Phantom traffic jams, and why lane expansion isn’t really successful(cnet.com)
cnet.com
Phantom traffic jams, and why lane expansion isn’t really successful
https://www.cnet.com/roadshow/news/traffic-talk-ford-vanderbilt-2018/
88 comments
Thank you. I am mystified by the apparently widespread mental framework that roads don't actually do anything other than attract masses of people in cars, inexplicably like insects toward bright light.
It's true that roads convey people & permit economic activity.
But that road down the way that is so congested right now? If we expanded it, and it could convey more cars at the same level of congestion- where did they come from?
The latent demand represents potential traffic that either:
1) is doing just fine taking a different route right now
2) isn't particularly economically valuable, i.e. not worth waiting through traffic for
3) things like future sprawl enabled by wider roads today
In order, #1 is not particularly valuable to reroute, it was doing ok. #2 is not very valuable either, though perhaps a little. #3 creates new demand, but the only thing we get in return is lower density.
So I guess in sum the point is that all the most economically valuable travel is already happening, right now, and by widening the road we add only the travel that has such marginal economic value it's not worth doing today. So the more we widen the road, the less & less economic value we get in return.
But that road down the way that is so congested right now? If we expanded it, and it could convey more cars at the same level of congestion- where did they come from?
The latent demand represents potential traffic that either:
1) is doing just fine taking a different route right now
2) isn't particularly economically valuable, i.e. not worth waiting through traffic for
3) things like future sprawl enabled by wider roads today
In order, #1 is not particularly valuable to reroute, it was doing ok. #2 is not very valuable either, though perhaps a little. #3 creates new demand, but the only thing we get in return is lower density.
So I guess in sum the point is that all the most economically valuable travel is already happening, right now, and by widening the road we add only the travel that has such marginal economic value it's not worth doing today. So the more we widen the road, the less & less economic value we get in return.
>2) isn't particularly economically valuable, i.e. not worth waiting through traffic for
With schedules and business operating hours, many economic activities are impossible due to hard time constraints.
I don't see how your argument would be less applicable to a city with two dirt roads.
With schedules and business operating hours, many economic activities are impossible due to hard time constraints.
I don't see how your argument would be less applicable to a city with two dirt roads.
I know you didn't mean it, but honestly your analogy is kind of brilliant. It does feel that way as somebody who enjoys walking.
Just aiming to maximize the throughput of private cars is not sustainable. The primary way to alleviate congestion problems should be making people less dependent on personal cars, both by urban planning that facilitates shorter commutes and by making mass transit more attractive and available.
That may be true, but not because of latent demand. Latent demand happens with mass transit and bike lanes too. It doesn’t mean we shouldn’t build those.
The point is to shift latent (and, indeed, actual) demand from personal vehicles to alternative means of transport. This is a net win when it comes to congestion because private cars have by far the greatest surface area per person transported (especially given that their passenger seats tend to be woefully underutilized).
There is a classic coordination problem regarding private/mass transport: as long as mass transit lags behind personal vehicles in convenience (and status), it is in the interests of car users that everybody except themselves switch to transit.
There is a classic coordination problem regarding private/mass transport: as long as mass transit lags behind personal vehicles in convenience (and status), it is in the interests of car users that everybody except themselves switch to transit.
I can’t speak to the aggregate, but going from two to three lanes in each direction on I-90 in northern IL has cut travel times for us by about 8% despite increased total traffic. This is because trucks passing now leave the left lane open and because the speed limit was increased 5MPH.
Yeah, I think that saying more lanes means more traffic is just ignoring the fact that the current road system never met the original demand. Also, we obviously have to plan for growth. We do have more people on the roads than we did in decades past. I haven't checked the numbers but I bet highway-lane-miles has not kept pace with population growth.
Well say you have a particularly busy 10 mile section of highway. You spend $millions adding lanes. Commuters are no happier because their commute time is unchanged. Riding on public transit drops which is worse for the environment.
Was adding the lanes a win for the city? The commuters? The environment?
It is counter intuitive, but if you think about it. Everyone has a commuting pain threshold that will cause a change in behavior. If a stretch of highway takes 10 minutes less because of additional lanes, then people will start commuting from 10 minutes further away.
Fortunately it's really easy to exactly quantify trip times, cars per lane, and similar. Quite a few studies have shown short term benefit, but within a year the traffic is within 0.1% of what it started with before adding the lanes.
Was adding the lanes a win for the city? The commuters? The environment?
It is counter intuitive, but if you think about it. Everyone has a commuting pain threshold that will cause a change in behavior. If a stretch of highway takes 10 minutes less because of additional lanes, then people will start commuting from 10 minutes further away.
Fortunately it's really easy to exactly quantify trip times, cars per lane, and similar. Quite a few studies have shown short term benefit, but within a year the traffic is within 0.1% of what it started with before adding the lanes.
> Was adding the lanes a win for the city? The commuters? The environment?
It was a win for the local economy.
By allowing more people to drive to and from locations they please, you are allowing more people to work jobs they otherwise wouldn't be able to, and purchase goods and services they wouldn't otherwise.
I'm rarely a fan of government spending; but, roads, police, and firefighting are probably some of the least evil.
It was a win for the local economy.
By allowing more people to drive to and from locations they please, you are allowing more people to work jobs they otherwise wouldn't be able to, and purchase goods and services they wouldn't otherwise.
I'm rarely a fan of government spending; but, roads, police, and firefighting are probably some of the least evil.
Giving people an all-you-can-eat buffet of asphalt encourages extremely inefficient resource allocation. Doing this without asking for anything to offset the negative externalities exacerbates the problem. So far the negatives of increased vmt have been completely ignored - they are not minor. Ask yourself - would you enjoy living on a busy street, or a quiet one? A wide one, or a narrow one? Which would you rather see your kid walk or cycle to school on? Which will have cleaner air? More noise? Would you rather your 85 year old parents or grandparents had to walk across an 8 lane road or a 2 lane one to get to the post office from their home?
You're comparing streets with roads, here. Fundamentally I agree with you, big roads divide communities and make cities unwalkable. But they are necessary.
A street is made for transporting humans first and foremost. It has sidewalks, driveways to houses are accessed directly from the street, and pedestrians can cross with relative ease, to the point that crosswalks are basically unnecessary.
A road is made for transporting cars first and foremost. There shouldn't be many residential driveways directly leading into a road, there may be sidewalks but they're not going to get used often, and you should rarely see pedestrians trying to cross. Crosswalks are very necessary on roads.
There's no getting around the fact that we need roads. Cars and trucks need ways to move long distanced quickly. The problem becomes when city designers try to mix streets and roads (to make a stroad), leading to massive arteries running through the heart of a residential neighborhood. Proper city design puts roads outside of pedestrian-heavy neighborhoods and fills them with easily and safely walkable streets instead.
There's nothing wrong with an 8 lane road, as long as that road isn't supposed to be used as a street as well.
https://www.strongtowns.org/journal/2018/3/1/whats-a-stroad-...
A street is made for transporting humans first and foremost. It has sidewalks, driveways to houses are accessed directly from the street, and pedestrians can cross with relative ease, to the point that crosswalks are basically unnecessary.
A road is made for transporting cars first and foremost. There shouldn't be many residential driveways directly leading into a road, there may be sidewalks but they're not going to get used often, and you should rarely see pedestrians trying to cross. Crosswalks are very necessary on roads.
There's no getting around the fact that we need roads. Cars and trucks need ways to move long distanced quickly. The problem becomes when city designers try to mix streets and roads (to make a stroad), leading to massive arteries running through the heart of a residential neighborhood. Proper city design puts roads outside of pedestrian-heavy neighborhoods and fills them with easily and safely walkable streets instead.
There's nothing wrong with an 8 lane road, as long as that road isn't supposed to be used as a street as well.
https://www.strongtowns.org/journal/2018/3/1/whats-a-stroad-...
we agree to an extent, but we've shoved roads and motorways where we should have streets. Santa Monica ridiculously slapped a 25mph sign and unguarded crosswalk on a Wilshire, a disaster of a road at the best of times, and of course people are killed on it now and then. How many nearby will die young from cancer,lung disease, etc - who knows?
Interestingly autism rates appear to be higher near freeways (like the ones we put in the middle of cities), yet the anti-freeways-in-urban-cores crowd somehow lacks mindshare compared to antivaxxers.
Interestingly autism rates appear to be higher near freeways (like the ones we put in the middle of cities), yet the anti-freeways-in-urban-cores crowd somehow lacks mindshare compared to antivaxxers.
If you haven't been familiar with the organization Strong Towns before, you should really look into it.
They've got tons of articles that are right up your alley on the dangers of just putting a 25mph speed limit on a road and calling it a "street".
This one is my favorite: Just Another Pedestrian Killed https://www.strongtowns.org/journal/2014/12/3/just-another-p...
They've got tons of articles that are right up your alley on the dangers of just putting a 25mph speed limit on a road and calling it a "street".
This one is my favorite: Just Another Pedestrian Killed https://www.strongtowns.org/journal/2014/12/3/just-another-p...
Thanks - I'm familiar with them. This is an issue I care deeply about, and one of the key reasons I emigrated.
This is 2018. A reasonable, sustainable policy would aim to decrease, not increase, the number of kilometers people drive around in private cars.
Not if electric cars take over and we have clean abundant wind, solar, and nuclear power.
Building electric cars is not exactly good for the environment either. They're also loud and kill people in accidents a lot more often than subways do.
And they do nothing for the congestion problem either. Private vehicles are simply a horrendously inefficient way to move people around, regardless of the source of motive power. Not to mention how inefficient (and people-unfriendly) use of land highways and parking lots are.
A lot of people here in Portland don't seem to grasp that. They go, well adding an extra lane will just make that lane full... yeah... exactly otherwise whats the point of building another lane? Mass transit is a great idea in theory, but when it doesn't run 24/7 to areas you want to go or has a problem with homeless and other crazy people harassing you [1, 2, 3, 4] it just doesn't work out. A good solution for a lot of places with tight space is to just stack the highways. Doesn't make sense endlessly making roads wider on both sides - it can only go so far.
1. https://www.kgw.com/article/news/crime/trimet-barber-gets-2-...
2. https://www.oregonlive.com/portland/index.ssf/2017/05/police...
3. http://www.wweek.com/news/2017/06/02/man-pummels-max-train-o...
4. https://www.oregonlive.com/portland/index.ssf/2018/05/crime_...
1. https://www.kgw.com/article/news/crime/trimet-barber-gets-2-...
2. https://www.oregonlive.com/portland/index.ssf/2017/05/police...
3. http://www.wweek.com/news/2017/06/02/man-pummels-max-train-o...
4. https://www.oregonlive.com/portland/index.ssf/2018/05/crime_...
Mass transit isn't perfect, so it shouldn't be had? That seems like flimsy logic when a number of the issues can be fixed.
Move to smaller over-the-road mass transit for off-hours or have fewer busses. You don't need to have full-size busses at 2am when you can use the smaller busses that schools use at times, for example, or simply have a bus that runs once an hour. You can make sure that major employers are serviced by busses either regularly or at times when their shifts start and stop (factories, for example). You can use van-sized taxis for smaller areas and adjust the routes to meet needs in smaller areas or have electric, automatic trams that run at specific times through the day and less so at night. Intercity travel can be more restricted if the demand isn't there.
Amsterdam has folks scan their ticket for the trams on their way out ... so they can see where folks are going to and from, in general, to better meet demand. Something like this can surely be done in different ways.
Even more, you can create safe walking and cycling paths so folks are more likely to be able safely get to public transport. Have bus shelters many places. Perhaps don't expect it to make a profit. Perhaps tax vehicles to help pay for them when they are available to the user - after all, if you own a car when public transport will adequately fit your needs, the car is a luxury. Perhaps utilize more tolls to help pay for them.
I've never had someone harass me on a bus where I am at now. In fact, I didn't have this issue in the states either when I was lucky enough to live in a place with busses. It would seem the solution would be to actually take care of folks that are homeless and mentally ill. It would make sense to call the police if someone is causing an actual disturbance to others. These aren't problems with public transportation, these are problems with society.
Move to smaller over-the-road mass transit for off-hours or have fewer busses. You don't need to have full-size busses at 2am when you can use the smaller busses that schools use at times, for example, or simply have a bus that runs once an hour. You can make sure that major employers are serviced by busses either regularly or at times when their shifts start and stop (factories, for example). You can use van-sized taxis for smaller areas and adjust the routes to meet needs in smaller areas or have electric, automatic trams that run at specific times through the day and less so at night. Intercity travel can be more restricted if the demand isn't there.
Amsterdam has folks scan their ticket for the trams on their way out ... so they can see where folks are going to and from, in general, to better meet demand. Something like this can surely be done in different ways.
Even more, you can create safe walking and cycling paths so folks are more likely to be able safely get to public transport. Have bus shelters many places. Perhaps don't expect it to make a profit. Perhaps tax vehicles to help pay for them when they are available to the user - after all, if you own a car when public transport will adequately fit your needs, the car is a luxury. Perhaps utilize more tolls to help pay for them.
I've never had someone harass me on a bus where I am at now. In fact, I didn't have this issue in the states either when I was lucky enough to live in a place with busses. It would seem the solution would be to actually take care of folks that are homeless and mentally ill. It would make sense to call the police if someone is causing an actual disturbance to others. These aren't problems with public transportation, these are problems with society.
I’m not saying it should exist at all. It needs to be better here. American mass transit is just plain awful.
WTF. Mass transit is a great idea period. It works in numerous places around the world. Even if it doesn't run 24/7 or everywhere it runs most of the day and to many places. Meaning less cars and less need to "stack the highways"(!!!), meaning less space used for parking and roads and more space available for humans.
It doesn't work for outlying areas. The further you get form a hub, the more varied the travel plans are. Express services work for connecting hubs. But it's the 30-60 mile out suburbs that are a challenge.
Note, they are only a challenge if bureaucrats and big business continue to pour money into centralizing resources in giant cities. In a modern Internet connected economy, there is very little reason to colocate so many government services and large corporations.
Note, they are only a challenge if bureaucrats and big business continue to pour money into centralizing resources in giant cities. In a modern Internet connected economy, there is very little reason to colocate so many government services and large corporations.
Also: induced demand is a good thing. Even if travel time were the same with three lanes instead of two, that's 50% more people moving down that road. That's a quantifiable social benefit.
If 100 people have decided they can now commute 30 miles instead of 20 miles because of increased road space, you haven't really enabled a whole lot. Granted, those people can live further away from their work, possibly on less expensive land.
You have, though, made it harder for the economics to work in favour of apartment buildings and public transit, and as a result made life more difficult for people who choose not to, are unable to, or can't afford to drive. Because people in the suburbs will vote in favour of thigns that make their motoring experience more pleasant, they will demand new construction have off-street parking (even of the residents don't drive), making units tens of thousands of dollars more expensive and less pleasant to walk near, and free or dirt cheap parking in city centres. They will force transit stations to have places for cars to sleep, not people (aka park and ride). Meanwhile, those people who don't drive still bear the negative externalities of the extra vmt and their taxes pay to enable it for others.
In addition to the economics, you've made the public realm more hostile to walking and cycling; car-friendly places are human-hostile.
More vehicle-miles traveled corresponds more emphysema, more extreme weather events, more noise, more light pollution, more cancer, and just more dead people in general. It's impossible to say who or when or how (except in the case of crashes) but every substantial road expansion kills some of the people nearby.
You have, though, made it harder for the economics to work in favour of apartment buildings and public transit, and as a result made life more difficult for people who choose not to, are unable to, or can't afford to drive. Because people in the suburbs will vote in favour of thigns that make their motoring experience more pleasant, they will demand new construction have off-street parking (even of the residents don't drive), making units tens of thousands of dollars more expensive and less pleasant to walk near, and free or dirt cheap parking in city centres. They will force transit stations to have places for cars to sleep, not people (aka park and ride). Meanwhile, those people who don't drive still bear the negative externalities of the extra vmt and their taxes pay to enable it for others.
In addition to the economics, you've made the public realm more hostile to walking and cycling; car-friendly places are human-hostile.
More vehicle-miles traveled corresponds more emphysema, more extreme weather events, more noise, more light pollution, more cancer, and just more dead people in general. It's impossible to say who or when or how (except in the case of crashes) but every substantial road expansion kills some of the people nearby.
Even if travel time were the same with three lanes instead of two, that's 50% more people moving down that road. That's a quantifiable social benefit.
Well, maybe those 50% achieve some benefit to themselves and society once they arrive at their destination. Keep in mind, if those people happen to be arriving at jobs they had previously, that's not a greater benefit than before.
However, the movement of people down the road itself is a social detriment. Most people don't to spend time in traffic for kicks and the cost of fuel consumption, pollution, wear on highways and wear on cars is significant. I mean, if somehow I could spend only five minutes in my car per day and still get to my destinations, the benefit would clearly be far greater.
Well, maybe those 50% achieve some benefit to themselves and society once they arrive at their destination. Keep in mind, if those people happen to be arriving at jobs they had previously, that's not a greater benefit than before.
However, the movement of people down the road itself is a social detriment. Most people don't to spend time in traffic for kicks and the cost of fuel consumption, pollution, wear on highways and wear on cars is significant. I mean, if somehow I could spend only five minutes in my car per day and still get to my destinations, the benefit would clearly be far greater.
> That's a quantifiable social benefit.
Only if you know how the occupants in these additional vehicles were transiting previously. If they were previously carpooling, or if they moved away from a good (but slightly less time efficient) public transport option, then the social benefit would be somewhere between minimal and nil.
Only if you know how the occupants in these additional vehicles were transiting previously. If they were previously carpooling, or if they moved away from a good (but slightly less time efficient) public transport option, then the social benefit would be somewhere between minimal and nil.
But presumably the new occupants compared their existing old travel method, and deemed the new one better. If the old method had some "social benefits", it must mean that the old method's social benefit came at a cost in benefit to the occupants. Otherwise, they would not have chosen the new method!
> the new occupants compared their existing old travel method, and deemed the new one better
You’re leaving out the taxpayers subsidising their commute.
You’re leaving out the taxpayers subsidising their commute.
Just because an individual makes a particular choice when given the opportunity, doesn't mean it must be a social benefit.
There definitely seems to be a couple of functionally optimal road sizes, where extra pavement dramatically improves flow and by doing so at least appears to greatly improve throughput.
In our town, the city has been switching from tight two lanes each way to one lane each way, plus turning lanes and bike lanes, and these new layouts are significantly more pleasant to drive with none of the old left-turn snarls of the past.
In our town, the city has been switching from tight two lanes each way to one lane each way, plus turning lanes and bike lanes, and these new layouts are significantly more pleasant to drive with none of the old left-turn snarls of the past.
How long ago were the lanes expanded? According to the general concept of latent demand, I would expect it to take perhaps a few years for the latent demand to manifest.
About two years. I agree it should continue to increase over the next several years as homes and businesses are located with the wider highway in mind. However, if latent demand takes too long to manifest, the lane expansion has saved driving time for a significant fraction of its life and shouldn’t be thought of as ineffective.
Adding a lane to an urban highway is immense. There is not only the cost of building the actual road, but the cost of acquiring the land which often requires displacing people from their homes via eminent domain and the cost of the disruption caused by the construction.
Just providing some benefit is not enough to justify the cost; it must provide an enormous benefit.
Just providing some benefit is not enough to justify the cost; it must provide an enormous benefit.
I just got a new Subaru Outback with their Eyesight Adaptive Cruise Control. I've gotta say it is fantastic and just the sort of fairly simple and safe technology that could ease traffic jams. Certainly outside the city where I live, the way it nicely follows cars at a safe distance seems to make a lot of sense. As soon as the lead car moves away, it neatly accelerates to the preset speed and smoothly decelerates on the engine braking (or even real braking if needed) when it sees the gap closing. It's very obvious on my long commute that probably only 30% of cars seem to be using just regular cruise control, based on how their speed varies on the clear open road.
Adaptive Cruise is life-changing. I couldn't go back to normal cruise.
A major downside to Adaptive Cruise is when another driver uses your gap to change into your lane, and your car slams on its breaks because suddenly their car is just inches away from yours.
Recklessly aggressive drivers are learning they can take abuse these safety features in the cars of other drivers, making roadways potentially more unsafe.
A major downside to Adaptive Cruise is when another driver uses your gap to change into your lane, and your car slams on its breaks because suddenly their car is just inches away from yours.
Recklessly aggressive drivers are learning they can take abuse these safety features in the cars of other drivers, making roadways potentially more unsafe.
Different cars have different behaviors here. My father has a Toyota Camry, and his will slam on the brakes, while mine, a Mazda 3, will much more gradually slow down and slowly increase the space until the gap is restored. It also seems to have a much more fluid sense for how big that gap should be.
I love my Mazda ACC and it will be a major criteria point in future car buying decisions.
I love my Mazda ACC and it will be a major criteria point in future car buying decisions.
What are you saying, you want it to crash into them for you? That's a major downside?
I'm not parent, but just a note from me: I drove a brand new Volvo XC40 last week that had adaptive cruise control. If someone just cut in to the front of me the car slammed on its brakes even if it wasn't necessary. The car could just as easily slow down gradually until necessary space was free in front of the car (as I usually do when I don't have adaptive cruise control).
Try Mazda, they're much better in that regard.
I rented my first Prius recently, and it had this. I absolutely love it. Adaptive cruise is one of those great features that both requires less attention, and teaches better habits.
Cruise control would increase road throughput but, like adding extra roads, it would probably be offset by extra cars in the road anyway
Bill Beatty's comprehensive explanation from 20 years ago is still the best: http://trafficwaves.org/
All he's doing here is obviating braking. In theory it could make for slightly reduced trip time as some stop-and-go's have more inefficiency than just moving slowly. He's reducing the pressure of the flow. But he's doing nothing about the actual causes of the slowdowns, which is on-ramps, off-ramps and rubbernecking. So while you reduce stop-and-go, you're still traveling at maximum capacity, just like in stop-and-go's. A kink in a hose doesn't go away if you reduce the water pressure.
Lane expansion isn't likely to eliminate rush hour, but it will increase the number of vehicles served, and can also reduce the duration of peak congestion. As vehicle populaton grows, congestion will grow as well, however.
A recession is even better than construction at improving traffic, however.
A recession is even better than construction at improving traffic, however.
The article has a lame explanation of latent demand. Sure some switch from surface streets to the highway. But people also decide to make more trips, commute further, or use public transportation less. So the result is after an adjustment period of about one year there's almost precisely as much traffic per lane as before.
Your point about recession is interesting. As we came out of the previous recession after the crash in 07/08, I noticed commute times increasing in the DC metro area.
My argument is and has been that highway lane miles have never met demand. More lanes bring more cars simply because the other routes we're equally congested/slow before and now there's a slightly better option with the new mega highway.
My argument is and has been that highway lane miles have never met demand. More lanes bring more cars simply because the other routes we're equally congested/slow before and now there's a slightly better option with the new mega highway.
Here in Bangkok the size of the (already legendary) traffic jams is certainly a great proxy for the economy as a whole.
If a lot more people drove manual transmissions, or at least learned that you can slow down by just raising the accelerator a little rather than hitting the brakes, things would be a lot better. It really doesn't seem like a hard thing to learn.
I believe it's the reaction time lag, not using the brakes instead of the throttle to slow down. Basically people suck at staying in the ideal position half way between the car in front in the car in back. Because of that they brake hard as they come close to the car in front. The next car has to brake harder... till you end up with cars stopping. So it's the human reaction time combined with cars accelerate (generally) slower than they brake.
This. There are lots of studies that have pointed to slow re-starters as the prime culprit for traffic. And it's only gotten worse with cellphones because now there is something to longer distract people when the come to a stop.
Slow-restarters and sudden late-brakers. This is really all about an oscillating wave being carried vs mean flow speed. The best part when you get the hang of it is that in many situations when you can anticipate the "slow-restart/sudden late-brake" of a dozen or so cars before you (and a couple behind you), you can act as a dampener as you strive to keep near the mean flow speed, and delightfully see the phantom jam downright evaporate in your rear view mirror.
I think it's that most drivers just don't pay attention. I'm able to maintain consistent following distances as speeds change if I pay attention to doing that.
The other problem is that if you maintain a gap as traffic speeds are slowing, someone will cut in front of you to gain a car length. Then you have to slow down using the brakes.
The other problem is that if you maintain a gap as traffic speeds are slowing, someone will cut in front of you to gain a car length. Then you have to slow down using the brakes.
Human reaction time is always there, but driving a stickshift generally teaches you to be more observant & plan ahead more- to anticipate and try to avoid extreme braking or acceleration situations.
I drove stick for years, and your parent is right on. And, back in an automatic, I've naturally regressed, I freely admit.
I drove stick for years, and your parent is right on. And, back in an automatic, I've naturally regressed, I freely admit.
My favorite related quote is that adding lanes to decrease traffic is like letting your belt out to lose weight.
For more info see: https://en.wikipedia.org/wiki/Induced_demand
For more info see: https://en.wikipedia.org/wiki/Induced_demand
Right. So when the subway gets crowded, the ideal thing to do is reduce the number of trains per hour so that people stop trying to ride it.
That's like saying, "we shouldn't build more hospitals, because we don't want to make more sick patients". It's only "true" if you grossly botch your before and after counts.
But you are ignoring the human and economic benefits of increased travel.
A better example would be someone pulling their belt so tight they can barely move, and then letting it out.
A better example would be someone pulling their belt so tight they can barely move, and then letting it out.
But we let the belt out all that way, and the belt still touches their waist!
Then there is more demand. Keep going. There is nothing that has unlimited demand.
You can't reduce demand by constraining supply, all you are doing is making people unhappy.
If your policies are very free, you'll find people will figure out how make things more efficient. For example inexpensive delivery will become possible if there was enough space for traffic.
I know this because I've lived in dense cities (no delivery service - it's simply too hard to find parking, and thus too expensive) and wide urban areas (every store offers delivery because it's cheap and easy).
And that's exactly the opposite of what everyone thinks will happen - they want cars off the road, and replaced with delivery, without actually making delivery practical (i.e. inexpensive).
You can't reduce demand by constraining supply, all you are doing is making people unhappy.
If your policies are very free, you'll find people will figure out how make things more efficient. For example inexpensive delivery will become possible if there was enough space for traffic.
I know this because I've lived in dense cities (no delivery service - it's simply too hard to find parking, and thus too expensive) and wide urban areas (every store offers delivery because it's cheap and easy).
And that's exactly the opposite of what everyone thinks will happen - they want cars off the road, and replaced with delivery, without actually making delivery practical (i.e. inexpensive).
Limiting the use and possession of private cars would go a long way into reducing environmental impact and income inequality.
That'd be great, but no one wants to actually build any workable alternatives, at least in the US. A lot of people would probably be happy to give up their cars (or at least keep them parked until they need them for weekend trips) if they had good public transit available to use instead. But US cities are not doing a very good job with that.
I'm curious what the specific mechanisms are that you imagine would cause what you claim.
First, how exactly are you going to limit the use and possession of private cars?
First, how exactly are you going to limit the use and possession of private cars?
I wonder if toll payment based on amount of time spent in a lane would improve things? Right/exit lane would be free and incrementally increase up to the high cost left lane. The lane stripes would have connected rfid sensors to measure amount of timr spent in each lane.
Might work because less lane switching means less slow downs to adjust to new cars entering a lane. And people who don't absolutely need to will remain on right lanes while leaving the left lane for those committed to the lane or for those using it momentarily to pass other cars. Heck,you can even change lane pricing dynamically based on traffic and time of day.
Might work because less lane switching means less slow downs to adjust to new cars entering a lane. And people who don't absolutely need to will remain on right lanes while leaving the left lane for those committed to the lane or for those using it momentarily to pass other cars. Heck,you can even change lane pricing dynamically based on traffic and time of day.
This would almost certainly cause extreme congestion on the exit lanes and as a result a huge speed differential between lanes.
Moving to the faster lanes would be dangerous. Cars merging back into the slower lanes to exit would cause permanent jams in the second lane at every turn-off. Nobody would benefit in the end and the road would be horribly unsafe.
Moving to the faster lanes would be dangerous. Cars merging back into the slower lanes to exit would cause permanent jams in the second lane at every turn-off. Nobody would benefit in the end and the road would be horribly unsafe.
That's why I suggested dynamic pricing and charging based on time spent. The right most lanes would congest but they would also have exiting cars.
Second from right could cost a bit more than free (1cent per mile for example). It would be an incentive to use it for passing but also to avoid right lane congestion for a negligible fee. Third and fourth from right can cost $0.20 and $0.40 per mile which discourages long term use but also allows those in a hurry who can afford the fees to leave the highway faster and allow for cars in right lanes to use less congested left most lanes to pass slow drivers and lane exit/entrance in their lanes.
Think of it from a network traffic CoS and QoS point of view (also,sorry for the late response,hope you see this)
Second from right could cost a bit more than free (1cent per mile for example). It would be an incentive to use it for passing but also to avoid right lane congestion for a negligible fee. Third and fourth from right can cost $0.20 and $0.40 per mile which discourages long term use but also allows those in a hurry who can afford the fees to leave the highway faster and allow for cars in right lanes to use less congested left most lanes to pass slow drivers and lane exit/entrance in their lanes.
Think of it from a network traffic CoS and QoS point of view (also,sorry for the late response,hope you see this)
Again, a nice idea in theory but it will just concentrate congestion towards the right, which would exacerbate the speed differential between lanes and make the road much less safe.
It would exacerbate economic disparities, forcing poorer people—who are already chronically time-poor as well as money-poor—to spend more time in their car for the convenience of rich people. Rich people already have innumerable advantages in our society, we don't need to go around looking for more to give them.
(As a comparatively rich person, I appreciate living in a society surrounded by people that are—in the main—well educated, healthy and not overly stressed, and I happily pay a lot in taxes to help maintain that. Which is why the USA holds zero appeal to me; you couldn't pay me to live there.)
It would exacerbate economic disparities, forcing poorer people—who are already chronically time-poor as well as money-poor—to spend more time in their car for the convenience of rich people. Rich people already have innumerable advantages in our society, we don't need to go around looking for more to give them.
(As a comparatively rich person, I appreciate living in a society surrounded by people that are—in the main—well educated, healthy and not overly stressed, and I happily pay a lot in taxes to help maintain that. Which is why the USA holds zero appeal to me; you couldn't pay me to live there.)
Overall,there would be less lane changing and cars would leave the freeway faster. The pricing can be adjustes or subsidized for low income drivers.
If you pay a toll anyways,this is an opportunity to save money and use the tollway,where as you would normally take the slow and long unpaid route. This would actually make toll roads accessible to the poor. No matter how I look at it,if you add lanes,they should be paid for so that it would be more affordable abd practical and actually incentivize drivers to cause less jams.
Also,traffic jams are not exclusive to the USA. Maybe it should be modeled in a more politically convenient way? How about fuel tax deductions? You get n deductions where you lose more deduction points by using left lanes? That way there is no added cost to the driver,the tax payers as a whole pay for it.
If you pay a toll anyways,this is an opportunity to save money and use the tollway,where as you would normally take the slow and long unpaid route. This would actually make toll roads accessible to the poor. No matter how I look at it,if you add lanes,they should be paid for so that it would be more affordable abd practical and actually incentivize drivers to cause less jams.
Also,traffic jams are not exclusive to the USA. Maybe it should be modeled in a more politically convenient way? How about fuel tax deductions? You get n deductions where you lose more deduction points by using left lanes? That way there is no added cost to the driver,the tax payers as a whole pay for it.
I find that on my commute most of the time when one lane is moving slow it’s because the vehicle at the front of a long line of cars, usually the left two lanes, in traveling 5-10 mph slower than other traffic. Speed differential and the use of brakes seems to cause traffic jams, so having vehicles all traveling the same speed or at least moving out of the way if they’d prefer not to go with the flow would seem to solve these issues.
"A long line of cars ... traveling 5-10 mph slower"
In other words you're saying that they're closely packed. If the distance between cars is reduced by 50% and their speed has reduced 10-15%, the throughput of that section of road has increased substantially.
As you said, it's speed differential and the use of brakes that causes traffic jams. By that reasoning, the left lane should go the same speed as the right. That would also discourage the use of lane changes which is generally what causes people to use their brakes on the freeway...
And remember the square law of braking. Because the safe following distance increases so much when you increase speed, faster speeds on the freeway decrease throughput rather than increase it.
In other words you're saying that they're closely packed. If the distance between cars is reduced by 50% and their speed has reduced 10-15%, the throughput of that section of road has increased substantially.
As you said, it's speed differential and the use of brakes that causes traffic jams. By that reasoning, the left lane should go the same speed as the right. That would also discourage the use of lane changes which is generally what causes people to use their brakes on the freeway...
And remember the square law of braking. Because the safe following distance increases so much when you increase speed, faster speeds on the freeway decrease throughput rather than increase it.
No. I’m saying that when faster traffic comes across these drivers they either change lanes immediately, if possible, which causes problems or they stay behind that driver, but still with many cars hunting for a faster moving lane behind them. I don’t find the slow moving line to be any closer together at all mostly because of people constantly changing through that lane.
If the far right or middle lanes are traveling faster than the left and also moving smoothly then someone’s doing it wrong.
If the far right or middle lanes are traveling faster than the left and also moving smoothly then someone’s doing it wrong.
> And remember the square law of braking. Because the safe following distance increases so much when you increase speed, faster speeds on the freeway decrease throughput rather than increase it.
Got an article to explain this? It seems counter-intuitive.
Following distances should be based on time, not physical distance. It shouldn't matter if cars are going 30 mph or 300 mph. If everyone maintains a 3 second following distance, there should still be a throughput of 20 cars per minute.
Got an article to explain this? It seems counter-intuitive.
Following distances should be based on time, not physical distance. It shouldn't matter if cars are going 30 mph or 300 mph. If everyone maintains a 3 second following distance, there should still be a throughput of 20 cars per minute.
At 10mph, safe following distance is 1.8 seconds.
At 60mph, safe following distance is 3.6 seconds.
At 90mph, safe following distance is 4.4 seconds.
http://www.adventures-in-education.com/math/StoppingDistance...
https://en.wikipedia.org/wiki/Braking_distance
http://www.adventures-in-education.com/math/StoppingDistance...
https://en.wikipedia.org/wiki/Braking_distance
These problems only occur when the load get close to the effective maximum load. It's easy to look for blame in driver behavior, but I think misleading -- it's like taking that first crystal to form in 0 degree water, and blaming it for ice formation. You may be able to reduce crystallization (and thus sneak in a couple more Joules) by distilling the water (adaptive cruise control) or cleaning your glassware (on-ramp metering or "do not change lanes" signs), but really you're just fighting the statistics of phase changes.
Note that I don't have a ton of supporting evidence for this sort of thing. One imaginary project that I would love to see if I had the time or the patience for deep research is to just gather a data set and make it public of anonymized traffic behavior -- over a large stretch of highway, have periodic cameras that track every car by license plate (if possible) or car shape, aggregate and remove the identifying data, and output a data set that allows you to slice it by time or by moving segment to directly observe these things.
Note that I don't have a ton of supporting evidence for this sort of thing. One imaginary project that I would love to see if I had the time or the patience for deep research is to just gather a data set and make it public of anonymized traffic behavior -- over a large stretch of highway, have periodic cameras that track every car by license plate (if possible) or car shape, aggregate and remove the identifying data, and output a data set that allows you to slice it by time or by moving segment to directly observe these things.
I was hoping for more detail. The argument against adding more lanes is that making the freeway faster than surface streets causes people to prefer the freeway to surface streets.
Tying lane count back to fuel economy, average door to door time, and number of vehicles served would be much more productive.
Similarly, the article recommends hov meters on freeway entrance ramps, but doesn’t mention that they create jams sbout a half mile after the entrance. Again, some sort of metric involving “how many people”, “how long”, and “how much gas” would go a long way.
Fun fact about the Bay Area: my car gets about 75% of epa freeway mileage driving from SJ to Santa Cruz (over a mountain), but only 50% from San Jose to Mountain View (in traffic). This is if I try my best to conserve fuel in the traffic case. If I drive aggressively, the gap widens.
So, reducing conjestion could substantially increase the miles per gallon of the existing silicon valley fleet of cars.
Instead, the “environmentally friendly” nimbys keep voting in people that pay for traffic quiescence “upgrades” that further snarl traffic and increase the local carbon footprint.
Tying lane count back to fuel economy, average door to door time, and number of vehicles served would be much more productive.
Similarly, the article recommends hov meters on freeway entrance ramps, but doesn’t mention that they create jams sbout a half mile after the entrance. Again, some sort of metric involving “how many people”, “how long”, and “how much gas” would go a long way.
Fun fact about the Bay Area: my car gets about 75% of epa freeway mileage driving from SJ to Santa Cruz (over a mountain), but only 50% from San Jose to Mountain View (in traffic). This is if I try my best to conserve fuel in the traffic case. If I drive aggressively, the gap widens.
So, reducing conjestion could substantially increase the miles per gallon of the existing silicon valley fleet of cars.
Instead, the “environmentally friendly” nimbys keep voting in people that pay for traffic quiescence “upgrades” that further snarl traffic and increase the local carbon footprint.
Here in New York I would attribute traffic jams in equal parts to the endless construction, massive congestion, and a huge number of terrible and/or elderly drivers. All it takes is a couple of slow/bad drivers to clog up the expressway for miles, even if the congestion is low and there is no construction.
If another freeway lane attracts drivers up from the surface streets, then that must be good for the surface streets. I assume the new drivers must have determined that that's a more efficient way to get from A to B.
Better idea- expand the lanes but give them to a light rail system instead.
One solution is to restrict lane changes.
All it takes to create a phantom traffic jam is someone to change lanes where the cars behind need to slow or brake to accommodate them.
All it takes to create a phantom traffic jam is someone to change lanes where the cars behind need to slow or brake to accommodate them.
...if the only benefit of building more lanes was reducing mean transit time at rush hour. You could also just build a one lane road and have the toll so high that only a single driver could afford to use it; that would be optimally “beneficial” in this model.