14 ms·
Jevons paradox
- jihadjihad 3y agoA similar concept is that of induced demand [0]. When cities expand the number of highway lanes, for example, sometimes they find that traffic and congestion actually increase. 0: https://en.wikipedia.org/wiki/Induced_demand https://en.wikipedia.org/wiki/Induced_demand
- CalRobert 3y agoIndeed, it's used as a great example of Jevon's Paradox at https://youtu.be/pCzCJzwrB_c?si=N8PgXPv3fCEDvdfN&t=103 https://youtu.be/pCzCJzwrB_c?si=N8PgXPv3fCEDvdfN&t=103
- autoexec 3y agoThe way induced demand is framed sometimes as an argument for not expanding busy roads really bugs me. It's not that adding a new lane to a busy road suddenly creates the demand to drive on it. The "demand" already exists and simply isn't being met by the inadequate infrastructure. If there's an aquarium at the end of a busy road, and I'm extremely offended by the sight of fish, no amount of added lanes is going to "induce" me to want to get up on a Saturday and drive to the aquarium. However, if there's a opera house at the end of a busy road and I love the opera and want to go every weekend, inadequate infrastructure can make the experience driving there painful enough that I'll just stay home, but my desire to go there hasn't changed. I'd still love to see the opera and adding enough new lanes will mean I'm able to go. If congestion increases immediately after new lanes for traffic are added it's probably because the existing demand to travel down that particular road has been being ignored for way too long. As long as you're a fan of allowing people to go to the places they want to visit, then adding new lanes of traffic is a really good idea even when it results in more cars on the road. The goal of the larger road should be to allow more people to get to where they want to be, not to reduce the number of cars using it.
- MalcolmDwyer 3y agoNo, 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.
- autoexec 3y agoBusy 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.
- firedaemon 3y agoMostly 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.
- autoexec 3y agoI 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.
- pixl97 3y agoI see this happen with cloud services all the time... Customer has on prem services that they do X operations with, we'll say 10,000 daily as an example. They look at the metrics and moving to cloud for that many operations would be cheaper. But very soon after moving to the cloud and an expandable infrastructure, operational costs explode. Suddenly they are not doing this 10k times a day, they are doing it 20 or 30k times a day. Whereas before before kicking off whatever job they were doing they were careful to ensure it was important and cleaning up mistakes in scripts doing while 1=1 loops because it had a measurable impact, after the backpressure disappeared, they'd do this operation without thinking about the costs.
- tomrod 3y agoDuring my early graduate school days, I recall being fascinated by these paradoxes caused by wealth and substitution effects. It's a big deal! We become more efficient with something, and the reduction in demand results in more utilization due to low costs in some cases. It's a major consideration for environmental and humanitarian policy.
- schiffern 3y ago> I recall being fascinated by these paradoxes caused by wealth and substitution effects For your intellectual safety, just be sure to exclude "population growth" from your list of effects. People really really seem to hate it when anybody points that one out. https://en.wikipedia.org/wiki/Thomas_Robert_Malthus https://en.wikipedia.org/wiki/Thomas_Robert_Malthus If history is any guide, they'll strawman your "-ism" until even the most general observations about feedback loops and population become twisted into a freakishly specific (and easily dismissed) reinterpretation. Even today, most people think that historical increases in food production since the 1800s "disprove" the Malthusian thesis.
- toolz 3y ago> Malthus considered population growth as inevitable whenever conditions improved Interesting how the most developed nations have the lowest birthrates. I don't think his findings in the late 1700s have withstood the test of time. edit: the trend also applies for population growth, not just birthrates
- Paul-Craft 3y agoMalthus could not have foreseen the development of the Haber-Bosch process or the green revolution.
- schiffern 3y agoHe literally did. Read him. Malthus foresaw and mentioned the possibility of improved farming techniques. After all, it's not like improvements in farming hadn't happened before! In reply to this predictable objection, Malthus..... well, I suppose you'll just have to read it to find out. :) https://archive.org/details/essayonprincipl00malt https://archive.org/details/essayonprincipl00malt
- intalentive 3y agoAndy giveth, Bill taketh away.
- RachelF 3y agoYep, I was about to say that. Much modern software is written in incredibly inefficient non-compiled languages, but it is easy for developers. The same goes with app size, cos hard drive space and bandwidth are almost free.
- martopix 3y agoWhen I first learned about Jevons Paradox, I thought of what a colleague once told me: "I believe that low-power neuromorphic chips is one way computational neuroscience can help use less fossil fuels by reducing the power consumption of AI". Because of the paradox, I now believe that the effect would rather be to make small AI devices more widespread instead, with no impact on total power consumption.
- pixl97 3y agoAt the end of the day Earth has a thermodynamic budget in order not to experience run away heating. Moving to things like fusion may delay the global warming caused by carbon, but our hunger for more energy will eventually lead to energy use beyond what we can radiate away. Hopefully, 1, we get to the point we have to worry about that. And 2, we'll have cheap and easy access to space where we can move entropy intensive processes away from the planet.
- CalRobert 3y agoAs much as I was glad to see lighting become more efficient, as someone who enjoys seeing the night sky I was worried what the impact of virtually free lighting would be. And indeed, at least in rural Ireland the proliferation of extremely bright uplights which are on all night has not been great for seeing the milky way. Similarly, I worry that if we ever _do_ get fully self-driving EV's it will only serve to make it comfortable to commute from, say, Yosemite to Mountain View - and who wouldn't want that? Yet I rarely see increased congestion come up in discussions about FSD.
- borissk 3y agoInteresting question. Beijing has huge problems with traffic jams and the CCP has has taken some extreme measures to limit the number of cars there. E.g. they practically don't issue registration numbers for ICE vehicles and there's a queue for EV registration numbers, that currently goes to 2035 (if you join the queue now you'll get a reg number in 2035). Some cities with heavy traffic in Europe and China have implemented rules, like only cars with even registration numbers are allowed to drive on Mondays.
- Sharlin 3y agoLighting is simply something that must be regulated, just like everything else that has negative externalities. It’s not that different from noise or other kinds of pollution.
- 3y ago
- nologic01 3y agoAn important factor seems to be whether the underlying good or service is of the satiating variety [1] and how quickly does such satiation kick-in. E.g. producing cheap and plentiful food will increase demand up to a point, but once everybody becomes obese that process stops (and mutates into waste). If you can generate cheap and plentiful jetpacks, pretty soon the sky will be dark with swarms of human mosquitos criss-crossing the biosphere. But at some point people will have exhausted the time they want to dedicate to this activity. On the other hand a generic useful good such as low entropy energy that can be repurposed for a wide variety of uses might actually never saturate. That is probably what happened with fossil fuels. [1] https://en.wikipedia.org/wiki/Economic_satiation https://en.wikipedia.org/wiki/Economic_satiation
- skrebbel 3y agoThis is currently playing out with solar panels. They get cheaper and you need fewer for the same production, which increases demand and that continues the cycle. Just pointing this out cause the example linked is coal (bad) but it also works for things that are generally thought of as good.
- rozab 3y agoI guess the implicit point the poster was getting at (and there always seems to be one when people post wikipedia articles on HN) is that the massive increases in compute efficiency (ops per watt) have lead to massively increased power consumption, often in pointless applications like crypto mining
- skrebbel 3y agoYeah but if that fuels increasing innovation and scale in solar panel manufacturing then it’s still bottom line a good effect.
- nickstinemates 3y agoI was introduced to this concept when working with Intel on a business deal. This seemed to be a guiding principle in our discussions as well as an ethos a lot of the Intel Capital had strong belief in.
- jevons_lover 3y agoHa. I’m married to a Jevons so scored some solid boyfriend points with this letter to The Economist (March 2017)… https://ibb.co/b5VLByn https://ibb.co/b5VLByn
- hathawsh 3y agoI enjoy how the game Factorio makes paradoxes like this clear and obvious. In Factorio, when you replace your coal-driven power infrastructure with a big array of solar panels and batteries, pollution (which attracts bugs in the game) falls temporarily, but then other polluters more than replace the original pollution. The only solution in the game is to keep going and figure out how to do better.
- rodrigosetti 3y agoIntelligence is a resource of this type
- valyagolev 3y agoParkinson's Law is probably the same thing, but framed in much more hostile way (now it's the worker being lazy/creating extra work/yak shaving...)