Building Bigger Roads Makes Traffic Worse | Hacker News Reader

Building Bigger Roads Makes Traffic Worse

wired.com

212 pointsdata_scientist230 comments

In addition to the correlation found by these researchers, I suspect the Braess Paradox[1] and other complex network phenomena are also at work.

[1] Adding more roads to a network of roads can make traffic worse (and viceversa, closing roads can make traffic better). See http://en.wikipedia.org/wiki/Braess's_paradox or watch this friendly explanation: http://www.youtube.com/watch?v=ZiauQXIKs3U

Edits: moved links and parenthesis to footnote, and added "In addition to the correlation found by these researchers" and "other complex network phenomena" to make my point clearer.

This is not the same statement. If I have one path which has a certain amount of bandwidth and subsequently double it (making the road twice as large), the overall speed of the system must either increase or stay the same (the latter if the system is under capacity).

That's the paradox! The difference between a traffic route and a bandwidth-pure pipe (apologies for idiosyncratic terminology, i'm not a trained traffic planner) is that cars will choose their routes selfishly. The way I think of this difference is that even on a straight road that doubles its number of lanes, and we've all seen this, cars will change lanes with the idea that they'll get through traffic. Unfortunately, this lane changing, especially in gridlock situations, slows other cars down more than their overall speed is increased. Thus the overall bandwidth of the road, such as it is, is reduced. Add brake ripples and slow reaction time when moving in bumper-to-bumper, and the reduced overall speed should be obvious.

The real problem is that the roads don't seem to be engineered with a "bandwidth" mentality. Instead I see interchanges where 2-3 lanes from a new road are being piped into an existing 2-3 lane road. Did the existing road get expanded to 4-6 lanes? Very rarely, because you would see a lot more 10-20 lane roads running through downtown.

Instead, it seems a lot of "traffic" studies are just minor tweaks around the edge, add a lane or remove a lane from a 4-6 lane main corridor and the traffic patterns basically don't change that much because its massively oversubscribed already.

actually they manipulate the flow of cars to the smaller roads by making u wait on the offramp longer when its busy iirc, after watching a few docu's on the engineering i think theyre engineers are really good and bright

the book "selfish routing and the price of anarchy" by tim roughgarden is about this. its really interesting for its own sake, though i dont know how relevant it is to this particular thread's article

The video doesn't do a good job of explaining this at all.

Ultimately the reason why adding more roads to a system produces more traffic is more people take cars vs. public transit.

No, its much more complex than that.

Adding road capacity makes other types of trips possible. It's not a simple "hey, I'm taking the bus today because they built a lane". It's more like "Hey, I don't need to worry about uprooting my family or spending 6 hours on the bus to get a job in place X vs Y."

Think about it in computer networking terms. Netflix and Office 365 didn't exist in 1995 because it wasn't possible to make those sorts of solutions work then. Today, we need to worry about network congestion because millions of people are streaming movies.

Very much so. New and bigger roads cause all kinds of associated growth. New businesses and home pop up because it's now practical to get to them in those locations. This increases demand for the roads, so they eventually become congested again. But not just because people like to go from place to place for no reason other than to go.

It would seem then one way to spur economic growth would be to build more roads, transit, etc. Just as building faster internet spurs more internet companies.

Been there, done that... it's called 1946-2006.

It's probably the same reason why we just get bigger hard-drives instead of cleaning stuff we store in our current storage capacity.

> Ultimately the reason why adding more roads to a system produces more traffic is more people take cars vs. public transit.

More precisely, more people take cars vs. not take cars. "Not take cars" can be public transit, human powered transport or even not moving at all.

This correction might seem pedantic, but I think it is important. One of the lessons of these "paradoxes" (more roads = more congestion) is that one must take into account the whole traffic system.

I'd add that another salient factor is people's housing choices relative to jobs and other destinations -- there is an interplay between road-building and housing construction.

No the video is about the The Braess Paradox.

You're talking about the Downs–Thomson paradox. that's a related, but different, problem.

[1] https://en.wikipedia.org/wiki/Braess%27s_paradox [2] https://en.wikipedia.org/wiki/Downs%E2%80%93Thomson_paradox

That's in addition to leading people to move further out -- which also intensifies the traffic at the re-entry points to the thinner roads at the center, which are march harder to increase in capacity.

bigdubs: that video explains the Braess paradox, not the correlation found by the researchers. I modified my comment to make that clearer. Thanks!

Please explain how he can include a route to switch between routes that takes zero time. That does not make logical sense at all. It kind of ruined any premise he might have started with.

A common example of this is an intersection. But even without that example, I don't know how that could ruin the explanation. Zero switching time is just used to simplify the analysis, the overall phenomenon should be clear.

I think that was just to keep the math easy. The principle holds if the switching time is non-zero, but the math becomes less clear.

The principle only holds if you are comparing switching between north-south or south-north. If you compare switching to not switching, it doesn't hold.

Let the switch cost something like 1 unit and the analysis on the Wikipedia example still holds - at the switch, drivers face a choice between a 41 minute route and a 45 minute route. Yes, this paradox is a property of the specifics of the network in question, but it may hold for real networks is the claim.

>This means raising the price of driving on a road when demand is high. During rush hour, drivers would have to pay a fee to use the most congested roads.

Drivers already pay for the cost in lost time. It's a 'natural' market.

The main problem with this article is the idea that the goal in traffic policy is to reduce congenstion. No, it should be to maximize utility. Sure, if you reduce roads, add congenstion charges then people will find alternatives to cars. That doesn't imply that utility is increased.

Public transportation is typically worse than using a private car. You can't decide when to leave, you pretty much need to plan your trips around public transportation. They don't take you directly from A to B etc. Public transportation becomes effective only when population density becomes high enough.

> Drivers already pay for the cost in lost time. It's a 'natural' market.

Yes, but it's a negative sum market. The lost time the drivers are paying is lost, it is destroyed, hours that could have been spent in productive work, or in enjoyable leisure are being wasted. Setting up the market to work in dollars rather than hours avoids this. The dollars are transferred but no resource is squandered.

This is typically true in North America, but it is not a fundamental law. Particularly, in Europe, but also for some trips in Sydney, Vancouver, et cetera, I have found public transit superior to a private car. The key is to have as wide of a network as possible, frequent service (every 5 minutes so you don't need to think about schedules) and a average speed that is higher than the car (doable with trains, dedicated bus lanes and so on). Anywhere that there is enough traffic to cause congestion ought to be a viable place for public transportation . . . population density is not necessarily the driving factor. In fact, in some situations, a high population density might allow people to move less as they are closer to shops, work, friends, et cetera, rendering transit less important.

No, the problem with public transit in North America isn't some fundamental flaw with public transit, it is because we tend to half-ass it.

Population density is key to mass transit otherwise you have to spend a lot more money to provide frequent and convenient service. You can't get around the need for high enough ridership to make a transit system financially viable. Sparse population density = higher per rider cost = more difficult to offer frequent service without losing a lot of money.

In North America part of the reason we don't have as much mass transit as other parts of the world is because we like our urban sprawl and that makes it difficult to make the economics work out.

> make a transit system financially viable

I don't know about other parts of Europe, but where I'm from public transit is most definitely not financially viable. Even in the big cities, it produces millions of euro of losses each year. Those are covered by taxes.

The problem in North America isn't [just] the suburban sprawl, it's also that you want to make a profit off of everything. Sometimes some things just need to be a public service because it's better that way.

Roads are (mostly) not free, and also cost millions of (insert currency unit here) which are covered by taxes. In particular, the asphalted road in a residential street occupied without charge by someone's expensive vehicle is being paid for by, among other people, me -- and as a non-car-owner, I gain essentially no benefit from the provision of this parking space. So clearly, we've opted as a society for subsidizing public services -- but only certain services.

I think the difference is in the perception that roads are static and once built just sort of stay there for everyone, whereas public transit has a lot of running costs. Sure roads have a lot of upkeep cost as well, but it's not as obvious.

Then again, the roads I've seen in the US are so shitty I'm [almost] certain there's no upkeep being done anyway.

The way this works in municipalities or low-population regions of states (which is what most people experience traveling through 95% of the United States' road miles) is that roads are left to deteriorate until such time as they can't avoid fixing, and preferentially might be completely rebuilt which takes a different flavor of money that can be subsidized by the federal govt.

Regions with affluent populations and dense metros essentially have their own road budgets and taxes to keep conditions acceptable.

The other problem in the states is that the weather varies very strongly; this coupled with snowplows and road freight does a number on roads in short order. Some stretches have to be resurfaced yearly. Doing this is expensive and disruptive, so sometimes it is only done in the worst spots if the other parts can wait.

Roads have to be maintained, the cost is non-trivial.

The roads in the US are absolutely a dream compared to say Indonesia, or even China where new roads tend to crumble after a few years of iffy maintenance.

I can think of a few benefits you get from your neighbor's parking space, but the most important is the increased economic output of your neighbor due to their happiness with suburban life.

The challenge for mass transit systems in the United States is that if the operating costs are so high that it requires heavy subsidization from taxes it becomes a very tough sell politically. "Our taxes are being raised by how much to subsidize the subway???"

Perhaps that's just a failing of how mass transit systems are sold to the public though. If your mass transit system takes several hundred thousand people off the road every day and significantly reduces gridlock it makes everyone's commute better and wastes a lot less time and gas. Maybe that's how these systems need to be sold rather than, "It'll run profitably after X years."

They also promote economic activity (people can get to work and spend less time commuting). Even cities in red states are striving for more of that (though to be fare, even those cities are mostly blue).

Mass transit in the USA used to be mostly operated by private companies, and yes they made a profit.

I don't think anyone really expects government-run mass transit to be profitable, I for one would be happy if it were less of an enormous money pit. Given the level of corruption in most public transit authorities however, I don't hold out much hope.

The private companies couldn't compete with public subsidized roads that came online in the 20th century. If you want to see transit companies making a profit, try highly dense Japan, and you'll still pay more for it.

As if the USA was a third-world country with a high level of corruption compared to say...China, which many conservatives/libertarians "praise" for its cost and efficiency.

'Drivers already pay the cost in lost time.' Yes, but not all of the cost. If each additional car adds 10 seconds to commutes, each driver will only pay a pittance more for her decision to hit the road. Meanwhile they are imposing large costs of 10 seconds * each driver on the road. Look up the economics of externalities again and you'll see that an additional congestion charge is well founded in theory.

I'm curious why this is being downvoted. Isn't congestion a textbook externality?

Assume 100 drivers each cause 10 seconds of slowdown.

Assume every driver is equally at fault.

Each driver is causing 1000 person-seconds of slowdown.

Each driver is receiving 1000 person-seconds of slowdown.

Seems fair to me, and each driver is paying their exact cost.

It's wrong to look at the externalities drivers output and ignore the externalities drivers input.

No, you have to look at the margin. The other 99 people are already there, and you have to choose whether or not to drive; you slow down everyone after you, but you don't slow yourself down.

Everyone before you chose to inconvenience all comers before and after. In a situation like this, especially because it's iterated every day, these people are clearly choosing to slow, and be slowed by, everyone coming after them. Being on the road first doesn't give you moral superiority.

In some situations the margin of newcomers is a fair way to look at it. But with drivers on the road, these people are all getting up in the morning and choosing to use this road at the same time. It is a continuous reaffirmation, day by day, that this is the path they want, no matter who they slow down, and they accept being slowed in return.

Each driver is causing themselves 10 and everybody else 990. Your analysis acts as if people make decisions as a group.

"Each driver is paying their exact cost." Yes, but on the margin each driver can only choose a small chunk of their cost. Seriously, have you looked at the enormous literature on this? One major annoyance on HN is the way CS people assume that they can independently grok the intricacies of any field.

I'm not trying to argue about the intricacies, I'm pointing out that your original example was terribly flawed.

Assuming drivers actually consider their effect on congestion, they are willingly accepting the congestion of those coming after just as much as they accept the congestion they cause. (Most probably don't consider it but that's beside the point) Most importantly, the drivers are choosing to use the road every day even after seeing the final congestion every day. That's not a marginal decision. They are choosing to accept the final congestion.

But even if you insist on a marginal explanation with actors incapable of anticipating future cars, the average driver is intentionally accepting and causing 500 with past cars, and unwittingly accepting and causing 500 with future cars. Versus your characterization of them only intentionally having 10 delay.

Framing it as a new car 'paying a pittance more' is nonsense. The twentieth car either doesn't travel or attempts to pay 200 to travel. Comparing 190 to 200 is meaningless, because they never had the chance to travel at 190.

You need to count the people who would like to use the road, but must choose their next-best option because the road is too congested for them to use (either that time, or in general).

There is a congestion externality on them despite not using the road.

That's fair, but it probably leaves drivers feeling at least half of the effect they cause, rather than the fraction of a percent implied by seizethecheese.

But that's highly inefficient. Everyone pays in the form of time, but some peoples' time is more valuable. Congestion pricing fixes that.

Some people's money is more valuable as well. Congestion pricing simply tips the balance of the system in favour of people who value time more than money.

Well except that the people paid the least per hour are also the ones who can afford to lose time the least and have the least ability to pay for congestion.

So the people who can least afford to lose time will avoid congested traffic and lose less time? Sounds like that could work out well.

If you wait an extra hour before commuting home because of the cost, I don't call that "lose less time".

> Public transportation is typically worse than using a private car. You can't decide when to leave, you pretty much need to plan your trips around public transportation. They don't take you directly from A to B etc. Public transportation becomes effective only when population density becomes high enough.

When PT is done right, it's much more predictable than private cars. Being able to leave when you want does you no good when you're dealing with unpredictable traffic.

Yep. Solid PT runs on a demand schedule too: during hours heavily traveled a train might run every four minutes; during the dead of night it may not run at all or run infrequency, like every forty-five minutes.

And to the GP's point, they most definitely take you from point A to point B. Those two points may not be where you are and where you want to go, though with a solid network it'll usually be within ambulatory distance.

;)

> Drivers already pay for the cost in lost time. It's a 'natural' market.

Except it's not: unless you're extraordinarily prescient (or have access to real-time, accurate-to-the-meter traffic data), the driver cannot act as a rational actor because he has imperfect knowledge about the transaction. And that's assuming he'd act like a rational actor anyways -- and anyone who's studied real life instead of theory will tell you that most actors are not rational except, sometimes, in aggregate.

> Public transportation is typically worse than using a private car.

Says who? Instead of spending time stressfully navigating through a slow maze of cars, you can read a book while quickly and predictably getting to your destination. (Public transit is substantially more predictable than traffic.)

I'd take a train over a car every time.

In big cities like Tokyo you can just go into the subways/train lines and demand is high enough that you'll only wait maybe 5 or 10 minutes. Not to mention that it's faster than a car most of the time (due to no red lights or the like).

And if you really need A -> B directly, there are taxis.

In the same cities owning a car is also inconvenient, it's not like you can park anywhere.

The fact that public transportation sucks in a lot of cities is not due to intrinsic qualities of these solutions, it's a consequence of shitty urbanisation planning that makes it hard to offer a good public transportation strategy.

Have you not ever been in a traffic jam?

'Traffic jam' is not a good concept to talk about. You need to talk about utility. In more concrete terms, you need to talk about travel time. Talking about whether or not something is a traffic jam or not is simply losing information.

I'm not interested in the definition of a 'traffic jam'. I'm not interested to know if I've ever been in a 'traffic jam' or not.

It's an interesting effect, and fairly obvious once you hear the explanation the first time. This article focuses primarily on the effect that roads have on their own congestion, but I think you can do similar analysis on the effects of roads on many other things.

This sort of idea is discussed a lot in the libertarian community. I don't intend to make the conversation about that; I just want to mention the ideas because they're similar. Basically, a common and oft-parodied criticism of libertarianism is "but who will build the roads," implying that massive road infrastructure is something that can only be produced by government through taxation and eminent domain.

One common retort by libertarians is to come up with a bunch of proposals for how private enterprise could create a very similar road system to what we have today. I think that's the wrong approach. I'm more interested in the ways that government roads, no matter how well-intended they are, can lead to inefficiency and perverse incentives.

I think the biggest and easiest argument to make is an environmental one, which is somewhat ironic, considering that libertarians are generally considered (by themselves and others) as a threat to environmentalism. And yet, I wager that public road infrastructure is one of the most obvious cases where government programs lead to (probably unintended) environmental problems. The interstate highway system is a massive blow to railroads, which have a vastly smaller environmental impact than road freight. The public road funds essentially subsidize the price of fast shipping via truck. Big trucks are probably responsible for a disproportionately large portion of road wear relative to the funding of roads that they provide.

Another interesting effect is that urban sprawl probably wouldn't happen without roads which are primarily produced by government. This is probably another big threat to the environment, since I would imagine that suburbs pollute much more per capita than urban areas because of the necessity to use automobiles.

The Chesapeake Bay-Bridge Tunnel is one of the largest bridge-tunnels in existence. It's 17.6 miles long, and was financed by toll-backed bonds, and continues to be maintained with toll money. The point being that large infrastructure projects can be financed by the free market, provided they require the users to pay for them in the future, allowing investors to profit.

Here in the DC region, demand-based pricing was used in the widening of interstate 495. It works. My favorite aspect is that buses and vehicles carrying 3 or more people can use the lanes for free, while all the assholes riding by themselves have to pay handsomely for the privilege of creating congestion at rush hour. http://en.wikipedia.org/wiki/Virginia_HOT_lanes

One of the funny things about the libertarian idea of making everyone pay for road use is that technology has made it much more feasible. We don't have to use toll booths anymore. Instead just stick a transponder on the window. One of the best remedies for situations in which overuse of a public good is causing environmental problems is to make people pay to use said good.

Everyone paying for road use is not a libertarian idea. It's the basis of the gas tax, for instance, which funded the Eisenhower interstate system--not usually held up by libertarians as their ideal.

Likewise the Chesapeake bridge-tunnel was built by a government organization funded by municipal bonds. Bond funding does not imply private ownership and is not really a libertarian idea.

The 495 HOT lanes were a public/private partnership, so again not exactly a libertarian ideal. Also, to date they are underperforming financially.

The best example of a libertarian-type road project in the DC area is probably the Dulles Greenway, which was built and is operated by an entirely private company.

I think the "roads" problem is just a simple and easy example of the general class of collective action problems that require a single authority to manage. Here is another example: the design of a network cannot be modified arbitrarily.

It's rational for a private company to insert a road into a system to benefit itself at the disadvantage to others (cf. negative externalities). Indeed it's rational for several to do this making ad-hoc modifications to the network, the net result of this is significant inefficiency.

Collective action problems and (negative) externalities are the main cases where we require a single-point-of-trust / authority to coordinate a system. Libertarians say that private companies can do this, but when you ask "well what do these companies look like?" you invariably get something like a small government.

The moral question then arises: we have these single-points-of-trust, do we have any democratic control over them? The answer, in a libterian system is No (money != vote).

The libertarian solution to these problems then, seems to me, to hand them over to private tyrannies with no public accountability to perform largely the same function as a government. Its at this point the "NOPE!! LEAVE ME ALONE!" tends to come out.

The collective action problem sounds good in theory, but the concept breaks down when you realize that the formation of a government (at least a democracy or republic) is also a collective action. The idea is that we can't form a market to efficiently create certain things like roads, but that somehow we can form a government which can efficiently create those things. Market failures, like the failure to produce a public good, are not defeated or circumvented by forming a government, because government is also a public good.

> It's rational for a private company to insert a road into a system to benefit itself at the disadvantage to others (cf. negative externalities). Indeed it's rational for several to do this making ad-hoc modifications to the network, the net result of this is significant inefficiency.

It's also rational for a government to insert a road into a system to benefit "itself" (i.e. politicians, construction companies that are "friendly" with government, etc.) at the disadvantage to others. What is your response to my environmental argument? That's an argument about specific negative externalities that are produced by government that I would not expect to be produced by private infrastructure.

> The moral question then arises: we have these single-points-of-trust, do we have any democratic control over them?

It's just as reasonable to ask that about government roads. Does government's decisions about roads reflect the desires of the population? I haven't seen any data, so I don't really know, but I would suspect the answer to be mostly no. But markets often do have the quality that their products reflect the desires of the population.

> The collective action problem sounds good in theory, but the concept breaks down when you realize that the formation of a government (at least a democracy or republic) is also a collective action.

Formation of a government may have these problems, but it's done once every now and then, and then in between we have stable periods of consensus. Yes, some will disagree with this, and some with that decision, but your solution involves spreading this same mess, akin to government formation, over every single activity where a consensus would be beneficial!

Your parent didn't claim government was the most efficient way. Your parent just claimed the public has more control over government, than a company operating as a pseudo-government serving similar functions.

Does government's decisions about roads reflect the desires of the population? I haven't seen any data, so I don't really know, but I would suspect the answer to be mostly no.

Well, if that isn't a convincing argument...

> Your parent just claimed the public has more control over government, than a company operating as a pseudo-government serving similar functions.

And I think the claim that public has more control over government needs to be backed up by some sort of data. Of course, you'd need to define the term "pseudo-government" as well.

> Well, if that isn't a convincing argument...

It's not meant to be a convincing argument. That's why I explicitly said that I don't have any data, and that I simply suspect something. Do you have any data? There are plenty of studies on whether the public approves of the general behavior of its government, and at least in the USA the numbers aren't great. My suspicion is that the numbers wouldn't be better when it comes to roads, based on the frequent mismanagement of roads and the number of complaints I here about roads. Again, not data, just suspicion.

> And I think the claim that public has more control over government needs to be backed up by some sort of data. Of course, you'd need to define the term "pseudo-government" as well.

For pseudo-gov read "multinational". As far as data goes, I cant think of a principle more well-evidenced. A single person who is not a consumer of a corporation has no influence whatsoever over the decision making of that corporation.

I can go to my local representative and act in my local community (petition, etc.) in which my local representative has a stake and he can do things for me.

The tremendous amount of public shaming the government goes through (compared to private corps) should be evidence enough that there's a considerable amount of public accountability.

You're still not showing any data, and I would question your conclusions. Just because it seems like you have a say in politics doesn't mean you actually do. If, for example, public policy has no or little positive correlation with public desires, or the public is extremely disappointed with government (both of which are the case in many polls and studies), I would question the presumption that people really have a voice in politics.

The 407 in Ontario is privatized (well, PPP, actually) and they add lanes all the time. The traffic is almost never gridlocked.

The main reason highways exhibit the behaviour is because they are not load-costed, as the 407 is. The cost per peak traveler on the 401 (an alternative-ish public highway of similar size / etc) is around $10k per car per year, due to the immense cost of building new lanes, the slow down that costs, and the additional slowdown of everyone else on the road. And that cost was when I was studying engineering back in 2007, so it's almost surely higher.

I think "cost per peak traveler" is a pretty useless metric. But even still the math doesn't work. 401 is 4+ lanes, 500 miles long, meaning peak travelers must be up near 500,000 cars. Now, apparently the Ministry of Transportation maintains the 401 (and much more than that) and their entire estimated budget for 2013-2014 is $1.5 billion.

Cost per use seems a lot more interesting. It's hard to say exactly how many cars ride it every day, but Wiki has a table of average daily utilization at 11 points along the system. If we simply add those up, then in 2008 they counted an average of 1.1m cars per day on the highway, or 401 million riders per year. This almost certainly significantly understates utilization.

If the Ministry of Transportation spent their entire budget on the 401, that would work out to about $3.70 per ride. In reality they are probably investing less than $0.25 / ride to maintain and extend the highway.

There is very little, perhaps nothing, that can compare to the cost-benefit of building freeways for your nation.

> can lead to inefficiency and perverse incentives.

What is the alternative. Can you show a single developed country with a successful private transportation system? You talk about trains. Most countries with functional and effective train system are not libertarian today. In most places they are mandated, standardized, subsidized by governments. Bullet trains, freight, safety rules, etc.

Showing that public roads lead to inefficiency and perverse incentives doesn't point to a solution. It just points to a problem. You still have to find a solution that works.

I think this argument is bit of cop-out, well we can't make argument how libertarianism would work so we can poke at something government subsidized and tell everyone how it is broken.

To me it seems this is not unlike Communism argument. "It is so nice in theory, we'll just share and distribute the wealth blah blah". "Ok, can you show me an example of a successful Communist country?" "Well no, _but_ let me point out some problems and inefficiences with the capitalist system".

> Can you show a single developed country with a successful private transportation system?

Not on a national scale, no, unless you count any passenger rail or air systems as "successful" and "private." But that's not really a sufficient argument. Obviously, it would be nearly impossible to compete with taxation and eminent domain at the scale they are used in modern developed nations.

> Showing that public roads lead to inefficiency and perverse incentives doesn't point to a solution. It just points to a problem. You still have to find a solution that works.

What do you mean I have to find a solution that works? Does that mean I don't get to propose new solutions and argue for why I think they should work?

> I think this argument is bit of cop-out, well we can't make argument how libertarianism would work so we can poke at something government subsidized and tell everyone how it is broken.

I can make an argument for how libertarianism (or rather, a specific societal organization that some might consider to be "libertarian") could work.

> To me it seems this is not unlike Communism argument...

Well, I don't think communism works well in theory or in practice. I am advocating for markets, and there is a great deal of theoretical and practical evidence for why markets perform well.

> no, unless you count any passenger rail or air systems as "successful" and "private." But that's not really a sufficient argument.

Well actually that is not bad. I mean air systems are private and they work. So a good argument. Not without heavy regulation but nevertheless. Now, one can argue that looking at just aircraft flying in the sky is not enough, one has to obviously look at airports. And there it is back to government allocating and sponsoring development.

What about passenger rail? I can't think successful modern national-level private passenger rail systems.

> Does that mean I don't get to propose new solutions and argue for why I think they should work?

Sorry, I meant the difference between just showing that highways produce some inefficiencies. That is not hard to do, it is a simple argument in a way. Some imminent domain is taken, government paid for some of it. Makes rail more expensive, so can't quite compete with it. Is that a basic rehash? A more interesting argument is to propose something that works. Say, privatize roads. How would that work and why are there no large scale examples of it, if it is supposedly a good idea.

> I can make an argument for how libertarianism

I guess I misunderstood your argument as more being about showing an inefficiency in a government run project. That is not too hard. But what to do about it. It might seem "privatise" is the clear answer. But the other answer is "not have it at all" or have something else, or even enhance government control over it.

We are almost there with the post office. It is inefficient, outdated, and dying. But it still delivers mail to places where it makes no financial sense to deliver for someone like FedEx. If USPO shut down, I think necessarily in many places there would anything in its places. Maybe that is the correct answer or maybe it isn't.

This is probably another big threat to the environment, since I would imagine that suburbs pollute much more per capita than urban areas because of the necessity to use automobiles.

Maybe. Or, while we're speculating, maybe spreading out the impact (even if it is in total greater) is good. Highly concentrated human activity causes very dramatic damage to the environment. If you spread the same activity, or even more activity, over an area many times as large the impact may not be felt.

There's quite a lot of studies on this that indicate the opposite:

"It has been shown that cities of a high density, such as, for example Hong Kong, have a far lower transport energy demand per capita than low density cities such as Houston, by a factor of 18. On average, when comparing 10 major cities in the US with 12 European cities, European cities are five times as dense but the US cities consume 3.6 times as much transport energy per capita. The conclusion often drawn from such data is that dense cities are low energy cities." [1]

You are correct that there is more immediate impact to the environment from dense urban areas, but it's outweighed by the gains in energy savings from sharing resources (energy, transportation) effectively. This is why contemporary urban planning tends to push compact, dense neighbourhoods over suburban sprawl.

[1] http://www.sciencedirect.com/science/article/pii/S0378778802...

I think you completely misunderstood me.

Let's compare the following two scenarios:

- 10,000 people live in log cabins in 1,000,000 acres.

- 10,000 people live in 100 acres.

Let's suppose the people living in log cabins use a hundred times as much energy as the second group.

Clearly from an energy standpoint, that isn't good. But the second group has completely and utterly decimated 100 acres, while the first has minimal impact on the land.

Basically, in summary, even if consumption goes up as we spread out, the amount of impact experienced by a given area of land decreases as our impact is diluted over a larger area.

Phrased differently, it's easier for native organisms to survive if humans are spread out, than if we bulldoze their habitat and build highrises.

10'000 people living in 10'000 log cabins, each of 50m² (5 × 10 m² sounds reasonable to me) will cover more space than 10'000 people living on 100 acres (= 0.4km²). Furthermore, if you consider a single spot of 0.4km² populated in an area of 4047km², you will likely find that its impact on the vast majority of those 4047km² is absolutely minimal, whereas, if my WolframAlpha-foo is not completely wrong, there is no point further than 400m from the next human settlement in your first scenario.

This is a false dichotomy; in reality, impact on the natural landscape does not decline with housing density.

Suburbs are not little log cabins surrounded by wilderness. They are big houses with big yards, garages, roads, schools, shopping malls, parking lots, etc. At even lower densities, in many rural areas houses are surrounded by farmland, which is also a substantial alteration of the natural landscape.

In addition, energy has to come from somewhere, and energy development also impacts the land through mining, drilling, refining, transporting, and using the energy sources. Even solar and wind have impacts on land use and must scale with demand. So it's incorrect to disregard energy use, because it will impact the land somewhere.

> Phrased differently, it's easier for native organisms to survive if humans are spread out, than if we bulldoze their habitat and build highrises.

This is just 100% wrong. The habitat gets bulldozed (or plowed under) in any case when humans move in and start developing.

9 feet for one lane road between 10,000 cabins in a line 209 feet apart leads to 431 acres of road area. i.e. your less dense population already destroys more road area than the denser city did. It is an unfortunate reality that dense cities are orders more efficient for a modern society than spread out ones.

Phrased differently, if we have to bulldoze anyway, we should bulldoze the minimal surface area.

Let me know if messed up the calculation, I never used acres or feet.

The second scenario leave 999,900 acres completely untouched while the first scenario probably impacts at least 5,000 acres.

The answers libertards give to each other and those we give to authoritarians are often different.

To other libertards, we have to come up with a well-reasoned, thoughtful argument, with support and contingencies.

To authoritarians, we have to use a glib sound bite. Who will build the roads? Without government, who would possibly make the hamburgers and blue jeans?

In reality, a libertarded city might not have any roads at all, like Kowloon Walled City before they tore it down. It wasn't necessarily a nice place to live, but it got on well enough without roads.

Some libertards are kooky enough that they would build a city around ziplines, sleds pulled by teams of housecats, for-hire piggybackers, and go-go-gadget hatcopters. If there were any cars at all, they would be made of hemp, have built in pistol holsters instead of seatbelts, and be powered by Stirling engines burning moonshine. Visit New Hyperbole City today!

But seriously, either someone will build a road at their own expense, or everyone else will adapt to the lack of a road. What you certainly will not see are paved county roads with one mailbox every half-mile or so. A massive system of autobahns will not spontaneously appear, and it certainly would not be free to use. Rail networks, as you mention, would almost certainly more efficient use of private funds. People living where roads are not economical would simply buy off-road-capable vehicles and chainsaws instead of paving their own. Libertarded CONUS would start to look a lot like Alaska, except with more fanboats, ATVs, mudboggers, dune buggies, ultralight fan-gliders, mountain bikes, zeppelins, and such. People who like cars would have to pack in closer on fewer roads, and pay for their use directly.

There was more or less a whole EconTalk episode about this recently; one interesting point from it was the (bad) concept of STROADs:

http://www.strongtowns.org/journal/2013/3/4/the-stroad.html#...

Thanks for this; it's a valuable concept I couldn't have articulated before. It seems I see stroads everywhere, and they do suck for pedestrians, cyclists, shoppers, businesses, and humans, but there isn't an obvious way to get from here to the improved future. At a certain population density, land simply is less valuable, no matter the road organization. I can imagine a town trying (and enforcing!) the street concept in a couple-block area, but how effective can that area be when everyone drives to get there? I almost suspect we'd need regular bus service, but at low population densities how does that work?

Use automated vehicles to supply an on-demand middle tier of transit. That layer could possibly provide transportation bridging low-density regions to a higher capacity fixed public transit nodes.

The Stroad to street transition seems difficult to manage. The main street in my hometown underwent some of the transition, going from 2 lanes of parallel parking and 4 traffic lanes to diagonal parking with 2 traffic lanes and a turn lane, but the wide patch of asphalt is still there, and business isn't really thriving. I guess the general economic malaise there makes it hard to score development concepts on their own merits.

It's interesting to look for the accidental victims of development, stuff like houses backed up against a road that probably started as a street. In areas where there is lots of room, the sensible thing is probably to just tear them down (or relocate them if that made sense), but people are too busy looking at the immediate impacts and ignore the needlessly low quality housing that will exist for decades (your link points out that the housing also makes the road less effective, which I hadn't considered).

Large swathes of Houston are made out of stroads.

Having done some grad work in this area, I'd say this is a fairly superficial take on the topic. This certainly describes one area on the supply/demand curve, and perhaps even a pervasive one given that we essentially stopped building highways 40 years ago, but it's not inherent to the problem.

I think it's very possible to jump ahead of this effect, but the capital outlay would be impressive, and as a society we apparently aren't permitted to accomplish Big Things anymore, so it's likely we stay in this weird bit of the function for a long while.

Who stopped building highways? In the Dallas area we've been building and expanding them non-stop for decades now. My experience is people have simply moved further out, increasing traffic flowing into the newly expanded/built roads, and overall congestion has remained constant.

But I think many of us would agree about accomplishing Big Things. Here's the dirty little secret: the US Government has almost always funded those "Big Things" and taxed the rich heavily to do it. The Interstate highway system was funded by a massive (as a percentage of the price per gallon) gas tax, along with a 90% post-war income tax rate on the rich.

Even the trans-continental railroad was funded thanks to guarantees, loans, etc from the US Federal government (including sending the Army out shoot Native Americans and bandits who attacked the railroads). No private entity was willing to take such a large risk - it could have bankrupted even JP Morgan if it went bad. Without "Big Government" support the line never would have been built.

I'm all for bringing back much higher taxes on the rich - tax anything over $1 million income at 50%, over $10 million at 60%. Same for cap gains. Use that money to fund massive expansions of roads, build subways under every major city, heck start a government-funded bank that gives anyone who wants to start a business a loan (it's what the Chinese government does).

We could do a ton of "Big Things" if we were willing to tax the uber-rich to pay for it.

Taxing the rich simply incents them to behaviors that avoid the tax. Government revenues increased dramatically when the high marginal tax rates were cut. Unfortunately government expenditures increased even more dramatically.

I would contest that many roads are not built or expanded until they are needed and due to that the demand fills out the road on completion. If you were to take a 3 lane road that is at 70% capacity and make it 6 lanes I very highly doubt you would get a 6 lane road at 70% capacity. They reference a % thought an entire city, on a more Macro level a single bigger road can indeed ease congestion, though It may cause other roads leading into it to become congested. The only way out of the trap AFAIK would be to build excess capacity thought the entire system.

It's a lot more complicated than that, if I remember my research on this subject correctly. It is definitely true that in a perfectly abstract system that if your roads are 70% utilized across the board, and you double capacity, you won't stay at 70% capacity.

However, what does happen is that when you take a road at 70% capacity and double it, so it has a lot more room, some people will move from other 70% utilized roads to your road, and yet more people will decide to drive instead of taking the bus or train, and the utilization of the whole system will remain largely flat at ~70%.

This was seen time and time again in NYC, where even at the peak of construction of roadways in and out of the city under Robert Moses, every new bridge, tunnel, and highway was at >80% utilization within days or weeks of completion. IIRC, the total amount of major roadways in and around NYC was increased by 5-10x from the 1940s to 1950s, and yet traffic congestion and utilization actually increased over this time.

It took a number of years of urban planning research to identify that as you increase the number of roads, more people opt to drive in lieu of mass transit or not making the trip at all.

So, i would argue that in an abstract sense you are correct, but in the context of a major city, a single major road change does not ease congestion, as it simply changes traffic flows and increases demand.

% capacity and % utilization are measures of congestion - not of demand. If your roads are only meeting 10% of latent demand and you double their capacity, guess what? They will be just as congested as before.

Leave it to "years of urban planning" to determine a proper solution is to make roads so poor that people decide not to use them.

> Leave it to "years of urban planning" to determine a proper solution is to make roads so poor that people decide not to use them.

What a weird level of snark. At least in NYC, demand is effectively infinite. Close to 2 million people commute into Manhattan every day, not counting commercial vehicle transportation. Building enough roads to satisfy all of those people is theoretically possible, but leads to plans like these:

http://gothamist.com/2010/01/16/1924_traffic_congestion_solu...

I can't find it now, unfortunately, but there was a serious analysis done of what would be required to accommodate all of the potential demand for car transit into Manhattan, and it was insane how many bridges and roadways were required.

Thats' why you can't look purely at demand. As you make it easier to drive into a city, more people will opt to do so, thereby increasing both demand and capacity.

You even see this in US cities like Atlanta that are not islands, where increasing demand led to more and more road construction, which drove more and more people to live further and further outside the city, since they could just drive an extra 15-20 minutes each day. In the end, the only thing that capped demand for Atlanta was rising gas prices, making it more expensive to commute.

That's why you talk about capacity and utilization - demand is effectively infinite until other costs and considerations reduce demand.

NYC is the most extreme case. A good local example of is the San Mateo Bridge, at rush hour the San Mateo Bridge / 92 interchange backs up the 101 all the way to Woodside (several miles) but clears up after that. By the time they get around to expanding the bridge or adding another one (ha, when pigs fly) the demand will cause whatever the replacement is to be jammed as well as I would assume many more people would live in the East Bay which is cheaper.

Yeah, this explanation is so obvious to me, I just don't understand how they can try to spin it any other way.

There are MILLLIONS of drivers in LA, I doubt we could build enough roads to make it so that there is never any traffic.

I wonder if we're missing something here? Maybe the other and parent are leaving out some key details, but I doubt it.

I guess we would have to do something that requires effort like actually examine why people need to drive at all. The real problem is probably workers being denied the opportunity to live close to their places of employment due to predatory pricing and poor services downtown. But then we couldn't put the blame the average working stiff or use it as an excuse to decrease public works spending.

There's another question that we are starting to battle recently: how much actual capacity can we honestly add to the road system?

Near my parents' home, there's an intersection of a 18 lanes (!) of highway + access road crossing over a 9-lane surface street. I just don't know how you grow that any further, without the prohibitively expensive option of more vertical stacking.

I've always thought of our current car-based transportation system to be amazingly inefficient when measured in energy consumption per passenger and travel time. But I'm noticing it's also extremely greedy with land use.

It's not just greedy with land use: it's literally hostile to humanity. In places in the US with a large driving culture it is almost a death wish to walk anywhere (because none of the drivers are expecting to have to watch out for a pedestrian).

> The only way out of the trap AFAIK would be to build excess capacity thought the entire system.

I agree. This makes a lot more since.

The authors are taking good data, but making a big leap on their assumption to the cause.

Roads don't "create" traffic. New roads in most major cities are never built until the old ones are already 300-500% over capacity.

If you wait until your 500% over capacity, and then double the lanes, your still way over capacity, so the old and new roads are still "congested". This traffic isn't "new" -- it was always there, you just couldn't see it, because a 200% over capacity road, and a 500% over capacity road, looks like the same gridlock from the sky.

No, it doesn't. Why can't some of you understand basic economics?

For simplicity, let us have City Core (C), the suburbs reachable with a one hour commute at the worst (S), and the rural area/farmland (F) further out, and ultra-rural deliverance areas beyond that (D).

No one wants to go buy a piece of Farmer Joe's land in zone F and build a house when it takes 3 hours to drive to work. Therefore relatively few people live out there. Because so few live out there, many other people don't want to live out there (not much shopping, no entertainment, few schools, etc).

Now the road is widened. In C, it goes from six lanes to ten. In S, it goes from four to eight. In F, a new closed-access highway with four lanes is built.

What is the result? Why spend $650,000 on a house in S when you can build the same house in F for $300,000? Which is in fact what a ton of people do. Over a period of ~20 years, developers buy up the farmland in zone F, subdivide it, and build houses. People move for the cheaper/larger housing; businesses move in, retail is built, and now we have zone S'. You have four lanes of traffic from zone F connecting to an expanded eight lanes in zone S, which was originally four lanes. Can you spot the problem? Hint: 8 - 4 = 4. Thus the "effective" number of lanes is unchanged for the people living in zone S, even though we presume both S and C have increased density by building taller buildings, renovating to subdivide houses into apartments, etc.

Same for the people in zone C, where the four additional lanes of traffic from S' eat all the excess capacity from the expansion.

What was zone D becomes the new zone F'. Soon as they build some excess road capacity (damn hippies and liberals! what a bunch of Phooey, we should just build more roads) all our problems will be solved!!!

---

So yes, in a place like NYC with massive over-demand for transportation, you are correct. But here in Dallas, we've been building tons of new roads all over for the past 50 years and it hasn't changed much, it has merely enabled people to buy up farmland and move further away. Plano in the 1960s was all farms, now it's been 100% build on (literally, the city has no further room to grow). Same for Allen. McKinney is almost completely built now. Any further road expansion will require tearing down skyscrapers, double-decking, or tunneling which magnify the cost of building new roads by an order of magnitude or worse.

Dallas is living proof that building more roads won't help congestion. It might lower home prices or allow your city to continue expanding, but it won't help with commute times or congestion beyond a small temporary bump.

Except that roads do create traffic. Your line of thinking was exactly what drove NYC's road construction boom in the 40s and 50s - new road is built, utilization briefly goes down, then, as soon as people realize that there's now a faster, more convenient way to get into the city, they start using it. Utilization goes back up.

Critically, many of those people are people who might have taken mass transit or carpooled. They are literally new demand in the system driven by new capacity.

Over a longer time period, there's an even deeper effect, where new roads and perceived availability of road transit drives more people to move to previously undesirable parts of metro areas, which increases population which increases demand.

> Except that roads do create traffic. - new road is built, utilization briefly goes down, then, as soon as people realize that there's now a faster, more convenient way to get into the city, they start using it. Utilization goes back up.

That's not "new" traffic. Traffic was already critically underserved previously, and is now only majorly underserved. The original road was 500% over capacity, and now the new road lowers it to 300% over capacity.

>They are literally new demand in the system driven by new capacity.

No, the demand was always there. No one went out and bought a car because a new road was built.

These are people who we're already underserved by transportation previously, but weren't counted (because demand so outstripped capacity that they couldn't be easily counted originally).

> Critically, many of those people are people who might have taken mass transit or carpooled.

This is a huge assumption. For instance, this can only even be possible in a handful of places like Chicago or NYC. There is no meaningful mass transit in the most of the major US cities to even pull traffic from.

If public transit is a good option, people will take it regardless of how many roads exist. (Very few people purposely want to make life more expensive or more inconvenient for themselves).

> as soon as people realize that there's now a faster, more convenient way to get into the city, they start using it.

From what you've said, it sounds like your advocating for people to willingly make life more expensive and more inconvenient for themselves. Your saying "new roads make driving too convenient and nice, we should take roads out so driving becomes even more terrible than public transit already is. Force people to make a bad choice, to avoid a terrible one"

- -

There's absolutely no need to race-to-the-bottom on transportation. We should build more roads and build more public transit, simultaneously. When public transit becomes cheap and convenient, many people will willing choose it. For those who don't, new roads and additional capacity alleviates that problem as well.

Your logic assumes that driving is inherently bad, and that we should "punish" drivers as a result. That's terrible, there's nothing wrong with driving, and there's no need to punish people for driving. There happen to be some bad externalities to driving (exhaust, traffic accidents), but those problems are being worked on (electric zero-emission cars, self driving cars, etc)

- -

What your advocating for sounds really dangerous. Imagine if we applied this logic to any other market:

"The internet is congested. We can't just add fiber lines, if people had more internet speeds, they'd just use it up! We should destroy fiber lines, and cap all internet users to 1Mbps speeds, so that people start using the Postal Service and Blockbuster more often. "

"We can't run more clean water to the bad part of town. It might 'drive more people to move to a previously undesirable part of the metro area'. We should keep the water system broken -- or actually, just start slowly cutting water off to that area. "

> "The internet is congested. We can't just add fiber lines, if people had more internet speeds, they'd just use it up! We should destroy fiber lines, and cap all internet users to 1Mbps speeds, so that people start using the Postal Service and Blockbuster more often. "

Love this example. I can already see the wheels turning in some people who prefer a more centrally planned society.... Renting DVDs from blockbuster is more efficient. Each disc transported to an area enables hundreds of people to watch the movie. Sure they can't watch it on demand whenever they'd like, but the DVD rental model is so efficient. Sure, public transportation doesn't take you exactly where you want to go whenever you'd like, but its more efficient! (Yeah, the analogy isn't perfect, but it does make a striking point)

"There's absolutely no need to race-to-the-bottom on transportation. We should build more roads and build more public transit, simultaneously. When public transit becomes cheap and convenient, many people will willing choose it."

The problem with this is that roads take room, and have externalities (cutting up neighborhoods, noise, pollution). Some of these are being worked (as you note) but there are no full technical solutions to these problems. People won't stand for it.

For these reasons, your internet and water-use analogies are inaccurate parallels.

I'm surprised you're offering these arguments. I live in LA and the opposition to new freeways in already-built-up areas is pretty much ingrained at this point. (Conservatives hate the destruction of neighborhoods, and liberals hate the environmental consequences). Limited in-place widening is still being done (with complaints) but 50 years of experience have taught something of a lesson. People generally don't believe these wider roads will significantly lessen congestion.

> The problem with this is that roads have externalities

Public Transit does too. Cutting up neighborhoods, noise, pollution all happen with busses and light rails. The externalities are almost identical in either case. (You can bury your light rail, but you could also bury your roads, etc).

> For these reasons, your internet and water-use analogies are inaccurate parallels.

That may be true, but I don't see how yet.

People oppose fiber lines all the time because "telephone poles look ugly" or "I don't want to be in the electromagnetic field" (even though there isn't one).

> People generally don't believe these wider roads will significantly lessen congestion.

To be clear, I'm not advocating for bad roads. I'm advocating for more roads, and there is a big difference. New freeways should be far above ground (so that there's no division of neighborhoods) or below grade, just like new public transit (usually) is.

I'm also not advocating for more strip-mall like roads. (4-8 lane at-grade streets typical near malls, WalMart, etc). There should be lots of big freeways, with lots of small exits to smaller urban streets (downtown) or smaller suburban streets.

A good example of this is the buried freeways near the Washington State Convention Center and Mercer Island - http://northwesturbanist.wordpress.com/2014/02/17/lets-bury-...

I think we are too far apart on this one. The buried freeway idea seems like a boutique solution for certain niche cases (the big dig in Boston is another example).

But the freeways relevant to me (in LA) are much too large and pervasive to put underground. It's over 100 miles of freeway, 5 lanes each way! And this is only in the most densely built-up core. Undergrounding them is lunacy!

By the time that all got done, we'd be in robotic helicopters or uploaded. Underground trains would be much more realistic (less tunneling for a given throughput), but even so, we have to be content with aboveground light rail.

>Public Transit does too. Cutting up neighborhoods, noise, pollution all happen with busses and light rails. The externalities are almost identical in either case.

Very true! However, they are significantly lower per person transported.

If people were rational you'd be right, but from happiness surveys we know that people underestimate the (negative) effects of a long commute by car. So it's entirely possible that when a new road is built, a family might move out of the city centre to a larger house in the suburbs, expecting this to make them happier, when in fact it makes them less happy.

> there's no need to punish people for driving. There happen to be some bad externalities to driving (exhaust, traffic accidents), but those problems are being worked on (electric zero-emission cars, self driving cars, etc)

They're being worked on, but they're not solved yet, and in the meantime US road taxes don't come close to covering those costs, and the way we treat road deaths as "accidents" (despite a larger, more predictable death toll than e.g. environmental pollution) means the compensation payouts are disproportionately low.

Who cares if it's predictably risky? Skydiving accidents are still accidents. It's bad to be complacent, but it's still an accident.

When you kill yourself because of a predictable risk it's on you. When you kill someone else because of a predictable risk that should qualify as negligent homicide, at least (and lower-risk things than driving do).

Okay, that applies to pedestrian deaths. But the vast majority of driving deaths are based on mutual choice to drive. As long as nobody's drunk it's rarely a 'kill someone else' situation, rather it's a 'two drivers interacted in a way that went wrong' situation.

Negligent homicide should not apply if the negligence was mutual.

That's not the whole picture. The article doesn't directly mention the economic benefits from more people being able to access an area. My guess is that it works something like this:

Let's say you change a 3 lane road to a 6 lane road, and double the capacity. While some people from nearby roads will decide to start using the new road, a lot of increased usage over time will come from the new development that's made possible by the improved capacity. Strip malls will open up, new subdivisions or condos will go in, business centers will open, etc. Which is all fine, but it means that ultimately road usage will increase until some pain-of-driving threshold is reached, at which it'll level out.

They count actual traffic though right? They aren't relying on qualitative assessments, they put those hose things by the side of the road.

Exactly this. Basically the 'demand' for driving is already > than the 'capacity' for driving, so adding capacity doesn't decrease congestion, it stays at 100%. This was made explicitly clear in the crash of 2000 when it felt like roughly 25% of the work force seems to have moved out of the Bay Area, traffic congestion got a lot less.

If the theory of 'induced demand' was accurate, then traffic congestion would not have changed, and that would have been something worth writing about.

This exactly. This article didn't even try to objectively validate the report. It's also provably false in other areas like already not very busy roads in the high desert in SoCal. Add a new road to Apple Valley and I bet it gets used but very little there's just no demand.

This report was highly biased since I'm sure no one built a road where people didn't need one and looked if demand were the same. Instead this is only about building roads where they were deemed to be critical enough to warrant a ton of spending and where the built out supply probably lags demand.

I came here to comment the same thing.

This really sounds like an over engineered explanation and usually the simpler explanations turnS out to be more accurate.

To use the author's own analogy, it's more like a pipe that is being overwhelmed being replaced by a slightly larger pipe that will continue to be overwhelmed.

That said, I hope it encourages people to find creative ways of solving the traffic problem.

Well they tend to focus on one road at a time, either ignore the feeders or the effect the newly expanded feeder has on surrounding roads.

Common around where I live is expanding two lane roads to three lanes and going back to two the closer to the city you get.

I think you are arguing "a priori" - you have a set of given principles and then you derive conclusions. While very powerful when done by smart people, a priori reasoning is often wrong, and should always be trumped by real data. I think one of the points of the article is that the data point to a flaw in the way we intuitively reason about the situation.

So more lanes don't reduce congestion, but this article ignores the fact that they obviously DO increase quality of life, by giving people the freedom to live and work where they choose. We should build more lanes for that reason alone. The reason congestion doesn't get better is simply that there is pent-up demand for driving. The solution isn't "tough luck, adjust your life to fit the traffic." It's to build more lanes, so people can choose to live the life they want to live.

they obviously DO increase quality of life

If you only consider the aspect of freedom, maybe. But there are more aspects of quality life to consider imo. Spending hours in a car, practically standing still, causes stress. More lanes? More stressed people, higher volume of traffic jam - is that still freedom, sitting inside a metal cage while you can't go forward or backward? Increased number of cars driving means increased pollution. More lanes means less space for buildings and nature. Etc. I get your point, but I think there is more to it.

This argument boils down to: "People don't know what's best for themselves, so we shouldn't give them what they clearly want." I can sympathize with that somewhat, but I think that's a dangerous direction for a government to go in.

> they obviously DO increase quality of life

Perhaps not, if you take into account the resource costs of road infrastructure (which are fairly hidden by government) and the health impact of automobile usage (from pollution, commuting hours, and obviously road accidents).

The thing is, massive public subsidies for roads might be the reason that so many people need to travel far from their homes to their work. You're assuming that the causal relationship goes one way, but it could just as easily go the other way.

Short term maybe, but long term it is really just a waste of resources and certainly not a better life for the next generations.

Promoting alternatives outside big cities and spending the money there to attract companies instead of spending it on roads would make much more sense.

By all means, then pay for YOUR ROADS which get YOU to YOUR MCMANSION SHITHOLE. Those of us who choose to be more rational about where we live, rather than adhering to an absurd ideal of a homestead that simulates a farm while having a modern, urban job far away shouldn't have to finance your silly fucking lawn and your boring, awful neighborhood.

I have 2 kids, and I've made urban living work. Courtesy of Virginia's former governor (now on trial for felony corruption charges) our roads are no longer funded by gasoline taxes, and are instead funded by a sales tax that people like me, who walk to work, have to fucking pay. If I had my way, all of you commuter idiots would have to pay a toll for all roads. Its absurd that I can't ride a subway for free, but you get to use public infrastructure while the rest of us pay to maintain it.

Your "freedom to live and work" is actually a freedom to take from urban people and give to yourself to finance your unsustainable, silly lifestyle.

If you want to live in the country, then fucking work in the country. Grow something other than a useless, decorative plant that requires extensive fertilizer and watering.

The solution:

Tough luck, adjust your life to your dumbass decision to live in the suburbs.

As someone who mostly agrees, and has found other posts you've made to be a useful contribution, I think this one strays from that.

Even if it's personally satisfying, no minds are being won here.

You're right. Thanks for the much needed reality-check.

I don't commute.

And Yahoo! employees used to be able to work like that.

You are outnumbered by several thousand.

If you were one of the people that lives close to work, you now have to compete for jobs with people from further away. Which means that you now have to probably drive 30 miles for a job, since that one a mile down the road is taken up by someone who lives 30 miles away.

On a smaller scale, I grew up in the middle of two medium-sized towns, the one north of me had factories, and the one south of me had other factories. In towns A and B, there were about as many people working from the other town as there were local. But without the roads, everyone would be more likely to work a couple miles from home, and probably be happier (type of work, pay rate, etc were about the same between the two).

Good luck with that. What if it is your back garden that get tarmacked over?

In the UK we tried this under the Thatcher government. Margaret Thatcher was allergic to trains and wanted to encourage private ownership of cars as much as possible. Vast swathes of South East England were turned to roads so that happy motorists could get around a few minutes quicker.

Ultimately the Conservatives had to give up on the road-building schemes. It was too expensive to do and the roads really were only encouraging motorists.

The crux of your argument has been tried and tested in the UK. Despite the vast resources and ideology that went into it, it was a battle that was lost.

So what now for the UK? Cycling and buses. It is now socially acceptable to cycle and the government are encouraging it as much as they can in London. Soon it will be 20 mph everywhere in London's boroughs (average speed for cars is ~ 10mph though). There is congestion charging to keep cars out of the capital. With only so many roads, for the commuter, cycling is actually pretty quick. You can also get a bike tax-free with payments spread out over a year.

Buses have been quite transformed in London. The night bus service means that people in service sector jobs can get in or out of the capital at odd hours. Years ago night buses were a nightmare. Now they are clean, safe and quite comfortable if you have some phone/book/tablet for entertainment. The cost is not bad either. People that used to not consider a bus will nowadays take them without hesitation. You can check the times online so you never have to wait an aeon.

Even though we have Conservatives in power - the sworn enemies of public transport - transport in the London area is better than it was. There have been 1/2 million extra Londoners arrive in the last decade or so, no new trains and no new roads. People do spend far too long on the commute - two hours each way is common, but, what do you do when a city gets to be huge?

Personally I wish that the cycle routes were more like 'motorways for bikes' with it possible to go ten miles or so from the suburbs to the centre without having to stop for any lights. Such routes would make commutes by bike all the way a lot more viable. For instance, according to the maps I have 13 miles by bike. I could do that in an hour if there was no start/stop + danger. As it is I cycle for 20 minutes to the train and then get a 35 minute train journey. Those figures sound good, however, I also spend some money on the journey (time=money) plus I have a 10 minute walk at the other end and I like to get to the station ten minutes before the train departs. Therefore I actually spend 1 hour 25 minutes door to door, with a little longer on the reverse journey. So that is 5 hours a week that I could save if I wanted to. What stops me? Those crazy idiots that set out every day to add to the congestion they know is on the roads. I have a quiet ride through the park and some quality time with my phone on the train, thereby avoiding the nutters that are in the dark ages of driving everywhere, single-occupancy style.

I have only read the linked article, not the study, but it seems to me to be a bizarre conclusion that if capacity is increased and utilization remains at maximum, that the capacity caused the utilization.

It seems to be a much simpler scenario that demand far outstrips capacity, and until you can meet the demand, saturation every lane added will continue to be saturated.

I'm sure amdahl's law applies to lane adding here, on top of the fact that all the highways I've seen merge from one side or the other. I don't think I've seen any that merge in the middle, though I guess they probably exist.

It's interesting that they make the point that congestion pricing will induce people to take alternate routes.

The recent update to Google Maps seems to do the same thing. It's much more proactive now about suggesting alternate routes. Here in Seattle, I recently bypassed I-5 by taking an astonishingly complex and twisty route that was all prompted by Google Maps, and I got to my destination five minutes sooner than Google was estimating I would had I stayed on I-5 (although that's impossible to verify since traffic congestion changes throughout the duration of a drive).

All that said, while it was fun (also especially when I avoided a Tacoma I-5 snarl that I hadn't previously known how to avoid), I don't think we actually want to redirect highway drivers onto local streets. Local streets aren't designed for highway overflow; they're designed for local traffic.

Isn't the real answer to design more decentralized town areas so people have less reason to drive long distances, for reasons other than recreation?

Try the math. For a given number of people, centralizing reduces everyone's commutes. The key is to not separate housing from commerce and business.

sounds like a great solution to the problem if you have the luxury of solving it before it is a problem. i don't think that demolishing and rebuilding the metropolitan Seattle area is a viable solution to the traffic issues that area faces.

Funny to see this on the HN frontpage – the Downs who originally suggested the "fundamental law of highway congestion" being confirmed in the paper [0] is my grandfather. Thanks for submitting this, it's very interesting work.

[0] http://www.nber.org/papers/w15376.pdf

Driverless cars have the vast vast potential to fix traffic congestion problems. Traffic mainly results from humans' inability to gauge the spacial and speed based tolerances required to prevent braking and as a result sending a backward brake wave through the cars behind it. Add to that fact that there are no game theory advantages to a single driver to driving properly (other than an altruistic realization that if you decide to drive with better tolerances and other people do the same, the congestion is abated)

Actually driverless cars have the potential to add a vast number of cars to the roads. Today, there's a limit of one car per driver. In the future, one car per same-day package delivery.

True, I'm thinking in a paradigm where humans in cars are replaced with driverless cars. The externalities that you point are valid but I still think that the efficiencies gained by having diverless cars would outpace any externalities (see my lengthy comment about minimum time per car added to a traffic network.)

That seems dubious. There are a finite number of cars you can get on the road at a given speed. If the total travel time is roughly a constant as this article implies, autonomous cars will be equivalent to adding a few more lanes, meaning there will be no long-term change.

Traffic results when car densities pass a certain threshold and the speed the cars are traveling at prevents humans from being able to react in time or appropriately.

You might have even seen this phenomenon in real life: Driving on a highway, suddenly you get stuck in traffic. As you pass through the traffic zone, there are no cars pulled over, no accident, no disruption in the quality of road. What most likely happened was that someone accidentally slammed on their brakes which caused an increase in car density. If there are enough cars that are in bound to this traffic zone before the density can disperse, the traffic zone will persist.

Also notice that traffic zones often appear in situation where there's a positive increase in elevation. What happens is that if people aren't using cruise control (or their cruise control, control system isn't tight enough) their speed will decrease as they ascend the hill, causing some congestion and as like if the density of traffic is high enough, the cars behind them will have to slow down. You may think that if the cars behind them are also ascending the hill they'll also slow down in the same way but some cars may use cruise control while others don't (resulting in congestion) or the differences in each car's torque, power-output, efficiency curves differ causing each to slow down at different rates (often leading to the least capable car governing the rate)

Use a worst-case scenario to analyze. Assume you have a network of diverless cars all linked together that maintain equidistant space between each other. If a new cars needs to join their network / pack the minimum amount of delay that the cars (behind the entering car) need to encounter is the (Length of the incoming car + gap tolerance between the other car) / (the sustained speed of the pack) =25/5280/65*3600 ~= .25 seconds per car. Now do a thought experiment that the high way you drive on was completely full (like we said in our original problem statement) do you think if you added one additional car to the highway, would everyone's (who's downstream of the merge) commute time be only increase ~.25 seconds? I'm very confident it's be much more than that.

Traffic management is a task that needs high coordination, cooperation and reaction times above what unassisted humans are capable of.

The bigger opportunity might be in traffic management. If a control system tries to get humans to do things, they will cheat at nearly every opportunity. Computer guided cars can be programmed to cooperate a little more.

If everything is networked, dynamic tolls also become easier. I guess a lot of people react badly to the concept, but it creates the ability to simply prevent heavy congestion.

This website gives a simplistic visualization of what I'm talking about

http://www.traffic-simulation.de/

Decreasing the time gap and space gap decreased the amount of traffic

A phenomenon very similar to "induced demand" was observed in 1865, when increased efficiency in coal consumption lead to increased demand for coal. It's called Jevon's paradox:

http://en.wikipedia.org/wiki/Jevon%27s_paradox

Yet when I visit a small country town with a giant road through the middle, there is still no traffic there. If big roads induce traffic independent of any other factors then this should not be true - the mere presence of the big road should have created traffic.

While I understand there's a really point to this paradox (it really happens), I hate the simplistic conclusion people reach that it is always wrong to make a road bigger. Sometimes you really do need a bigger road. Sometimes you really need more public transport or something else. It's complicated, and we need better models of it. But reducing it to one line slogans doesn't help that.

That's kind of the point made about halfway through the article where he states if you expand a road to 100 lanes you won't ever see any traffic.

In those country towns with a population of 30,000 but a super road through the center that could carry 100,000 people a day the road _is_ the same as if that 100 lane highway had been built in a major city.

In the article they state that there are more factors at work but they do need a title for the article that will draw people to read it.

I think highways like the 401 in Toronto prove that the "100 lane" theory is pretty much bullshit. https://en.wikipedia.org/wiki/Ontario_Highway_401 It's arguably the busiest highway in North America based on a number of metrics, even more volume than LA's perpetually busy Santa Monica Freeway.

They've been expanding it ever since they've built it and in parts it is 25 lanes wide. It's slammed all the time. Traffic jams at 3am aren't unheard of if they're doing road work and it's slimmed down to half capacity.

It was once merely an 8 lane highway, so the idea that tripling or quadrupling the size of a highway magically makes traffic jams go away is so far from the truth.

There's a parallel highway running north of it that's paid by tolls which isn't as busy, but only because they can adjust pricing according to demand.

Counterargument here: http://pc.blogspot.com.au/2014/06/more-roads-doesnt-mean-mor...

"By comparison, building expensive transit systems aimed at getting people out of their less-expensive cars generates zero economic benefits if it generates no new travel. Only new travel generates economic benefits, so people who argue that new roads induce new travel are actually arguing that new roads create economic benefits.

If congestion is the issue, then–as Mann briefly mentions–the solution is congestion pricing. But Mann doesn’t understand the difference between true congestion pricing and New York City’s proposal for cordon pricing. Cordon pricing is more a way of raising revenue to fund urban boondoggles than a way of relieving congestion.

Even UC Berkeley planning professor Robert Cervero believes that the induced demand argument is “wrong headed.” “Road investments by themselves do not increase volumes,” he writes. “Only by conferring a benefit, like faster speeds, will traffic increase.” Provided that benefit is greater than the cost–something that could be assured, Cervero says, through proper pricing of roads–then it is a good thing."

Unused road space is staring everyone in the face and only a small percentage of commuters are using it: motorcyclists. If you just took 10% of the cars off the road and had them riding and lane splitting instead, congestion would be reduced exponentially. The good news for Californians is that lane splitting is actually legal unlike every other backwards ass state.

Dallas is building a congested demand pricing project for the I-635 corridor. They are calling it managed toll lanes. http://www.txdot.gov/business/partnerships/current-cda/635-l...

I-495 through the VA side of the DC suburbs uses High-Occupany Toll lanes - demand pricing on top of HOV lanes. If you have passengers, you can use them for free. If you're solo, you pay the toll. They use a switchable EZ Pass transponder to make it work without toll booths at each end.

Colloquially, they're referred to as "Lexus Lanes".

As far as a I can tell, they haven't done much to reduce congestion on the non-toll lanes. But, if you can afford $12/day (peak rate, round trip) for a 10 mile stretch of interstate, you get to avoid the dirty, unwashed masses who are stuck in the poor lanes. :)

Well, there is the same concept in LA, the 405 got "pool" lanes, usable for free when you got 2+ people in the car. The issue with this kind of lane is that it's the farthest left, and if your trip on the 405 is only a couple miles you gonna struggle most of your time to reach this lane and then leave it and reach the right lane before your exit when the road is jammed!

The concern when this got deployed in the Seattle area is that it would be poorer people who are most likely to use these. Example: somebody trying to get to the daycare by 6PM so they don't get charged by the minute for being late.

Why is this a concern? It sounds like a completely rational decision for someone of any income level to pay money for a faster trip to avoid an even great amount of money in fines...

Well, at least there'll be multiple toll lanes. Some cities have demand pricing (Atlanta, for one) where there's only one lane and it's separated from the main highway by solid painted lines or barriers. So if you are in the lane behind some slowpoke, you're unable to pass them.

I drove from Austin to Dallas the other weekend -- what a mess I-35 is. And it's been that way for 30+ years. High speed rail could help, if there's good transit connectivity at both ends, and the trains are frequent enough.

I live inside the Washington D.C. beltway in Alexandria, VA. Traffic here in town is slightly better than what I had gotten used to in Philadelphia, but that's likely because A) I never have that far to go and B) I never drive during rush hour anymore. In the District, it's a fucking Boschian nightmare. DC typically ranks 2nd or 3rd worst traffic in the country, duking it out with LA and Miami.

And all around me, as I look out my windows, all I see are skycranes build condo projects.

On my walk to my local coffee shop, I see a political campaign poster touting the candidate will work to block some proposed parking garage project that I never understood why old people hated.

There is a new metro station going in less than half a mile from my place. Except it won't be done for another 5 years or something. And really, the other nearest station is only a mile away over flat road. I don't understand putting metro stations less than half a mile from each other. That just seems more likely to cause traffic than it is to alleviate it, as it increases the latency of trips.

There aren't any major road projects going on. There are one or two projects to fix up a few ramps that were crumbling. There is something about an express lane all the way to the Dulles airport. But nothing about the 95 expressway seems to be addressed.

Clearly, someone is planning to pack a LOT more people into this area, but has given no thought whatsoever where to put their cars, either while in use or not.

So I've just accepted that this is reality now. There is absolutely no reason to complain about traffic, because nobody is doing anything meaningful about it. I try to work my life to not be involved with it at all. I do most of my shopping online. I am a freelancer who works 100% from home. I live in walking distance of a bar and a coffee shop, and I'm even cutting back how often I go. Unless I can finally convince my wife to move out of this place, I'm not traveling anywhere during daylight hours.

I see it as basically Malthusian: people will keep driving until it is so unpleasant that they don't want to do it any more. The tipping point for most people is around 45 minutes. Here's a book with a footnote that mentions that the same 45-minute rule applies worldwide and though out history: http://books.google.com/books?id=uuqIA-ce1CoC&pg=PA408&lpg=P...

Dolts like the article author completely miss the point. Yes, the number of people on the road will increase as we add lanes, but that doesn't mean that adding more lanes is useless. More people are able to get where they want to go, that's a good thing.

Building alternative high-volume modes of transportation is more effective than building larger roads.

Adding more lanes is useless from a urban planing point of view because it doesn't alleviate congestion.

    Finally, businesses that rely on roads will swoop into cities 
    with many of them, bringing trucking and shipments.

Indeed, roads are most often built to bring in new large corporations that create jobs. Depending on how one measures "quality of life" this could be considered a very good thing.

I was expecting something more interesting such as the correlation to car driver behaviour and inefficiency with stop-go traffic when not leaving a buffer in front of the next car (something that was written up very well a year or so ago) and how this logic is effected with wider roads. Instead, got some pretty worn out theories.

I'm not sure I'm a buyer of the congestion demand pricing. There's a very real cost already involved in traffic. Time lost in the car, quality of life, stress. Anybody who lives in the Bay Area, LA, NY, DC understand that timing your entire life around traffic is the regional pastime. So the argument then has to be an incremental one. If we then layer a more direct money tax of some kind on top of all the crap that already exists then that will be the straw that breaks the camels back and gets enough people off the road. Hmm.

That's a sunk cost though. When someone gets in their car to drive somewhere, they don't think "this trip is going to cost me $7.50" because they pay for their gas at a different time, they pay for their insurance at a different time, they pay for repairs at a different time, and vehicle depreciation is largely invisible. They don't equate time lost to money lost either.

However, if you have a big visual reminder that the road you're driving on will cost you money as soon as you drive on it, then people will be less likely to either take that route or drive altogether.

I hear you but I'm going to respectfully push back. When you live in a congested area you really do plan much of your professional and personal life around traffic. When to start work. When to leave work. When to schedule a meeting. When and where you'll go this weekend/vacation/kids. So this will make it more painful but it's hard to know how much this will be the final tipping point.

We're about to see a large-scale experiment of demand pricing in a major metropolitan area: Dallas, Texas is implementing this very thing as part of the LBJ Express project.[1] It is scheduled to be fully completed by the end of 2015, but portions of it are already open to traffic, with the demand pricing in place.

It will be interesting to see what happens. For the record, I can see the construction from my office window. I've used the express lanes a couple of times now, and traffic has always been clear on it when I did so. Whether this is due to pricing or simply because it is so new I am not sure. Either way, there was no traffic jams for me (during afternoon rush hour, mind) on the occasions I used them.

[1] http://www.lbjexpress.com/FAQs.asp

These guys are partially idiots. Obviously there is more demand than road if use is directly proportional to supply as they claim. The problem is the way we design cities and lay out the roads. A regular grid where we put STOP lights? seriously? Go, stop, go, stop. And then we put businesses right on the corners to increase congestion right where it's already bad. When cars get tested for EPA MPG ratings, the average speed on the city drive cycle is slightly under 20mph due to all the stop and go. Hybrid cars get better city mpg than highway by using regenerative braking to recover energy from all that stopping and making drag the dominant factor. I don't think traffic circles are relevant, but designing the layout to minimize disruptions to flow altogether. I've been contemplating this a lot, but can't write it all here right now. Imagine if every drive were 10-20 percent longer distance, but you could average twice the speed over the journey! You'd save gas and time. And if your journey takes less time, that's fewer car-minutes on the road which means less traffic and they could have fewer lanes. The problem is even if we could do the design correctly, nobody is going to build a city from scratch - except in China maybe.

Building a city from scratch and designing it primarily with private automobile traffic in mind is a textbook case of Doing It Wrong. The only sane approach would be to build a subway (or maybe streetcars that have right-of-way) and then stick a city on top of it.

It seems like you're assuming that the only priority is moving cars around - not walkability or livability. In other words it seems you're assuming the only purpose of the built environment is to move cars. I argue that it isn't and in that case we should measure other stuff besides moving cars efficiently when building our cities.

No, I advocate efficiency as a goal weather it be cars, public transit, or something else. Personally I'd like easy access to both. But public transit is only viable with high density and I don't like the way we do dense either.

I envision traffic as a viscous fluid. Unlike fluids that we are used to, like flowing water, maple syrup, or honey, this fluid wants to flow in multiple directions at once. If you can imagine a stream that flows both uphill and downhill at the same time, and sometimes acts like thick molasses, other times like corn starch suspension, and other times like mercury, you are ready to think about traffic.

This bizarre fluid does obey certain rules, which are consistent and discoverable. Once you have all the rules, you can model an entire system and run experiments.

That said, I hate roundabouts. They seem to combine the worst elements of multiple solutions to make all drivers equally unhappy and inconvenienced. It is somehow okay that no one wins, because everyone loses.

They're also an eyesore.

Not to worry, over the next several decades large cities will build out effective mass transit systems. High-speed rail will eventually come to America after years of debate.

Look to China to lead the way. What's this low-speed maglev that they're building now?

http://usa.chinadaily.com.cn/epaper/2014-02/19/content_17291...

over the next several decades large cities will build out effective mass transit systems

Which, as the article points out, won't fix the traffic congestion, because as some people switch from driving to mass transit, other people will take their place by driving when they previously walked or just didn't make the trip at all. So mass transit can increase the total number of trips made, but it won't decrease road usage.

People like me, who use mass transit, don't care about traffic congestion. Queue all you want. I just want more mass transit because travel by mass transit happens in constant time regardless of the amount of travelers (within reasonable bounds. Carriages do get full eventually).

The other benefit is being able to do something useful on the commute too. I can usually get through all my morning's emails on the train into work.

This is much better than sitting behind a wheel for 1hr+ a day doing nothing.

http://www.theonion.com/articles/report-98-percent-of-us-com...

It is the onion, and is flippant, but as a general belief it touches upon a basic truth -- everyone would love if everyone else used public transit, leaving the roads available for our use.

This is occasionally used as an argument in favor of cheaper or free public transit: that even from a self-interest perspective, it makes sense for road users to subsidize public transit, as a way of effectively paying other people to get off the roads and out of their way. However most road users don't seem to see it that way.

As with most articles from the Onion, there is an element of truth, but if you look at a city with good public transit options (NYC is the only one in the US) over 50% of commuters travel by public transit. If you look at cities with adequate but not great public transit options (DC, Boston or San Francisco) more than 30% of commuters travel by public transit.

People drive usually because it's either faster and/or cheaper. Right? I'd rather not intentionally make it expensive to drive so solve for X... Faster and cheaper... Everyone wants to save money and get home sooner.

People drive usually because it's either faster and/or cheaper. Right?

Or because it gives more freedom--you're not restricted to particular routes or particular times.

Mass transit can beat driving on all these metrics (speed, price, and freedom of travel) when the population density is high enough that it's cost-effective for mass transit to have dense coverage, so that you're never more than a short walk from a train station or a bus stop. But many people (including me) don't want to live in a place with that high a population density. That's ultimately why mass transit hasn't caught on in the US the way it has in other countries.

I think people are basically more or less the same everywhere in that respect. Americans don't like public transit because it really sucks in America. In places like NYC where it doesn't they don't. But not every city with great transit has to be a stressful megalopolis like NYC - Vienna or Zurich for example are really nice and quiet (and clean and have low crime) despite having great transit.

It's kind of a chicken and eggs sort of thing but I'm pretty sure if you've suddenly by some miracle have really awesome mass transit in LA or Houston like you do in NYC, Tokyo & lots of European cities (even medium sized ones like Prague or Copenhagen) I'll bet loads of those freedom and sparsity loving folk will start using it and over time population and living patterns will adapt to the new conditions.

Americans don't like public transit because it really sucks in America.

Not everywhere. When I was in school in Boston, I went everywhere by public transit. I never missed having a car (I rented one the few times I wanted to travel somewhere like up to New Hampshire to see foliage). However, I also learned from that experience that I don't want to live in a place like that for the rest of my life.

not every city with great transit has to be a stressful megalopolis like NYC - Vienna or Zurich for example are really nice and quiet (and clean and have low crime) despite having great transit

I didn't find Boston to be a "stressful megalopolis" (I haven't lived in NYC so I can't speak to that). I just realized I didn't want to live permanently in a place that dense. I've also visited London and Paris, both of which have excellent mass transit (at least by US standards), and while I enjoyed visiting them, I don't think I would want to live permanently in those cities either. Maybe I'll get a chance sometime to visit one of the smaller European cities you mention to see what they're like.

Well Paris & London are the largest cities in Europe (together with Istanbul and Moscow, if you consider those Europe) so if you aren't after the "big city experience" it's not really surprising you wouldn't want to live there :)

I wouldn't either btw - and I live in Berlin which is also a big city (but still 1/3 of London and less than 1/2 of Paris in terms of metropolitan population - and still double or more the medium sized cities I mentioned).

Vienna/Zurich/Prague/Copenhagen/etc are more like living in Portland than in London or Paris (or NYC for that matter).

Vienna/Zurich/Prague/Copenhagen/etc are more like living in Portland than in London or Paris

Hmm, interesting comparison. Portland is smaller than Boston, I believe. I have visited it (briefly), but never stayed long enough to really get a feel for the city.

You don't need to live in a city at all. I grew up in a village in Sweden, population 4000. The village itself was walkable/bikeable, and there was a bus to the nearest city (pop 90K) until 3 AM every night, departing every 20 minutes during rush hour.

I thought that it went without saying that we were discussing mass transit in densely populated areas. If you don't live in such a place, you probably don't have 9 lanes of road either.

As for why it hasn't caught on, it's because it's incomplete and inconvenient for most people. Even where it works, it's still not great. Ever take a slow bus across town in Manhattan? It's faster to walk.

I thought that it went without saying that we were discussing mass transit in densely populated areas. If you don't live in such a place, you probably don't have 9 lanes of road either.

The 9-lane roads in metropolitan areas aren't confined to the densely populated areas; in fact, many of them are there only because so many people commute from the less densely populated suburbs to the densely populated inner city to work, instead of living in the inner city where they could easily take mass transit.

I wish, LA had a good public transportation network but it doesn't. I used to take the bus, I had to change twice and spend at least 1h traveling, while it takes me 12 to 20min using my car (20min on exceptionally busy days). It probably is one of biggest city that doesn't have an underground metro! I guess that's one of the rare thing I miss from NYC. Being able to move fast without owning a car!

The article fails to follow the logical process to the very end. What isn't being asked is: why are humans using that road when they weren't using the smaller road before?

Humans don't just drive for the sake of doing so. They drive because they have the need to do so. Building roads lowers the perceived time cost of driving, so more people drive, and more people drive at the time that is most desirable to do so. In some cases this results in increased congestion, but if the need of the population is met or exceeded, then the result is smooth traffic flow. Even if the bigger road is more congested, it may still have a greater net traffic flow, and that increased flow results in greater economic and societal good.

For example, because of new highway XYZ, I can now work somewhere that was formerly too far a commute.

The problem isn't solved by building smaller roads, it's solved by building even bigger roads, or building roads in a more optimal fashion (selecting a route for a new road that matches the transportation needs of the population).

This is exactly the sort of thing that I don't want a politician to read.

A study finds a correlation between construction of new roads and increase in driver miles. A policy-maker, never one for logic, transmutes correlation to causation. Now, it may be that they assume that increased demand for travel causes road construction, and assign resources to reduce the total cost for each trip. Good for them.

But they may also assume that more roads cause more travel, and decide to prevent trips by destroying roads. Boo. You could as easily prevent water from flowing downstream by building a dam. Then your stream becomes a reservoir, the downstream flow rate reduces while it fills, then the reservoir spills over the dam at the same rate it flowed before.

People want to move around. They want to travel freely between the places they can earn money, the places they can spend it, and the places they enjoy keeping their stuff. Making the travel more difficult makes them unhappy, while making it easier makes them more content. For the most part, the desire to move minus the cost of the trip results in economic activity. If that number is zero or negative, people stay put, and don't add to economic circulation.

Human movement is an economic engine. You can add friction to it and generate more heat, or you can remove friction and extract work instead

Generally speaking, you want people to take more trips, and you want the cost per trip to be as low as possible. That doesn't necessarily mean building more, wider, and longer roads. You may be able to get better results by siting the popular destinations more closely to each other, and ensuring that the routes between them are direct, non-viscous, and non-turbulent.

When you do something stupid like make 3000 people work in one building complex, far away from homes or other businesses, and have everyone sleep in a suburban bedroom community, far away from jobs and shopping, yes, the road that is the best available route between them will be packed every single day. You cannot economically reduce traffic by narrowing that road. The cars you see on it will go down, but that is because people quit their jobs or moved their households over the arduous commute. They went somewhere else, and will be earning and spending there, instead. It would be better to simply put the homes, businesses, and offices closer together, and diffuse the obvious trips onto roads in such a way that they are neither over or underutilized.

> But they may also assume that more roads cause more travel, and decide to prevent trips by destroying roads.

Politicians in the city I live in are literally doing this, installing "traffic calming measures" (destroying perfectly good road lanes that were just resurfaced)

I believe phase two involves generating clean energy by harnessing the power of Bastiat spinning in his grave.

I witnessed a nice microcosm of this demonstrating how fast the shift occurs.

Every day I cycle commute up the London A24/A3 (CS7) corridor. On the first day of the first tube strikes early this year, the congestion was insane. I mean the worst traffic I've ever seen in my life. An ambulance was literally stuck in traffic and no one within earshot of the ambulance able to move their vehicle enough for it to get through.

But then the second day of the strike traffic was a bit on the heavy side but mostly back to normal. I think a huge number of people realized that it was actually pointless to even attempt to go to work unless they were walking or biking. This effect even carried over to the second tube strike a month or two later.

The actual paper http://www.ademloos.be/sites/default/files/meccano_docs/The_...

Sounds a bit like Parkinson's Law: http://en.wikipedia.org/wiki/Parkinson's_law

"The demand upon a resource tends to expand to match the supply of the resource."

This article seems like a gross oversimplification. There are a lot more variables at play here than just what people tend to do with extra capacity. Terrain and flow dynamics, as well as driver behavior, have a lot to do with it.

Take 405 in West Los Angeles for example. On one side it has the Sepulvida pass with very steep decent going south and ascent going north. It is crossed by I-10 going to Santa Monica and Downtown, a little further in 90 goes to Marina del Rey, and then exits for LAX.

If you examine this section as an example, you have number of people either trying to get in or pass through this section during commuting hours. From the north, the sepulvida pass makes people slow down as they climb the hill. Not exactly sure why, but many drivers seem to be afraid of it. This recoils through the rest of the traffic, as each incremental car slowing down results in the car behind it slowing down slightly more.

Once you get down the hill, you have people exiting and entering with very short distances to decelerate or accelerate, causing more slow downs.

On the other side, you have a similar dynamic, but people are also trying to merge to or from 90 and I-10. With majority of the traffic heading either to Santa Monica, West LA, Century City or Hollywood/Downtown/something in between.

The city has recently completed a major expansion on the 405 in that area adding a carpool lane, but more importantly changing how you merge to/from several of the exits, including Wilshire blvd and I-10.

In my observation, the carpool lane did relatively little, but the smoothed exiting/entering with enough time to accelerate shifted the bottleneck from before Wilshire blvd, frequently spilling past the 90, to no ending around Wilshire.

The flow restriction now seems to be the dreaded 101 interchange, with a huge amount of traffic stuck in the Sepulvida pass. As the result, I can not get from the 90 to I-10 in about 10 mins, whereas before it was easily 30.

Point being, that if you maintain road size (or marginally improve it) and optimize the flow, you get a lot better returns.

Honestly, it would have been hard to make the Wilshire exit/entrance any pathologically worse. It was one of those ramps where the entering cars have to merge with the exiting cars and everyone was entitled and freaked out about it. (I used to live about 4 blocks from it.) It was awful, so I'm glad they've finally fixed it.

Well, sure, the article doesn't fully describe the intricacies of the situation... but neither do your anecdotes.

But nevertheless, the study mentioned in the article does show that building more roads does not make traffic better:

> "[the researchers compared] the amount of new roads and highways built in different U.S. cities between 1980 and 2000, and the total number of miles driven in those cities over the same period. “We found that there’s this perfect one-to-one relationship,” said Turner. If a city had increased its road capacity by 10 percent between 1980 and 1990, then the amount of driving in that city went up by 10 percent. If the amount of roads in the same city then went up by 11 percent between 1990 and 2000, the total number of miles driven also went up by 11 percent. It’s like the two figures were moving in perfect lockstep, changing at the same exact rate."

I'm pretty sure you missed the entire point of my comment. Basically with highly concentrated areas of work for people living in the sprawl that is greater LA, the bigger gains are not in bigger roads, but in facilitating better flows in existing corridors. LA freeways have a lot of room for optimization, the big problem is always cost and lag in implementation. The above "anecdote" took 2+ years to be built.

If you have never experienced 405 in West LA, I suggest you visit :)

apparently, correlation is equal to causation[1]

[1]entire article

I don't think that applies. Isn't the point (or one of the main points) of building more roads to make traffic congestion better? The study in this article shows that that doesn't pan out.

Correlation may not imply causation, but non-correlation is pretty strong evidence for non-causation.

What about a smartphone app that helped you avoid the most congested roads. I think Waze does something like this today and I'm sure that Google Maps can as well. We just need to have it integrated into the system so that we don't have to think about it.

I think this is where projects like Andriod Auto & iOS CarPlay come into play. Just jump into your car and say "Take me to Julia's house" and the route will be mapped out for you. I would love to not have to figure out the best way to get to the mall on a busy Saturday afternoon! Even better, an auto-driving car will actually take care of the moving bits as well as planning your route.

What a bunch phooey. Just more peddling of flawed studies to promote increased taxation.

If you want to reduce congestion try the obvious. 1) Mixed zoning, allowing more residential and retail to be close together. So people have less distance to travel. 2) Raise speed limits where possible so people can get to their destinations faster. 3) Build for the future. Don't just add a lane or two, add six. 4) And yes trains, do help. The problem in the States is that we hardly have any. Only the major metro areas have any sort of adequate rail system and even they are fairly slow, limited and antiquated. Which brings me to my last point: 5) Consult with Elon Musk.

I'm sorry but you are simply way off base. When a road is expanded, it does temporarily shorten commute times and reduce traffic. But that makes the cheap land farther away from everything more desirable, causing people to move out further, increasing traffic. Those new suburbs require a massive infrastructure investment to install water, sewer, gas, electricity, etc. Not to mention building surface streets, schools, libraries, shopping centers, etc.

We've been building roads at a massive clip around the Dallas area, sprawling out over huge suburbs, many of which support mixed development (as does the city of Dallas itself). When I first moved here, the previous two lane highway had just been turned into a six lane toll road and there were no traffic jams, not even during rush hour. It's been 8 years and now we have stop and go traffic every day for about a 30 minute window and that window continues to grow larger.

If you think about the design of roads it should become clear why this happens... You have many on-ramps dumping cars into a fixed set of lanes. If, at any point, there are more cars that want to enter than there is room traffic must slow to accommodate them. In reality it happens before the road is full as people leave space for braking, etc. When that happens at any point, it creates a wave of slowed traffic that begins to ripple backwards from that point, meaning anywhere further backward that is close to capacity begins to jam up, making it even more difficult for new vehicles to enter the roadway, amplifying the wave even more.

Even if demand were 100% evenly spread out (which it never, never will be simply by the fact that brand new areas have empty fields or wooded lots - it takes time to build offices, start businesses, etc), the friction of merging & exiting the highway would eventually require more cars to exit than are entering, the exact opposite of rush hour conditions.

Building more lanes merely delays the inevitable; that's assuming you can, which in Dallas is impossible - most of the current roads (and 100% of the worst congested ones) would require tearing down skyscrapers or 100% tunneling / elevated highways because there is no more space.

Can you please explain how taxation plays into this? Are you implying that decreased investment in roads coupled with increased investment in public transit would increase taxation? What's your argument?

I think studies which measure traffic congestion in the US based on road building over the last couple decades IMHO are flawed.

That is because I believe how and where you build the road is more important than how much road you build. For example, where I live the main roads in the city proper haven't been upgraded in the last 40 years. Yet, new roads funneling drivers from farther and farther out are being constructed all the time. This simply means that the first 1/2 of many drivers commutes involves sailing along a new road until they hit the old roads, and then puttering around at 5mph (or worse) for the last 5-10 miles of the journey.

>> If we gave drivers some extra incentive to avoid the most congested hours, we could better utilize the roads’ capacities.

There is already enough incentive, the incentive is to avoid seeing your hours waste away on traffic. Most people don't really get to choose their hours, with or without incentive, otherwise they already avoid the most congested hours. So the suggested incentive doesn't really make a difference when it's not a matter of choice, except by taking money away from people without adding any value.

So we need to incentivise the people who set the hours.

Weird title, seeing as how the article states that bigger roads do not make traffic worse, it stays the same:

>> New roads will create new drivers, resulting in the intensity of traffic staying the same

Good point. And it's still not completely clear that the roads are keeping up with the growth vs pulling it in. There were a couple of good points around larger interstate type services picking your town for HQ if you have extra capacity but in general it still seems too abstract.

Car traffic is harmful, so having more of it is bad.

I wish there were a starter packet for using the metro. I hate it when people say, "where do you want to go? Oh, that's simple, you just take the Orange line to Such-and-such station. Why couldn't you figure that out? It's so simple." No, that is not simple. That's an insider's language designed to call out the outsiders. I get there and there isn't just an orange line. The orange and green lines overlap. Don't get on the wrong train at the right terminal! Or maybe it's like the purple line, where you have to take the orange line to get to it, and everyone else already knows this, but they don't tell you that, they just say "take the purple line" and I'm left standing there freaking out that I can't see a purple dot on a sign in a poorly-lit cavern.

If I drive somewhere, all I need to know is where it is and where I am and I can eventually get there within 50% of the normal time to get there. Let's take this road that points towards where I want to go. That didn't work? Okay, let's take the next road that points to where I want to go. Unless you're driving around the Pentagon, which I'm fairly certain has had its road systems designed into the shape of a PENTAGRAM, then that will get you to where you want to go in the majority of the US.

And grand total is still faster than walking to the metro, waiting in line for the train to show up, and waiting for it to stop at every spot in between.

Google maps / transit routing smartphone apps do a pretty good job of this. I've not had trouble coming into a new city, asking Google maps to give me mass transit directions and just following them.

Most cities have real time bus/train trackers now that help you plan when to leave so you don't have to wait long.

Yes, Google Maps is how I manage it.

>I wish there were a starter packet for using the metro.

A map?

What else does a starter packet need to contain?

When you go to sign up for your driving permit, they hand you an entire book on road signs, markings, and traffic signals. They don't just explain what is around your neighborhood and on your daily commute, they explain how the entire system works and how to read it fluently.

Later, they test you on it. Obviously, testing people on riding the metro would be ridiculous. But there is a huge difference in having a map and signing system in place versus knowing how to read them.

Just ask somebody. If the first person can't help you, ask someone else. Most people you'll see taking public transportation have taken it before.

Is that any different to understanding road signs/given road directions?

Yes, I took a semester-long class for (among other things) understanding road signs in highschool, and my parents took several hours of their own time to coach me as well. Learning to drive and understand the road system is not a trivial task.

Yet for the metro, if you don't know what you're doing and you don't know anyone who knows what they are doing, you're essentially dumped in the deep end during all-swim. The maps and the keys for understanding the maps are all printed in high-traffic areas (of course, where they will be seen by the most people), and the other passengers are often quite annoyed by "noobs" who stand in those high traffic areas and gawk at the maps.

This article is rubbish

> What’s Up With That: Building Bigger Roads Actually Makes Traffic Worse

> For instance, Paris in recent decades has had a persistent policy to dramatically downsize and reduce roadways. “Driving in Paris was bad before,” said Duranton. “It’s just as bad, but it’s not much worse.”

Bigger roads are worse, and smaller roads are worse?

There's a point here somewhere but they've failed to make it.

I guess they get more foot traffic, and living streets (more shopping/social life etc). If you get that and less cars in traffic I guess I see a win there.

Then clearly we can make traffic flow more smoothly by shrinking existing roads. Right?

What? No? It doesn't work that way? Hmm, what a coincidence, it appears we're already at the sweet spot on the traffic-flow Laffer curve.

Either that, or what the authors of the study are trying to optimize is very different from what the typical driver is trying to optimize.

I'm a huge fan of toll roads. And I really like charging for use at certain times.

Perhaps you could reduce rates for each passenger in the vehicle. Maybe a four-passenger vehicle gets to drive free anytime. This encourages car pooling for those who can't afford to pay the tolls.

In fact it may create new jobs.

Well yes, people are irrational.

I personally think that a lot of jobs today don't justify driving around at all. What difference does it make if you code in the crowded building of your employer or at home and use tele-conferences to talk with each other? Maybe you would even be more productive if you got rid of all the stress caused by commuting around a substantial portion of the day.

I guess a good chunk of the office work done today could be replaced by remote work, but employers insist on everyone to get into the car, commute and thereby congest roads and burn up non-renewable petrol.

I am seriously wondering why the working population don't bill the time and expenses of commuting to the employers? Most entrepreneurs do. You pay $50-$100 for the plumber to come to your place (just to come there, work is billed extra). Why not office workers?

This is nothing new. Read Robert Caro's 1974 Pulitzer Prize-winning book THE POWER BROKER if you wanna know about roads and highways and freeways and how expanding them only makes traffic worse.

I've always thought something is definitely wrong with the way we think about roads when car pool lanes end up just as busy/gridlocked as the rest of the road.

I feel like there's a cause and effect relationship being presented here that isn't necessarily true, even if the correlation is true.

It’s the roads themselves that cause traffic. The concept is called induced demand, which is economist-speak for when increasing the supply of something (like roads) makes people want that thing even more.

So contrary to popular belief, invention is actually the mother of necessity. The question now is how best to take a systems approach to promoting alternative forms of transportation.

"The question now is how best to take a systems approach to promoting alternative forms of transportation."

If the principle, holds, obviously: build more.

I think that it might hold, at least for me. The biggest obstacles to my use of public transportation is the first mile and last mile, and the regularity of trips.

Using Google Maps as my source, a trip home will take between 1hr, 3min to 1hr, 30min, and still includes driving for the last mile. Granted, this is San Antonio, which is certainly a driving city. The same drive home is 14-17 minutes, depending on route.

In my calculus, drive vs. public trans. is a balance of which sucks least. Right now, I would say that the cost and altruism aspect of public trans. would allow it 200% of the suckiness allotted to driving. All else being equal, I would need a public trans. trip of ~30minutes or less.

Yes, I’ve been stuck in traffic on 405, but it’s not like you never get through.

Also: bigger/faster computers makes programs bloated/slower.

This actually makes sense in an unintuitive way.

well duh

Reply on news.ycombinator.com