Cars and Robust Cities Are Fundamentally Incompatible
theatlanticcities.com
theatlanticcities.com
Shopping is the main one. Cars allow shops to be very large (providing more variety and creating economies of scale), and the ability to site shops in cheaper locations contributes to lower purchase prices.
The other classic problem with public transport is that it cannot replace the point-to-point nature of cars. Sometimes that matters, sometimes it doesn't. But it matters for enough people that cars are the dominant mode of transport.
Self-driving cars will help some of this, particularly the car daycare problem. Though they will have problems of their own. Ever tried to get a taxi at rush hour? Ever been on a bus with suspiciously damp seats? People will still want their own cars and will still want to travel at the same times.
As an aside, one thing I really like about my city (Perth) is that there are a number of free bus routes in the CBD. You really do find CEOs riding the bus to get around.
I am especially bullish on that delivery service. Sure, they use a car, but one small truck making the rounds seems a lot more efficient than lots of small cars driving to the store.
Shopping is a very trite activity, owning a car doesn't make it much better. In fact driving out to some shopping center seems to make shopping a day trip, which is one wasted day. I prefer to just pick up a few things on may way home from the tube, and ordering more complicated stuff from Amazon (plus the weekly food deliveries).
What they do have is lots is support for shopping and public transport including wrapping packages so they are easier to carry (buying a printer they wrap a secure handle around the package to make it easy to carry home) and most stores offer relatively cheap and convenient delivery.
Let me put it in networking terms.
Cars provide a circuit model. You go from point to point over a single dedicated "channel".
Any kind and combination of public transport does not do that. It is a packet model and you, the human, are the packet. You must change modes of transport and hope that they sync up, aren't running early or late, fit your schedule and so on.
Except in the cases where point-to-point travel takes substantially longer (very dense urban cores), I think most people will always prefer cars.
(Yes, I know the analogy is imperfect because the car can be seen as a packet blah blah blah. But cars are not sentient. Humans are and it is humans who decide what mode of transport to use.)
As a society we seem to have mostly established the abstraction that energy is free and unlimited. From the recent explosion of amount of things humanity does we can see clearly how practically free energy is a desired state. However, we're not there yet. Until we really get there, forgetting that this is only a leaky abstraction is dangerous, as the leaks have deadly consequences.
None of that matters. Who decides which transport to use? Humans. What measurement do they use? It's clearly not litres of diesel per passenger-kilometre.
"Humanity" is pretty good at economising scarce resources. But people are pretty clearly prepared to pay enormous costs (fuel, parking space, commute time) to have a private car. Arguing with revealed preferences doesn't change them.
The summary is that the commutes are brutal, but the private cars get to work and get parked, off the road, and they're not a big deal. But cabs are on the road and take up space all the time. The 2,000 additional cabs will cause traffic to move 12% slower, including deliveries, buses, and commutes.
People commuting in/out of the city is not as bad as just getting around the city.
http://www.npr.org/blogs/money/2012/07/31/157477611/does-new...
http://cityroom.blogs.nytimes.com/2008/04/07/congestion-pric...
I live in Boston, and I'm fairly certain that the local Porsche dealership is catering to potential customers with very early morning test drives on Storrow Drive with demo cars. During the summer months I almost lose count of the number of 911s I encounter at 2AM (with a driver and a passenger, almost always both male) with window stickers on them, carving through the elevation changes and off-camber corners well beyond the speed limit. Last year it was so frequent I just got over to the rightmost lane whenever I would hear a wailing flat-six approaching.
The article tacitly admits this, with the lowest rate of automobile commuting in their data set still over 40%. Which makes the terribly overstated title even more puzzling.
If there's more support for cars, then those services get pushed outside of the city, or the city expands in area (if possible). Both trigger a higher need for cars. Venice can't expand, automobile traffic is mostly impossible, and also, the population is decreasing, especially in the historic old city, so likely not a good reference.
My observation is that cars "swarm" to the carrying capacity of the city, so that every city will have cars unless they are prohibited either by law or lack of drivable roads. After all, cars are really convenient, so people who can afford cars will use them until it's more convenient to use other means, like walking, biking, mass transit or (in a few places) animal-drawn carts. Therefore, seeing a lot of cars in a city isn't itself meaningful.
The question then isn't one of swarming, but of the response to the swarming. What are the effects of making it easier to commute by car? That's what this study appears to answer.
The large cities this is true for, that I have some personal experience with, are London and Amsterdam.
The difference is that they are also swarming with public transport and/or bikes.
The following is not an uncommon sight in your typical Dutch city: http://www.youtube.com/watch?v=n-AbPav5E5M
And in London using the tube / subway is just the fastest way from any A to B anywhere in and around the center (unless you can't walk half a mile), and the most convenient and care-free one (no finding a parking spot, no road taxes, no insurance, no fueling it up, no having it repaired if it breaks down, no winter tires, no ice scraping, no cleaning it,...).
I counted about 500 boats on the Grand Canal using Google's map imagery, either in motion or moored. (From the train station to Piazza San Marco.) The distance is ~3km. How do you map that to car traffic? Boats carry more than cars and a vaporetti caries more than a bus. After all, the boats carry a good fraction of the 18 million tourists per year.
Boats also need more space and go at different speeds. Did you factor in wake pollution? Quoting the quite work from WPI (http://www.wpi.edu/Pubs/E-project/Available/E-project-122107... ) "it is the only place in the world whose method of transportation has the literal ability to destroy the city." It also describes the effects of noise pollution, and the over 100 years of various attempts at traffic control.
Given all of these differences, how did you base your comparison of the congestion of cars in a normal city vs. that of boats in Venice?
Anybody who has actually been to Venice can clearly see the difference. I mean, they have ambulance boats and deliver boats and garbage boats, but boats are not a problem. Most of the boats you will ever see are the bus boats, or gondolas.
Did you read any of the links I posted? The first says "There has been an extraordinary high number of boat accidents, people ending up in the water and general confusion since many are not used to the complex canal traffic code where imaginary lines mark lanes that you are not supposed to cross."
Boats are definitely a problem. That WPI report I posted describes many of the problems with boats in Venice, including how they have "the literal ability to destroy the city" through wake damage, with pictures of what some of that that damage looks like and damage. Yet only 3% of boat traffic follow the posted speed limits.
(And I know that that report doesn't go into the full details. Wake damage also increases the amount of rotting in the buildings, because submerged wood and dry wood rot at a slower rate than wood which is both wet and exposed to air, which the wake in Venice exacerbates.)
My question to you was, how do you compare car traffic to boat traffic? Your response seems to be that you base it on the number of deaths or dangerous accidents. I thought surely that "destroy the city" quote would help you understand that you have an incomplete understanding of the issues.
BTW, if deaths and dangerous accidents were the key factors to "clogged" traffic, then NYC's traffic congestion could be solved with a 5mph speed limit.
You wrote "Most of the boats you will ever see are the bus boats, or gondolas". The traffic counts from that WPI report says that gondolas "represent a tiny portion of overall boat traffic in Venice" - except around the tourist area of La Piazza San Marco. It gives a breakdown. Gondole "account for 4% of the traffic in Venice" and "turismo boats account for only 40% of the total traffic." 44%+4% < 50%.
The majority of boat traffic, though just barely a majority, were service, merchant, and sport boats. (I checked. The vaporetti are classified as turismo boats for purposes of this analysis.)
So the reason that vaporetti and gondole are most of what you saw would be because you were a tourist, in the tourist areas, where the boat traffic more caters to you and people like you.
Doesn't all this suggest that you don't understand the traffic issues in Venice enough to be able to make a comparison with "swarming" car traffic?
http://www.canalvoyagers.com/2012/does-birmingham-have-more-...
I picked an arbitrary American city of similar population that I'm familiar with to compare. Harrisonburg, VA has over 300 "lane miles" of streets:
http://www.harrisonburgva.gov/sites/default/files/CMO/files/...
I'm not sure if a two-way road counts as one or two lanes for that count, but suffice to say that, despite having fewer people, Harrisonburg has roughly an order of magnitude more streets than Venice has canals.
Venice's canals aren't appreciably more crowded from what I remember, which supports my "order of magnitude" claim.
Alas, I couldn't find any statistics on the total number of boats in Venice, which really would be the best measurement here. But canal-miles seems like a decent proxy.
I pulled up a section of Harrisonburg's Main street, at some arbitrary time during the day:
http://maps.google.com/maps?saddr=S+Main+St&daddr=N+Main...
I counted 54 cars in use in that 1.5 miles/2.4km stretch.
The main road through town is US-19. In the section where it meets S-Main to N-Main I counted another 31 cars, though I think a few of them are parked. Almost all of these are private vehicles, and I presume with 1-2 people in the car.
That's about 85 cars per 1.5 miles, and probably around 100 people using the roads.
In Venice, I started with Ponte Scalzi, by the train station. There were 45 boats in use by the Ponte de Rialto and about 81 boats in total by the time I got to the end of 1.5 miles (by Ai Leoncini). This puts the traffic density at about the same. Boats are both bigger, slower, and don't have brakes, so this isn't a great comparison, but we're using your metric.
Of these 81 boats, at least 10 were vaporetti. There's easily 4 times as many people on the Grand Canal as there are on Main St. and US-19 combined. If they each used a 1-2 person boat then the boat traffic and corresponding need to dock a boat would be that much higher.
Take a look at Harrisonburg's downtown. See how much of it is dedicated to parking? I reckon it's over 1/2 of the city center! Now, look at Venice. There's places for boats to park, but nowhere near as much space dedicated for that. That's because many people in Venice walk to work or ride the vaporetti.
I contend that the boat traffic in Venice is high. The damage from ship wakes and propellers is well known. The problem is that it's so convenient to be able to go more than 5kt, and hard to say no to the 2 million cruise ship tourists who come in ships whose wake and pollution affect the lagoon and air, when the damage is slow to become apparent.
Fewer. Fewer roads, fewer cars.
But it really works strongly the other way around. The closer you are to others, the less you have to drive to see them.
Things like zoning and parking minimums increase the space between buildings (and therefore people), which necessitates more driving.
Very few cities have a good "park and ride" culture where you get safe, modern parking structures on the main approaches to the city, and then frequent express trains or buses into the center. Without this, it's more tempting to just stay in the car and put up with a little extra traffic and expense in the city center.
(In London, the closest thing you get to this is parking at Westfield in Shepherd's Bush and tubing it in, but Shepherd's Bush is too central to London to avoid adding to the worst of the traffic.)
Here in Milwaukee, for example, the downtown area has essentially been choked off - there are office buildings, there are some hotels, there are a few eateries and pubs which specifically cater to the office workers and business travelers (many of them aren't even open on the weekends). And there is. . . parking. Lots and lots and lots and lots of parking. If you ever go down there at night, the parking is really the most noticeable thing because of the sense of void it creates. All these great huge swaths of the very core of a major urban center that are just completely empty. The feeling is downright oppressive; it's like being in a shopping mall where 30% of the storefronts are vacant.
See, I was lumping myself in with that group. Park and ride schemes should, ideally, be attractive to commuters in the outer suburbs (many of the smaller British cities do a good job with this - Oxford, especially - just not any of the big ones I can think of).
1) They underestimate problems with their ideas, and then brush them off with solutions which, when summed together, make the visionary idea much less palatable. For example, the idea that we can abandon traffic lights and stop signs is complicated by the fact that not everybody can, will, or even should use a car to get around. But in the world without traffic lights and stop signs, pedestrians and cyclists are trapped by fear. Sure, they might technically be able to walk out into traffic and the wondercars will stop for them, but in reality, nobody will do that, because the cost of being wrong is death. In every conversation I have had with these technophiles, the solution always ends up being some massively costly network of pedestrian bridges at every intersection of the city. Yeah, not so palatable anymore.
2) They fail to understand that, like evolution, intermediate states have to be viable improvements on their own merits. Wheeled animals do not exist, not because wheeled animals wouldn't have superior mobility, but because no possible intermediate evolutionary state has incremental benefit over the previous legged state. How do you transition a 2 lane road into three lanes when only 50% of the cars are self-driving? What about when 99.9% are self-driving? There are thousands of imaginary future states that do not have credible intermediate states, and will likely never see the future state as a result.
With computer in charge it will be even more safer.
There are a few reasons for why it can work. One is the idea that drivers on a well delimited road become complacent. If it's a road designed for cars, then drivers go faster, believe they have priority, and pay less attention to non-drivers.
You can't just get rid of signs and expect things to work, because signs are the only signifier. For example, the curb is a physical indicator of the boundary between the place for cars and the place for non-cars. Also, if there are more non-drivers often near the streets (cyclists, walkers, people sitting at a cafe table) then drivers will be more attuned to events which are currently rather low.
It also depends on a relatively slow speed. That video for Saigon shows that the traffic is about that of a jog. At that speed, it's easy for drivers to slow down, and easily within the normal reaction time for humans.
If you go this route then you can't have broad, open, clear avenues since that encourages drivers to go a lot faster, which makes things more dangerous.
It broke off from driverless cars a while back, I think. I personally can't wrap my head around the idea that a personal chauffeur would get rid of city congestion. Who wants to wait more than 5 minutes for their car to come? That sets a maximum distance to wherever the car depot might be.
Technically they are apparently not wheels because the whole animal/plant rotates, but I'd argue that nature worked out the wheel principle just fine. In fact it seems an incredible cool solution to the problem that there aren't roads everywhere: the whole animal can transform from wheel form to walking form and back. It's almost like a biological Transformer (TM).
Even today, cities introduce pedestrian areas or limit the number of cars to certain areas, or they have special bus and taxi lanes. It seems likely that they would introduce special "only robot cars" areas eventually, as a first step.
The transition from traditional cars to driverless cars will be smooth, insofar as the cars are intermingling. But the supposed additional benefits that idealists like to portray as inevitable are nearly impossible to intermingle. The only way that they can appear is through a forceful transition...eg banning non-driverless cars.
It also seems to me that the transition to driverless cars will probably sneak up on us: "normal" cars will become more and more intelligent until suddenly they can drive without human input.
Think of a electrical utility truck. They need to do way more than just follow the rules of the road...and in some cases they need to break the rules of the road.
Many people will still want their own car for comfort and status.
Most people will still travel to and from work during the traditional rush hours.
> We analyzed a dozen historically dense, small cities from around the country in which the shares of residents getting to work by automobile range from 43 to 91 percent. We compared the rates of automobile use to the number residents and employees per square mile.
The underlying abstract says:
> Theory suggests that as automobile mode share increases in a city, the amount of land used for transportation also increases, whereas the land available for other uses decreases. This can result in a loss of activities from the city. This study compiles data from 12 cities in the United States to test these theoretical relationships. The findings suggest that on average each increase of 10 percentage points in the portion of commuters traveling by automobile is associated with an increase of more than 2500 m2 of parking per 1000 people and a decrease of 1700 people/km2.
It's saying that the more the city promotes car commuting, the fewer activities are in the city ("decreased robustness").
Small cities can also be dense cities. The paper included Cambridge, MA and Berkeley, CA in the study. Both have populations around 110,000 with a relatively dense population.
In other words think of congestion as the engine of public transportation systems. Now, looking a small city's inside major metro areas vs outside major metro areas congestion is vary different.
PS: Not to mention the relative student populations have direct effects on car ownership. Now, if they had looked at more city's this would have been less of an issue, but look at the 12 chosen.
New Haven CT: 129k Hartford CT: 125k Lowell MA: 107k Cambridge MA: 106k Berkeley CA: 113k
For Arlington, I don't know which city they're talking about. It appears, regardless of the state, that it's the odd one out, so I'm not sure why they included it. Arlington MA: 42k Arlington VA: 216k
> But Donald Shoup, a professor of urban planning at the University of California, Los Angeles, and the author of “The High Cost of Free Parking,” sees Santa Monica’s plan as a pale imitation of the more ambitious plans in San Francisco and Los Angeles.
> “I don’t see how this increases turnover — it just makes sure they get everyone to pay and they know how much they are getting,” said Professor Shoup, who is widely considered the intelligent-parking guru. “Anytime someone says something isn’t about money, it’s about money.” http://www.nytimes.com/2012/06/13/us/santa-monica-resets-par...
(It looks like LA is using variable rate meters, which he says is better at improving turnover and of getting to the real market rate.)
More at http://la.curbed.com/archives/2010/10/shoup.php:
> As an example in Santa Monica, which is starting to move in Shoup direction, but not quite there, we have some garages in Downtown that are in the red almost all the time during the week, and many blocks where on street parking is impossible, while hundreds of spaces are routinely unused at the Main Library underground lot and Civic Center garage at the same time. I know this because Santa Monica actually has a city website with live tracking of parking spots remaining at each of it's city lots.
> If there were stronger price incentives (meaning make high demand lots and street spots more expensive while making under utilized lots and spaces dirt cheap) and advertising of the availability of the under utilized spaces, instead of the same garages filling up to the max and people freaking out that we need more and more parking, maybe we'd realize we actually have plenty. Then instead of wasting yet more land use and city money for more parking capacity, we could start making more open space for people, such as more park space and plazas.
That was before the new meter system. Apparently, there's now an app for finding parking.
That's all I could find before I grew disinterested.
I don't own a car, so I haven't paid _too_ much attention to parking, but it seems like there isn't a whole lot of space down near the promenade, or basically anywhere in the whole touristy part of town.
I walked in some residential neighborhoods tonight, and lots of people had garages and driveways, so even one (admittedly long) block off of a thoroughfare, there was a large amount of on-street parking.
Based on the linked summary, "If the function of parking in these places was to enable growth and development, the data suggests they were abysmal failures." What this paper says is that increasing parking does not help the city, economically speaking.
If the thesis is correct - which the data seems to support - then having cars diminishes the options for those who want a "vibrant city." If you're in favor of options, then surely you should be in favor of no-car parts of town, no?
If you're in favor of more options, then (as the author comments on a link from his home page), do you think that "factors like suburban infrastructure and the distance to most business centers and even the distance between many stores have made people reliant on motor vehicles" and reduced the transportation options?
You write that you don't want to remove options. The issue seems to be that some options have already been removed, and should be restored.
At the very least, it would get the cycle-only weirdos and walking fetishists out of the parts of town that you want to drive to.
Which the hand picked examples seem to support.
Any time I see "country/state/city did X and got result Y" I am immediately suspicious of the lack of broadly sampled statistics. It's very easy to cherrypick examples for and against any policy setting you care about.
In any case, my question was 'Where do you get that the author wants to completely remove the option of driving?'
South Korea has much less land and much fewer resources, yet (relative to the rest of the world) it's just as wealthy as the United States. Ditto for Ireland, or UK (you could scream "but they were an empire!", but the question is -- how did they become an empire in the first place?)
Russia has land and oil. The situation there has greatly improved -- and having grown up there myself, I was still lucky to have had such basics an education, safe drinking water, and other utilities -- yet it's nowhere near as wealthy as South Korea or the United States.
Incidentally I disagree with your overall point (about zero-sum games, or that there's a binary choice between car-dependent suburbs and "dense urban living") -- but you're using a horrible argument to support it, too.
Try parking a car in Seoul. We are still talking about cars right?
> Ditto for Ireland, or UK
Have you been to these countries before? Really? Car usage is anything but cheap there, it's definitely moderated compared to the states!
> Russia has land and oil
I'm sure cars are cheap in Russia right? Why did they have to block used cars being sent from Japan to Vladivostok?
I really wonder if we are talking about the same thing (that cars are cheap to buy and use vs, ANYWHERE else). Even in Iran you have to buy the car before you can take advantage of cheap subsidized gas.
Small town, in my case, 90 minutes from the nearest mid-size city.
If you don't demand that I pay for rapid transit I'll never see, or use, don't restrict the kind of car or truck I can buy, then deal.
The whole point of infrastructure is that it benefits society as a whole. This is just as true of city infrastructure as it is infrastructure in the countryside.
(Not to mention there's a whole lot of countryside that doesn't produce food...or much of anything, really, other than a place for misanthropic people to retire to.)
I think it is possible you underestimate how much manufacturing happens outside urban areas.
Your argument about 'how we need each other' has a point.
Does that mean I get to vote on how you run your trains, down there in Big City, USA?
Also, cities of this size don't tend to have rapid transit systems, unless they are part of a large metro area. Instead, they have buses or sometimes trams. Your use of "rapid transit" suggests that you are using a 'Big City, USA' like NYC or Chicago as your reference, and not a medium-sized city like Green Bay, WI. This study did not examine the policies appropriate for a big city, but it's rather obvious that NYC can't handle even one car per household.
In any case, the debate isn't in restricting the kind of vehicle you can buy, but in how much a medium-sized city should subsidize your ability to be able to drive to and park downtown, compared to the other things it could subsidize.
For example, is it better to require more free parking through zoning laws, or to have more space for shops and restaurants though with more limited parking options? Is it better to have free parking, paid fixed-rate parking, or variable rate parking?
If "better" is defined as "enable growth and development" then this paper suggests that more free parking is not the optimal choice.
Unfortunately, that causes more problems for drivers than it does for transit riders (which is why we are in the current state that we are in). You see, the cost structure of automobiles is linear, and when you throw in density, exponential. The cost structure for transit is inverse to its ridership (within vehicle capacity limits).
In other words, small increases in ridership produce huge decreases in costs per passenger. So much so that the most heavily used transit systems in the world are completely profitable (Hong Kong's MTR is an exceptionally profitable publicly traded corporation).
Now currently, both roads and transit systems are subsidized. But since they are substitutes (in the economic definition), a small decrease in subsidies to roads will push more people to transit, thus requiring fewer subsidies to transit.
But we can't have that, can we?