more, then they will do so, but only up to a limit.

This is the primary objection I have to self-driving cars. By enabling people to do other things while they are travelling, self-driving cars raise that limit considerably. When self-driving takes off, so will congestion.

This is very true. Hopefully the Boring company can be successful in reducing tunneling costs by a factor of 10 or 100. Then cities will just have to fined the political will to allow them to be built. Maybe public, self-driving taxies can be made to travel close enough together so that they look like and can be called "trains". Then all the public transport supporters might not try and block them even if the tunnels are open to people's individually owned vehicles also.

Here's a nice articlr thst explains how induced demand is different from latwnt demand:

https://www.strongtowns.org/journal/2021/8/24/does-induced-d...

The article ought to do more to distinguish these.

Latent demand: Driving more often because we can now actually get to places when we used to give up trying.

Induced demand: Driving more miles per day because space for the roads pushed each building farther apart over time. Going to larger stores because small nearby stores failed to compete.

Based on your tone, it seems like you are critical of the article, but your critique is actually what the article is addressing. Specifically, "Effect 2" demonstrates a unique property of "induced demand" that doesn't fit into traditional discussions of supply and demand.

I have yet to receive a convincing answer why induced demand does not similarly apply to/or is not a concern for welfare programs, homeless shelters or public transportation. I find it suspect that this theory is apparently only applied to roads, and largely by people who commute by bike.

> why induced demand does not similarly apply to/or is not a concern for welfare programs

It does in certain circumstances, like when you'd get 90% of your paycheck if you quit your job, and so people quit. Which is why smart welfare programs provide benefits in a gradual way rather than at a single cutoff -- e.g. how the Affordable Care Act subsidies were designed.

> homeless shelters

Again, it does. Cities with excellent homeless shelters and care literally attract homeless people from parts of the country without. Remember the scandal about cities literally buying homeless people bus tickets to send them somewhere else, to try to address the problem? This is a known issue, and a good justification for why social services funding ought to be provided at the national level rather than at the city level.

> or public transportation

It does absolutely. Entire cities spring up after a rail line is built.

I guess my question to you is -- what are you trying to argue? Induced demand isn't just a property of roads, and I don't know any author who says it's exclusive to roads. It seems like a straw-man you've constructed.

But the concept is often raised in conversations about roads, because many people naively believe that widening a road will solve congestion.

The article literally discusses that in the second part.

I said convincing answer. My read is that by the theory of induced demand the best way to eliminate bike deaths is to eliminate bike lanes and therefore people on bikes.

The best way to eliminate car deaths is to eliminate cars, too. But obviously we're not doing that either.

I'm not clear what you're trying to accomplish by juxtaposing induced demand with transportation fatalities, when the two concepts don't seem to have any special connection.

> The best way to eliminate car deaths is to eliminate cars, too. But obviously we're not doing that either.

As far as I can tell that is exactly the outcome the induced demand folks are hoping for, or something close to it.

I'm sorry, who are "induced demand folks"? What does that even mean?

And I'm unaware of anyone proposing the elimination of cars as serious policy. At most it's limiting vehicles in parts of dense pedestrian city centers.

People who are generally "left-leaning" on transportation envision a healthy mix between public transportation, cycling, and cars. They simply want car usage to be balanced with alternatives, rather than cars dominating by default.

Virtually everyone recognizes that cars are necessary for a wide variety of uses. They're not going away, so I don't think you need to worry about that.

It gets applied to roads because people keep trying to “solve” traffic without pricing roads. Nobody likes being stuck in traffic, but unfortunately it seems most people are also angrily opposed to tolls which would balance supply and demand and allow roads to flow smoothly (for those willing to pay). There’s a faction of people who _swear_ that if we just built one more highway we’ll _solve_ traffic. That’s who the “induced demand” argument is aimed at.

To your broader point, all services work like this to a point; if you first suppose there’s insufficient supply to meet demand, then adding supply can induce more demand. There’s an eventual saturation point for everything though. Tear down all the buildings downtown and replace them with parking lots. There will be plenty of unused parking and no more rush hour as there’s no longer a reason to go downtown.

> Induced demand as a concept just doesn’t make sense. If demand is “induced” it just means that the supply wasn’t adequate to meet demand

Supply and demand aren't fixed -- they meet at an equilibrium price. There are latent trips that are canceled, change modes, or time-shifted to deal with traffic because of the high cost.

> The phrase “induced demand” is simply a part of the rhetorical war waged by anti car activists.

Induced demand is a standard transportation engineering concept. You might argue that it's been hijacked by the anti-car folks or whatever, but tramsportation engineers and planners who want to build car-heavy transportation systems are aware of and use it. They taught it to me in a very conservative civil engineering department.

Your entire comment leads me to believe that you did not even read the article.

Do you have evidence to support this? The effect of widening highways used to support "induced demand" typically seems to hold up.

You may also want to read up on induced demand in more depth, because it's about where the demand is, not about how many people are driving cars.

I'd like to see a study across all road construction whether excessive induced demand is a guaranteed effect. In all articles related to it I only see examples cherry-picked where induced demand negated the effects of adding lanes. However, it's not clear whether this is universally true, true enough to be a rule of thumb, or something that only occurs in specific situations.

Anecdotally, removing toll plazas in favor of plate scanners and adding bridge lanes to the Goethals Bridge in Staten Island, as well as widening the Staten Island Expressway, has significantly reduced traffic delays and improved safety for several years. There _are_ more vehicles using the road but not enough more to significantly negate the improvements.

https://www.panynj.gov/port-authority/en/press-room/press-re...

To me, the case against improving roads is not nearly as open-and-shut as Strong Towns and similar one-sided policy-pushers make it out to be.

I don't know the data (or this area particularly well generally), but intuitively I would expect induced demand to be seen in the majority of cases of roads being updated for increased capacity.

Not because of a necessity that any capacity increase must always cause it, but because the decisions of which roads to expand are made by humans who, with the exception of edge cases (maybe the reason for rebuilding was because of the state of the road and increased capacity was a secondary affect rather than the main motivation; maybe corruption led to a poor choice being made; maybe forward looking expecting future demand to increase years later based on other factors like planned new buildings) wouldn't spend money increase capacity unless road traffic has been increasing and looks like it is't about to start trending in the opposite direction soon enough to make expanding the road a pointless short-sighted exercise.

Aka if increasing road capacity wouldn't induce more demand, how likely is it to be in a situation where it's considered worthwhile expanding it?

It would seems that removing the toll booths dramatically increases capacity and safety without the need to add additional lanes.

Also, more vehicles are using the road, true, but you have to remember that it increases over time.

How fast does demand grow? Does it ever stop or slow down? I know for a fact that far from every highway and road in America is congested: the majority are not. Even many if not most roads in metropolitan areas are not. So you can’t extrapolate demand increases ad infinitum for the current moment.

It's also about altering development patterns (suburbanizating) that spreads stuff out more and forcing everyone to drive more or even buy a car in the first place even if they'd prefer a walkable town.

Induced demand has never made sense. Does building a hospital create more sick people?

If you're only thinking about the already-there population, no. But many hospitals have patients that travel to get there.

Building a Mayo Clinic hospital in Arizona results in more sick people traveling there for care than otherwise would have.

Seems like there are some natural ways this could cause a loop. Now we hire more doctors and train more new specialists. That burnishes the reputation even more. Now more people want to come!

(One thing to note is that it's generally always a localized increase, not an absolute one. The people living in the new suburbs that you built a new shiny road to still would've lived SOMEWHERE in the country. You didn't increase that country's total demand for transit. You just redirected it into "road transit from this particular suburb" from wherever it would've gone in the absence of building that suburb.)

The part I always miss, is how is this a bad thing when it comes to roads? Building 'hospitals'[0] for people to use until there are enough hospitals[0] that people no longer have incentive to flock to the new Mayo Clinic in Arizona seems pretty desirable.

[0]Roads

The article asked the same question rhetorically. The last section was their answer.

Thats because roads are not "supply" it is "price". Building a hospital and staffing it should make the cost of healthcare in an area lower because there is more "supply" of healthcare. Which in turn means more people would go get healthcare that couldnt afford it before. The sick people are already there, they just cant afford to get treatment.

Just like the people who want to drive somewhere are already there, but the cost of a clogged commute makes it not worth it for them to spend the time, so they do something else.

No, but it can make them visible and willing to be treated. If people cant afford private health care or are unwilling to travel long distances or wait for a long time on a public health waiting list, they might just put up with being untreated. This is quite common for things like joint replacement or chronic dental issues.

It increases the number of people getting treated for their sickness, who otherwise would not be.

You are describing latent demand.