Hacker News Reader

Comment by digler999 | Hacker News Reader

There was a good New Yorker article within the past 2 years that talks about the increase in consumption. One example it used was refrigerators. Back in the 60s, something happened where they got cheaper/more power efficient. Instead of a gradual reduction in power, it increased because people just put the old one in the basement to freeze more stuff. Same is true (I think) with traffic. If we were to make, say, traffic signals more efficient by synchronizing them, or somehow increase capacity on freeways, people would just drive more. Not saying all resources are like this, however.

View on HN

This is called induced demand, whereby something that is plentiful will be utilized since it was built. Commonly, people say "Build it and they will come!" which is all the more true for roads.

Building a freeway or expanding it actually creates more traffic, since it lowers the cost of driving (time & gas) and when building a new freeway you actually create new traffic flows in unexpected ways. Building extra lanes actually causes more cars than those lanes could ever conceivably handle to pour onto the roadway due to the latent demand [1], and the most viable path forward is to create other avenues for said latent demand for transport to be serviced.

Ideally, you will build better mass transit infrastructure, with a sizable light rail network and feeder bus lines in the suburbs. Traffic will not get better due to this in any way, shape or form, but it will lessen the number of additional cars that end up traversing the streets every year, and prevent some additional load.

What this will do though is prevent your local economy from strangling itself in its own traffic, since your community is now providing more than one transport option. The only enemy of transit is density in the US, where we require massive swaths of land be set aside for unused parking that ultimately severely disadvantages every mode of transport, since it forces cities to be more spread out and dense construction to be very costly (underground parking construction costs are insane vs building above ground sqft).

[1] - https://www.wired.com/2014/06/wuwt-traffic-induced-demand/

> If we were to make, say, traffic signals more efficient by synchronizing them, or somehow increase capacity on freeways, people would just drive more.

Or say, invent self-driving cars...

That will just clog roads more, doesn't matter how much you mechanize it. Self driving cars will induce massive amounts of demand piled atop the latent demand that already exists for driving on nearly every arterial and freeway in cities, and even if everything was perfect, with cars driving at an optimal 45mph with minimal stopping distance, you can only move 2200 cars an hour per lane [1] and that scales non-linearly, as you add lanes to highways and arterial streets, throughput drops exponentially due to lane switching.

Now, self driving cars could get us closer to that 2200 cars an hour figure, but there is so much pent up demand for a magical faster option (latent demand) that when those marginal efficiency gains occur, there will be a flood of new vehicular traffic on the roads that will fully negate said efficiency gains, and likely exacerbate said issues since the tolerance for time spent in traffic will significantly increase for autonomous vehicle owners due to being able to do another task while your car drives for you.

[1] - http://www.fhwa.dot.gov/ohim/hpmsmanl/appn2.cfm

Indeed, that's the point I was hinting at: self-driving cars have the potential to increase carbon emissions, even the electric ones, due to increased usage and congestion.

Has Elon Musk publicly addressed this issue at all?

Self-driving cars are inevitable, so they may as well be electric. But if Tesla pulls the industry into the self-driving age before the majority of cars are electric they could actually be doing more harm than good. Perhaps self-driving cars should be required to be electric.

If we'll use them as "self-driving taxis" more than personal cars — if you're not attached to a specific "your" car — some nice things might start happening:

(1) multi-modal mass transit/car mixing becomes more attractive.

(2) transit can optimize for high-frequency mini/microbus formats rather than least-number-of-employed-drivers. this is less optimal from road-occupation POV but can be a win if it makes more people actually use transit?

(3) required parking space drastically decreases. Reaping the benefits will take a LONG time, but we'll be able to go back to denser cities?

I'm not sure if that's the point you are making, but I'm 100% certain that self-driving cars will increase the number of cars on the road, not decrease it.

Reply on news.ycombinator.com