Roads
The next black swan for transportation: self-driving cars
Whether we are prepared for it or not, the next revolution in transportation will be here soon, and it won't be streetcars, monorails, segways, or electric vehicles. It will be self-driving cars, and the adoption of this technology will change everything we accept as a given in the field of transportation planning.
Photo by PMC 1stPix on Flickr.
There is a fundamental flaw in the practice of transportation planning. Our local and regional transportation models assume that 20 years from now, the transportation system will be largely the same, with slight adjustments on the margin.
But history shows that every so often, unexpected technology arises. The most obvious was the move from horses and horse-conveyed vehicles for personal and short-haul transportation to the automobile, but it also happened with railroads superseding canals and short-distance shipping, and airplanes superseding railroads for long-distance personal travel.
In only 32 years at the turn of the twentieth century, the primary power source for human mobility for 7 millennia became obsolete. If there was a comparable transportation planning field at the time, using our current system of forecasting, their models from 1890 would have estimated a linear increase in horse trips for the next 30 years, with some additional subway and horse-drawn streetcar lines.
They would not have imagined that there would be no horses at all. Instead, they would need a complex set of signal and traffic control technologies to address the new safety and mobility issues presented by automobiles.
Today we are at a similar inflection point. The occasional news story in Popular Science or the New York Times describes the wondrous technology that allows cars to operate under complex scenarios without any driver input. However, we don't hear a peep from transportation planning organizations about how society will adapt and plan for this change, even though it seems increasingly imminent.
Washingtonian recently interviewed Michael Pack, the region's foremost traffic technologist. The interview is alarming: "I ask how long before we can all stop driving and let the cars do the work," the interviewer asks. Pack responds, "Oh, a while, Maybe 30 years." Pack appears to be considering a network of interconnected cars that can talk to each other and have a situational awareness that allows them to travel at 65 mph and 6 inches from one bumper to the next.
If Pack's estimate is reliable, then that means we should start seeing the first generation of autonomous cars that replace the human driver with superior situational awareness in a matter of years. It also means we are already within the 30-year window where transportation planners should anticipate adoption of this technology and its consequences.
My best guess, based on publicly available information, is that within 7-12 years, there will be a commercially available autonomous vehicle sold in the US
In sharing this theory, I've heard my share of skepticism. But most people would accept on faith that within 100 years we'll have autonomous vehicles. Some would accept that they'll be here in 50 years, while few would accept that we'll have them in 5 years. So at heart, the discussion is not a question of if this technology will develop, but when, and whether we have to start thinking about it from a policy and planning standpoint.
Whichever manufacturer is first to roll out a consumer-ready version of this technology will have a blockbuster product, so the economic incentive to be first is enormous.
In a future post, I'll discuss some of the changes we're likely to see in a world with self-driving cars.
 | Will Handsfield earned his master's degree in public policy from the University of Denver in 2008, and has worked on transportation projects in Los Angeles, Denver and Washington, DC. Will bike commutes and lives with his wife and son in the Lincoln Park neighborhood of Capitol Hill. His posts reflect his personal views.   |
Comments
- Successful speed cameras require fair speed limits
- VDOT ignores own data, pushes widening I-66
- DC's divide need not be black and white
- Preservationists ask to shrink 3rd Church replacement
- Live chat: Matt Yglesias, Wednesday at noon
- Half-hour Metro headways are not acceptable
- "Degree density" maps show region's east-west divide
we don't hear a peep from transportation planning organizations about how society will adapt and plan for this change, even though it seems increasingly imminent.
A curmudgeon might argue that we won't see "self-driving" cars at all--for the same reason we won't see GPS-linked speed limiters. The hurdles involved are not technological ones.
by oboe on Dec 6, 2011 2:34 pm
by Lance on Dec 6, 2011 2:39 pm
Americans see cars and highways as the tried and true, and we're a conservative (in the non-political sense of the word) people in general. Electric, self-driving cars could be implemented at the the local level and wouldn't require the large, centralized planning that public transportation plans (particularly rail) need to be implemented and maintained. People could maintain the freedom, comfort, and safety of personal travel and continue to live in the suburbs while keeping our highways safer and functional and at the same time being kinder to the environment (the best of both worlds).
Electric, self-driving cars will be the end of the urbanist argument for densification as they will allow the suburban development pattern to continue indefinitely. You can revel in this or take sorrow in it, but it's the truth.
by Larry on Dec 6, 2011 2:47 pm
by canaan on Dec 6, 2011 2:49 pm
by Thayer-D on Dec 6, 2011 2:55 pm
by curious george on Dec 6, 2011 2:56 pm
by Not sure about this on Dec 6, 2011 2:56 pm
Well! It would certainly help deal with parking, taxi-driving, and more. What would be the price-per-mile of these vehicles? What would be the time saved? Would mass transit become obsolete, or would we consider new ways of doing mass transit? Already, slugging functions as an informal mass transit system in VA, increasing the efficiency of cars.
That said, I don't think the technology would be quite as much of a game-changer as people think. It would enable us to more efficiently store cars and help mitigate parking concerns, but the human and economic benefits of density would remain the same. You will still have to fuel or charge the car, you will still have to maintain it, and you will still like to walk places.
by OctaviusIII on Dec 6, 2011 2:59 pm
by mark76 on Dec 6, 2011 3:00 pm
Who would it appeal to?
by HogWash on Dec 6, 2011 3:02 pm
by OctaviusIII on Dec 6, 2011 3:03 pm
you are aware that there is govt $$ in this (its called intelligent vehicle/intelligent highways, BTW). There will likely be govt involvement in implementing it.
by AWalkerInTheCity on Dec 6, 2011 3:05 pm
by HogWash on Dec 6, 2011 3:05 pm
Several of the comments have already been addressed in Dave alperts earlier piece
A. mixing with older cars - there could be autonomous vehicle only lanes
B. transit - this could make buses more economical
Personally I think this could lead to benefits. My problem is claims it will lead to ALL the benefits, all at once. Shorter car car distances (and thus more throughput on roads) at the same safety level. More safety but at the same throughput. But more safety, no driver monitoring, AND short car to car distances - ALL very soon? hmmmm.
by AwalkerIntheCity on Dec 6, 2011 3:13 pm
And if you're saying that parking is no longer an issue because people can just send their self-driving car home, you've completely negated any benefit gained by that 6 inch spacing because you've DOUBLED the number of trips the car has to make. You're still going to get traffic jams if that's what people start doing. And by the way, 6 inch spacing only works on long stretches of highway without any on ramps feeding more traffic. Not exactly workable on an urban freeway.
Seems to me that self-driving cars will only bolster the urbanist vision. If you're not actually controlling the vehicle, I can only imagine that would further erode people's personal attachment. Bottom line is there's a real desire to live in PLACES where you can actually walk to things. Just look at real estate prices in the most walkable neighborhoods in the country and new, mixed-use developments. Those will give you an indication of where things are headed, no matter how we get around.
by Alex on Dec 6, 2011 3:18 pm
We can't rely on hype-o-thetical technological magic bullets to solve the horrible problems with car-dependent land uses. PRT was supposed to be some sort of magic bullet that would solve traffic problems. I've read stuff like this article before and how self-driving cars will ensure the end of traffic jams. I've also read about how programmers can't solve the problem of judgement. A computer locks up when the input isn't something it's programmed to expect and categorize. Driving is all about judgement. Computers don't judge.
Take this article and email it to Randall O'Toole. He and other Sprawl Lobby folks will be happy to trumpet self-driving cars as a reason not to invest in transit and traditional pedestrian-oriented human settlements until the end of time. Or, maybe email a PRT crank website. Wasn't that technology supposed to be ready "in ten years" back in 1970? This is more of the same.
by Cavan on Dec 6, 2011 3:19 pm
Blame the Christian Church.
by Drg998 on Dec 6, 2011 3:21 pm
Yes, train systems that move as many people as an 8-lane highway are a "boondoggle," but the $450B (and counting) Interstate Highway System is the free market at its finest! I also like how apparently suburban housing = better commutes, when clearly in reality suburban housing = longer, worse commutes.
Nobody's trying to force people to live in an urban high-rise. You just have to accept that if you want to live far outside the city you can't continue to expect city residents to submit to the destruction of their homes and ways of life just so that it's as easy as possible for you to drive your car from your home to your workplace.
Back to driverless cars:
How much have the basic fundamentals of car operation changed in the 80+ years since the Model T? By my estimation not much. And driverless cars still doesn't change the fundamental problem of ever increasing travel distances.
by MLD on Dec 6, 2011 3:23 pm
by Sandy on Dec 6, 2011 3:30 pm
I agree w/whoever mentioned that this wouldn't take more cars off the road since the cars (sent home) would have to make two trips. In fact, the car ends up making four total trips and although I have no figures, I can't imagine how doing so makes it more eco-friendly than a person driving in and going home.
I wonder if the whole idea is more a matter of who can afford it than an actual need to have this sort of vehicle.
by HogWash on Dec 6, 2011 3:30 pm
Assuming people be able to sleep or play on their tablets while the diverless car travels, doesn't that mean people will be more willing to spend more time in their car? This would make long commutes more viable and suburbs more desirable.
There goes taxi drivers. No jobs for them.
Will there be enough energy for the masses to own these and power them as energy constraints get worse over few decades?
How many lives could these driverless cars save?
Will the masses accept this technology?
by David J on Dec 6, 2011 3:48 pm
As an example of a computer that massively consumes and needs data, pretty cool. As an example of a self driving call, not very practical.
(short version -- they use lasers and radar to map out an entire street. Then the drop that into a computer, and the car compares the current version against the stored 3d map.)
Horses were far from the "primary" power source for 7 thousand years. Maybe more like 3, and even that is a stretch. Your two legs have always been the primary power source. And I'm pretty sure the 19th century saw widespread adoption of this thing called the railroad.
But onwards technology utopia. Just don't forget to stop at Taco Bell.
by charlie on Dec 6, 2011 3:49 pm
What does this actually mean for people? Well, it absolutely means that many will chose to commute from further away. On the other hand, the same people that prefer urban living, including myself, will continue their lives in the city... but with vastly improved taxi service. I really can't estimate exactly how efficient traffic would be, but it's entirely possible that there would be no need for public transit other than autonomous taxis and buses. Buses may be able to be smart enough to actually deviate from their route to drop people off closer to where they need to go without going off schedule much.
The possibilities, to me, are rather mind-bending.
by rowsdower on Dec 6, 2011 3:50 pm
by Roger on Dec 6, 2011 3:50 pm
The robot isn't mad that it has to share the lane with me. The robot knows how to parallel park and watch for bikes at the same time. The robot won't drop its focus to answer the phone or talk to its buddy in the back seat. And if I hit a bump and fall off my bike, the robot can hit the brakes before I'm even done falling.
But before any of that can happen, we need to make sure that the programmers have the regulatory and financial incentives to program for urban safety, rather than just programming to make I-95 less frustrating. So yes, this is definitely a technology we need to watch, even at this early stage.
by tom veil on Dec 6, 2011 3:50 pm
I agree with some of the above posters who say this may be the nail in the coffin of urbanism.
Plus, how are you going to be walking and biking when the other 99% of the population is whizzing around at 60 mph in self-driving cars? You'll see even more pressure than today to "get a car!" as they used to say around the turn of the last century.
by GTimmons on Dec 6, 2011 4:03 pm
The solution isn't in some huge government boondoggle of a train system, but allow the free market and the private sector to cater to the public demand, in the case, the demand for suburban housing and better commutes.
Oh, you'd like to impose self-driving, government-regulated, speed-limited cars wouldn't you? Well, you can take my steering wheel from me when you pry it from my cold dead fingers, you hippie!
by Oboe's Id on Dec 6, 2011 4:04 pm
Driving is now the cheapest it's ever been. It gets cheaper each year due to newer and newer technology which lowers and lowers both the cost to build better and better cars, lower and lower costs to operate such cars, and better and better ways to get energy to power such cars including finding and extracting the current standard fuel (i.e., oil.) And yes, of course I mean in inflation adjusted dollars.
by Lance on Dec 6, 2011 4:04 pm
by w on Dec 6, 2011 4:04 pm
wouldnt it be more logical to use this to argue againt more road lanes? IF we are going to need 1/2 as much asphalt to serve the same number of vehicles, or even less, why build as much?
@tom - some of the techs described only work if all nearby vehicles use the same control technology, so all movements are predictable. That wouldnt work sharing the road with human cyclists, I think.
@rows - bus drivers could do that - The reasons not to stay the same, I think.
When do we get perfectly safe rail transit with close headways? That would seem to be a much simpler problem, but we arent there yet.
by AWalkerInTheCity on Dec 6, 2011 4:04 pm
He acts as though everyone wants to live in the sprawly suburbs and only evil social engineers keep people in the cities.
Yet the most expensive, fastest-turnover, most in-demand housing in the nation is in the central core of big cities. Go look at prices in Dupont Circle or Penn Quarter if you don't believe me.
In a fight between mark76's screeds and the free market, I'll take the free market every time.
And if self-driving cars can also negotiate among themselves for parking, they will bring a new frontier in convenient city living. Imagine the city without the constant parking hassles and the congestion of people looking for parking.
by dal20402 on Dec 6, 2011 4:08 pm
Lance got you on that one! Driving is just going to get cheaper and cheaper as we build increasingly efficient cars. It's going to be the duty of government to get out of the way with its crazy transportation schemes (HSR, light rail, and so on) and allow the free market drive us (all puns intended) where we want to go. Urbanists need to face the fact that their mode of living is destined to be a niche market, at least in the U.S. The market and personal preference are undeniably for the auto and low-density suburbia.
by Larry on Dec 6, 2011 4:10 pm
(http://www1.eere.energy.gov/vehiclesandfuels/facts/2005/fcvt_fotw365.html)
by oboe on Dec 6, 2011 4:10 pm
While partly this is a function of just how awful suburban sprawl has gotten, really the main thing driving this is just changing tastes. People who are under 40 tend to want to live in more urban, walkable areas even when they have kids. That means more and more older people whose tastes run counter to their cohort have that option as well.
Of course, if you're 50+ years old and think "Key Largo" and "Pina Colada (The Escape Song)" are the pinnacle of musical achievement in the 21st century, you're going to think someone who listens to Moby or Wilco is an out-of-touch outlier.
by oboe on Dec 6, 2011 4:15 pm
Just wait until those couples get a taste of those top-notch inner-city schools. That'll have them running for a "car infested" suburb real fast!! :)
by craigster on Dec 6, 2011 4:18 pm
Well, at least that's what's happened in the 80s and 90s. Of course, the last ten years has seen a big jump in DCPS elementary schools with middle-class families in places like Capitol Hill. Also the explosion in public charters.
I'd say this is actually a symptom of how terrified most city-dwellers are about the prospect of moving out. Ten years ago the obstacle was finding a decent elementary school. Now middle school looms large. Should be interesting to see what happens over the next decade.
by oboe on Dec 6, 2011 4:27 pm
re: parking: I'm not sure the only options with this technology would be parking your car at work or sending it home. It would seem that after dropping us off at our destinations, robo-cars could park, very tightly and efficiently, in lots or garages located further than humans might want to walk, but a lot closer than home. Basically, it would open a lot of new areas for parking.
re: personal attachment to cars: Since all cars would move at the same speed and do the same sorts of maneuvers, people might not be as excited about the brand and performance of their cars. The main differences between cars would probably be in terms of interior amenities. Car ownership might even be replaced with some sort of car-sharing on a huge scale.
re: public transportation: This technology offers tremendous potential for safety improvements. Just imagine if we had WMATA buses, operated without any WMATA bus drivers running into lampposts, bicycles, parked cars, etc.!
by Mike on Dec 6, 2011 4:33 pm
2. WRT MLD's comment:
How much have the basic fundamentals of car operation changed in the 80+ years since the Model T? By my estimation not much.
I agree. And I laugh to myself when the anti-streetcar people rail against streetcars as a 19th century technology when for the most part, so are cars.
by Richard Layman on Dec 6, 2011 4:37 pm
One comment from @tom veil got me thinking of my bike commute this morning. Along East Capitol St. North of Lincoln Park, a Mercedes had demolished its front end into the back of a Tahoe. This stretch of road never moves faster than about 20 mph since it has a couple of raised crosswalks, and most mornings it is stop and go with MD commuters. The only explanation is that the Mercedes driver wasn't paying attention, and crashed at 10-15 mph into the rear of the Tahoe.
I'll jump on board with Tom and say I would have more faith in a thoroughly vetted computer system controlling vehicles than I do with easily distracted humans.
by Will Handsfield on Dec 6, 2011 4:41 pm
It is IMPOSSIBLE to say what the impact of smart cars will be without knowing their actual performance specs - able to mix with other vehicles, or not? Savings over 1/2 of road space? less? Only on exit free stretches of highway?
Given that we still havent managed safe driver free rail transit,AFAICT, I dont see how we estimate the realism of a given achievement schedule for these specs.
I mean we were also promised that broad band internet would eliminate commuting and make all locations equal - thats a more radical promise, with technology that is already widespread. yet it has not been fulfilled.
by AWalkerInTheCity on Dec 6, 2011 4:45 pm
IF you are talking about the same traffic conditions, maybe, yeah. IE take the safety benefits and forego any impact on congestion, and hence, urban form. But you're talking about 6 inch seperations at 60 mph. Suppose the computer is 10% better than an average human being - thats NOT enough, by far, to make that safe or feasible.
I would feel better about this tech, if i wasnt hearing that kind of failure to rigourously analyze costs and benefits.
by AWalkerInThecity on Dec 6, 2011 4:58 pm
by dcd on Dec 6, 2011 5:12 pm
2. Cops on segways aren't much different than cops on bikes. It hasn't been revolutionary. In most instances cops on Segways aren't superior to deployment on bikes, but it costs more and because it is a newer technology there is more pressure to adopt the Segway instead of equally serviceable and cheaper technologies.
Segways do make a difference on "walking" tours, agreed. Probably not to the point of "revolutionary."
by Richard Layman on Dec 6, 2011 5:45 pm
"Carsharing" or some sort of PRT system where you pick up a small number of riders along the way makes a *TON* more sense under that scenario. (Of course, balancing the most efficient route with demand remains a famously unsolved problem in computer science, and would be an extraordinarily difficult computational problem if you had a large number of vehicles on a complicated network of streets, but I digress)
You'd end up with less deadheading, and would be able to anticipate demand in areas that need more vehicles.
by andrew on Dec 6, 2011 5:57 pm
by Phil LaCombe on Dec 6, 2011 7:08 pm
by lee on Dec 6, 2011 7:20 pm
Just fast forward to the future with self-driving taxis gliding along city streets, and self-driving trucks roaring along our highways, while millions more stand in the unemployment lines.
by interguru on Dec 6, 2011 7:26 pm
by spookiness on Dec 6, 2011 7:36 pm
It's the sort of thing that techno-utopians get excited about, but that the Teabaggers of 2150 will fight with every fiber of their being.
by oboe on Dec 6, 2011 7:55 pm
Good point. Actually, there was a recent case where a local urbanism blogger moved out of the city because of the dearth of ample parking.
by oboe on Dec 6, 2011 8:01 pm
Less than the price of current cars -- about 50 cents per vehicle-mile, which is about 30 cents per passenger-mile. Which is one of the reasons why self-driving cars will lead to the virtual extinction of urban mass transit. No one's going to bother with buses and trains when they can get a direct, on-demand, door-to-door taxi ride for a fraction of the price.
by Bertie on Dec 6, 2011 8:02 pm
by Steven on Dec 6, 2011 8:09 pm
The 2010 Census found that the trend of suburbanization has continued during the most recent decade. Almost all cities are either growing more slowly than their surrounding suburbs, or are actually losing population. And even among cities that are growing, most of that growth seems to be coming from immigrants, not domestic migrants moving "back to the city." New York City is a good example. It lost a million domestic migrants between 2000 and 2010. Where is this alleged trend of "younger people opting for more walkable living closer to their jobs?"
by Bertie on Dec 6, 2011 8:13 pm
instead of drive-thoughts, there could be services geared towards vehicles that go to pick things up for you - like malls and groceries and take-out places that only driver less vehicles might ever enter. Do we really need stores in the city anymore when you can just send your vehicle out to some costco or target geared for automated vehicle pick-ups? it could do it overnight, picking up whatever's next on your phone's list. Maybe the store could be a big automated vending machine, so we don't need store employees. Maybe there will be no place left to walk to when you can just send your car to get things and nobody actually knows where the car is going or has even been there? People won't go places anymore, because cars could bring everything to them. Maybe it will be the death of travel for many people. Who will go get things when you have a servant to do that for you?
by lee on Dec 6, 2011 8:25 pm
Self-driving cars will give people who are currently dependent on walking/biking and mass transit the option of car travel instead, via cheap self-driving taxis. They will also make car travel much easier for people who already get around mostly by car, which is the vast majority of people. So self-driving cars will tend to promote, not reduce, car-oriented lifestyles and car-oriented urban development.
by Bertie on Dec 6, 2011 8:37 pm
by Gray on Dec 6, 2011 8:37 pm
Not just the tech of the cars, but the impact on congestion, cities, urban planning, etc
by Doug on Dec 6, 2011 8:51 pm
by Doug on Dec 6, 2011 8:53 pm
by burdlyn on Dec 6, 2011 9:01 pm
That's the magic thing about capitalism and free markets; as old jobs are made obsolete, new ones are created by increased demand due to greater efficiencies, and the net amount of wealth always goes up. The challenge is to re-educate yourself and adapt to a new environment!
by Chesney on Dec 6, 2011 9:09 pm
For the same reason that private ownership is the norm for almost all consumer goods, from houses to clothes to computers. People like having stuff to which they have exclusive rights for themselves and their family. It's more convenient. It's more private. It's more satisfying. Long-term leasing or renting might be popular options, but they're effectively equivalent to temporary private ownership. And self-driving cars will make car ownership even cheaper than it is now. In a world full of self-driving cars, insurance costs will be lower, fuel costs will be lower, travel time costs will be lower, parking costs will be lower. Maintenance costs have long been declining, as cars become increasingly reliable and durable. My new truck goes twice as far between scheduled service intervals as my old one.
"Carsharing" or some sort of PRT system where you pick up a small number of riders along the way makes a *TON* more sense under that scenario.
I doubt most customers of self-driving taxis would bother doing that. People rarely share taxis with strangers even today, when taxis are expensive. A self-driving taxi ride will cost less than a bus ride today. The government could pay the entire cost of the ride for less than it pays today in subsidies for a bus ticket.
by Bertie on Dec 6, 2011 9:18 pm
I think there actually have been some pretty phenomenal changes in cars over the last 80+ years: e.g., hydraulic brakes, automatic spark advance, electric windshield wipers and automatic transmissions, just to name a few of the earlier advances. But the changes have largely been evolutionary, not revolutionary. And I have to think the transition to driverless cars will be incremental as well.
It seems to me though that the effort to create the self-driving car is an admission that driving is a joyless chore. This is in tension with nearly all the marketing behind new cars. If cars are just appliances now, what happens to brand identity?
by tdballo on Dec 6, 2011 10:40 pm
A couple of anecdotes about the trend of middle-class people moving into DC:
Population, percent change, 2000 to 2010 5.2%
Median household income, 2009 $58,906
Median household income, 2000 $$40,127
http://quickfacts.census.gov/qfd/states/11000.html
by oboe on Dec 6, 2011 10:49 pm
And I doubt cars will ever become "just appliances." Not that brand identity doesn't matter for appliances anyway (love those Viking stoves, Dyson vacuums and Subzero refrigerators). Styling, comfort, features, power, capacity, etc. are probably still going to matter to people whether they drive the vehicle themselves or not.
by Bertie on Dec 6, 2011 11:14 pm
by Bueller on Dec 6, 2011 11:28 pm
That's not evidence of a trend of middle-class people moving into DC. As in New York, the (small) population increase may have come from immigration and natural increase (excess of births over deaths) rather than domestic migration, and the rise in median income may have come from lower-income people moving out of DC. Certainly, SOME middle class people have moved into DC over the past 10 years. But many more may have moved out of it. And DC is just one city anyway. You can't infer a national trend from changes in a single city.
by Bertie on Dec 6, 2011 11:35 pm
by Thayer-D on Dec 7, 2011 4:44 am
Taking multiple fares on one ride (you could just have multiple meters in one taxi, or computerize it) is a congestion reduction measure that should be supported.
But taxis aren't addressed by the Transportation Element of the Comp. Plan.
by Richard Layman on Dec 7, 2011 6:30 am
WRT driverless cars, it'd be relatively cheap to implement the technology for highways (relatively not actually). I think the cost to do it on surface streets, in cities, is likely to be far greater than what any locality can afford.
by Richard Layman on Dec 7, 2011 6:33 am
This is Census data, not anecdote.
But there is no question that by observation over the past 24 years of living here, in the past 10 years, there has been a significant amount of in-migration of white or mixed households in many middle class neighborhoods in the city.
If you believe Leinberger's research about housing preferences, he finds 30% of people prefer urban living, 40% prefer what we call typical suburban, and the other 30% can live in either, basically they don't care.
Plenty of people live in what we would think of as urban-like places in the suburbs, but because traditionally the cities have been the sinkhole of poverty, the suburban urban dwellers get to live in urban conditions without the problems typically associated with center cities (crime, poor schools, disinvestment in roads and other infrastructure).
E.g. my joke about Bethesda Row being urban without being "urban." Even with plenty of black middle class residents in Montgomery County, Bethesda Row feels mostly white/foreign high income ethnics.
by Richard Layman on Dec 7, 2011 9:11 am
Autonomous operation is only an ancillary part of the solution.
The next revolutionary solution will effectively address the energy inefficiency of personal automobiles.
There are (and always will be) many unfeasible ideas trumpeted.
All known attempts at this next revolution so far are failures. (This means the Volt, the Leaf, high speed and maglev rail, etc. and all the crackpot crazy crap like PRT)
The closest thing to success is the golf cart. (Calling it Neighborhood Electric Vehicle doesn't change what it is, and didn't really advance anything.)
The upcoming revolutionary transporation product is probably in somebody's garage under development and testing right now.
by Robert on Dec 7, 2011 9:20 am
As in New York, the (small) population increase may have come from immigration and natural increase (excess of births over deaths) rather than domestic migration, and the rise in median income may have come from lower-income people moving out of DC. Certainly, SOME middle class people have moved into DC over the past 10 years. But many more may have moved out of it. And DC is just one city anyway. You can't infer a national trend from changes in a single city.
I suppose I should have explained the numbers a bit more clearly: The numbers show a combination of 5% population growth coupled with a nearly 50% rise in median household income in DC over a little less than one decade. For your speculation about DC to be correct two things would have to be true: A) the birth rate among the poor would have to have stalled completely, and b) the net population growth per middle-class household would have to be off the charts. And I mean, like, rampant Mormon polygamy -style numbers.
I'm not saying that's impossible, just that as far as supporting your hypothesis, it makes anecdotal speculation look like data meticulously gleaned from a decades long double-blind randomized longitudinal study.
DC's population grew by 5% over the decade after about a half century of hemorrhaging people. The median income growth was eye-popping at this time. Combining these two data, we get overwhelming evidence for an increase of wealthy households. Since wealthier people tend to have fewer children than poorer ones, the raw population increase is likely more significant than it appears.
Finally, while there's significant evidence that this trend is a national one, certainly there are some cities that have benefited more than others. Without pulling up a bunch of numbers, it's likely that LA or Chicago benefits from this dynamic more than, say, Oklahoma City. But in order to show that there's a nationwide trend towards urban gentrification, and a suburbanization of poverty, we don't have to show that every American city has been as radically transformed as DC.
So as far as your point that "DC is just one city anyway", you're right. But, obviously it's the one most relevant to GGW.
by oboe on Dec 7, 2011 9:58 am
The AAA's own studies say otherwise. AAA issues an annual "Your Driving Costs" study, which has generally shown continual climbs, with very few exceptions indeed, the exact opposite of what Lance and Larry claim: http://prn.to/vwhp3a
2011: +3.4%
2010: +4.8%
2009: -0.1%
2008: appx. +3.4% (% not provided, but benchmarked at same cent-per-mile increase of 2011 study)
AAA also reported a tiny decrease in 2002, but reported that was the first decrease in a decade, i.e., costs had increased year-on-year from 1993 to 2002: http://bit.ly/rJkIcn
by Joel on Dec 7, 2011 10:16 am
It goes against some of our cherished notions as Americans, but the technological advances of the future are unlikely to be invented by someone working out of his garage or basement (excepting solely "software" type developments). The low-hanging fruit has already been plucked.
And why is HSR a crackpot scheme? Unlike PRT or Maglev, it's proven technology successfully implemented in numerous countries.
by JimT on Dec 7, 2011 10:36 am
I predict that a driverless car will be available from a major car maker in ten years -- say, 2020 or 2023. I do not expect a revolution, because the entire fleet will need to be replaced.
Regarding "who would want one?" I expect that the people that cannot drive, that they depend on others to get around, will be among the first adapters. Consider that MetroAccess" is there to provide transport for such people, which may be antiquated by this technology. More people will adapt when the autonomous driver is better at it than normal people.
by goldfish on Dec 7, 2011 10:42 am
******
There are plenty of heavy rail systems that run without operators.
The above two statement illustrate exactly the difficulty of widescale adoption of self-driving technology. As with metrorail, the difficulty is not developing the technology (self-driving metro cars are obviously possible), it's the cultural barrier.
What we've learned from Metro is if there is even ONE major accident in 30+ years, then the system is deemed critically unsafe and fixes/upgrades that take years and billions of dollars are necessary to satisfy the public. Similarly, it will only take one major failure for self-driving cars for people to become afraid of them.
Also, as we know with flying, people fear not being in control of their own safety, regardless of the statistics. I just don't see self-driving cars overcoming human nature's opposition before some other technology makes even self-driving cars obsolete.
by Falls Church on Dec 7, 2011 11:27 am
If we ever see this technology deployed, it'll be mandatory in the urban core, or voluntary on interstates.
by oboe on Dec 7, 2011 11:38 am
I know Oboe likes to talk about the coming suburban collapse caused by disinvestment in infrastructure, but if there's a collapse to be had, it's more likely to be an urban one (it's happened before). Some cities like DC and NYC have had a free ride on infrastructure until now because they had so much catching up to do to get to their previous population levels. For example, after 20 years of growth, NYC is just now at the same population level it was at in the 1950s.
During this catch up period, cities got a free ride leveraging unused infrastructure from the past. Also, cities didn't need to invest in school infrastructure because the only folks with kids in urban schools were powerless poor families. As cities start to exceed their previous peak pops and as the demand for reasonable school quality increases, cities will suddenly find themselves in a much more difficult situation.
Cities have aging infrastructure that has undergone decades of under-investment. For example, DC's $2.5B repair/expansion of its sewer system is just the tip of the iceberg. There's a lot more work DC needs to do in repairing and expanding its infrastructure and school system (the schools themselves are often falling apart, not to mention the far more difficult task of improving the quality of instruction inside those buildings).
I don't think there will actually be an urban collapse (especially not in DC or NYC) but growth for cities is about to become a lot more challenging.
by Falls Church on Dec 7, 2011 11:57 am
HSR operations depend on heavy subsidy, (just like Volt and Leaf sales.) Japan National Railways underwent "financial restructuring" in 1987 (i.e. bankruptcy) due to debt load from Shinkansen construction. The European HSR systems are propped up by their governments. If it were smart to do, then private enterprise would do it (without massive subsidy.)
"It goes against some of our cherished notions as Americans, but the technological advances of the future are unlikely to be invented by someone working out of his garage or basement (excepting solely "software" type developments). The low-hanging fruit has already been plucked."
That's an assumption that can be proven wrong, but never proven correct. The next "low-hanging" fruit innovation will prove it wrong (again.)
by Robert on Dec 7, 2011 11:59 am
In the context of this thread, where urban means high density walkable transit oriented rather than center city political jurisdiction the experience of Arlington may be relevant. AFAIK arlington is not and has not been in the recovery mode that DC and NYC are. Yet it has managed to achieve similar growth, and with similar increases in prices.
by AWalkerInTheCity on Dec 7, 2011 12:05 pm
Your comparison to New York illustrates the fallacy about thinking regarding the 'suburbs.'
New York's population hasn't changed much, that's true. But neither has its borders. And the city is largely built out.
Joel Kotkin and others have tried to use this to argue that the suburbs are the future, but one of their examples of a suburb that's growing was Jersey City. Jersey City is suburban only in the sense that it's not part of the central jurisdiction. Jersey City's built form is entirely urban. Kotkin's entire argument (echoed by O'Toole and others who don't think critically about it) rests on the peculiarities of anachronistic political boundaries.
As for cities being more likely to collapse because of infrastructure, I think that's entirely wrong and completely misses the point. The advantage of dense cities is that they use infrastructure more efficiently, therefore you can get more bang for your buck. The fact that various infrastructure networks already exist in cities and have to be built from scratch on the fringe is yet another point in favor of cities.
by Alex B. on Dec 7, 2011 12:11 pm
by Robert on Dec 7, 2011 12:13 pm
Why didn't private industry build the interstate highway system then? Must not have been a smart thing to do then.
That's an assumption that can be proven wrong, but never proven correct. The next "low-hanging" fruit innovation will prove it wrong (again.)
Well, we're waiting! :)
I'd say we already have the answer--bike lanes, buses, HSR, subways, pedestrianization, etc.
by Sandy on Dec 7, 2011 12:25 pm
The Brookings document you link to discusses population growth in certain cities' downtowns. That is not the same thing as city population growth, let alone city population growth due to domestic migration (i.e., a "back to the city" trend). "Downtown" is typically just a small area within a city, with a small fraction of the population. Chicago, for example, experienced significant growth in its downtown between 2000 and 2010, but the city overall lost 200,000 people.
And again, population growth is not the same thing as in-migration. The population of New York City grew 2.1% between 2000 and 2010, but this growth didn't come from people moving into NYC from the suburbs or other parts of the country. It came from immigration and natural increase. Domestic migration was in the opposite direction. A million more people moved out of NYC for other parts of the U.S. than moved into NYC from other parts of the U.S.
by Bertie on Dec 7, 2011 12:48 pm
by Robert on Dec 7, 2011 12:48 pm
yet people pay it. If consumers are spending more for something than its worth, doesnt that bring the whole free market thing into question?
(personally I think that given the options, the acela riders are making an entirely rational choice)
by AWalkerInThecity on Dec 7, 2011 1:00 pm
yes - isnt it clear that what we have is a movement that is to downtowns and adjacent areas, high density very convenient areas with multimodal options, NOT a necessarily a movement into particular political jurisdictions?
by AWalkerInTheCity on Dec 7, 2011 1:02 pm
I think you're missing the argument. I think the thought process about roads being private is that roads are a private enterprise because they require people to own cars to use them.
That (1) governments own the vast majority of roads in the country, (2) the road system is subsidized by general tax dollars (because the gas tax doesn't cover all expenses), and (3) that the government has to exercise eminent domain and seize private property to build portions of the system are all irrelevant facts. So long as private cars are involved the system _must_ be private.
The road system is absolutely a public system.
by Paul on Dec 7, 2011 1:03 pm
I'm not sure what technology you're talking about. The self-driving cars under development by Google and the major automakers do not require any new technology for highways. They're designed to operate on regular public roads alongside regular cars. Google's fleet of prototype self-driving cars has now driven 200,000 miles on public roads in all sorts of environments and conditions, from crowded city streets to rural freeways, without a single accident.
by Bertie on Dec 7, 2011 1:03 pm
You're correct, cities have natural advantages but the failure to develop honest and efficient institutions has meant that cities have not leveraged those advantages well. Poor governance and mismanaged institutions have meant decades of poor decisions and underinvestment which mean that cities have a deep hole from which to climb out of. However, their natural advantages and efficiencies should help them climb out of that hole.
In the context of this thread, where urban means high density walkable transit oriented rather than center city political jurisdiction the experience of Arlington may be relevant.
I agree. The future is probably the urbanification of the suburbs rather than their collapse. As the inner suburbs become more urban, the outer suburbs will become more sustainable.
by Falls Church on Dec 7, 2011 1:10 pm
thats not clear to me at all.
while I think Oboe is over the top, I dont think the issues he see's harming the outer suburbs, are really alleviated by anything happening in the inner suburbs. how does growth in Clarendon make Stafford County more sustainable? (note the tech corridor with its heavy employment concentration is a different ball of wax)
by AWalkerInTheCity on Dec 7, 2011 1:19 pm
there are several systems under development that do involve highway infrastructure. One might expect the first response to the clear, certain, success of the fully autonomous vehicles would be the end of research on the on systems involving highway infrastructure. I am not aware that has happened yet.
by AWalkerInTheCity on Dec 7, 2011 1:25 pm
Where's the profit in the interstate highway system?
What would the tolls be if all of the system construction, maintenance, financing, and administration costs had to be covered by the toll revenue exclusively? Toll roads were initially privately built, and profitable.
I have no idea what Paul's "thought process" is, and no matter what he says I assure you, he has no idea about mine either.
by Robert on Dec 7, 2011 1:28 pm
No, what you have is pockets of growth in certain areas of certain cities, often fueled by massive public subsidies or public works projects (sports stadiums, arts centers, convention centers, transit hubs, etc.) against the backdrop of a continued overall shift in population from cities to suburbs. And not dense suburbs like Jersey City, but typical post-war low-density car-oriented suburbs. DC is a clear example of this. The District grew modestly between 2000 and 2010. Its suburbs grew much, much more. Loudon County grew by 84%. Prince William County, 43%. Fairfax County, 11%. The smallest of those numbers is twice the rate of growth of the District.
by Bertie on Dec 7, 2011 1:31 pm
Yes, privately funded. Except for those land grants, and the extension of eminent domain powers to the railroads, and the financing help, etc.
Other than that, completely private.
by Alex B. on Dec 7, 2011 1:32 pm
oh dear me no. There was abundant state financing, including public private partnerships, etc. Than the big transcontinentals were subsidized with major donations of public lands (not land to build on, but additional land for the RRs to sell, to help subsidize construction)
"massive public subsidies or public works projects (sports stadiums, arts centers, convention centers, transit hubs, etc"
as has been discussed here, theres little evidence that stadiums make that much difference = I dont know convention centers do much better.
by AWalkerInTheCity on Dec 7, 2011 1:37 pm
that suburbs in a fast growing metro are ALSO growing does not show there is not a movement to the high density parts of the DC (note DCs growth is slowed by departures from less urbanist parts of DC). Note Fairfax is now pursuing high density urbanism to maintain its growth. PWC - well Im not sure how theyve done since the bubble popped - much of that growth may not have been sustainable.
by AWalkerInTheCity on Dec 7, 2011 1:39 pm
Eminent domain is a necessity for any real physical network to ever exist, and is not public funding.
Now the first telegraph line... that got a subsidy.
What specific examples of large government financial support for railroad construction can anybody provide?
by Robert on Dec 7, 2011 2:03 pm
Um, the entire history of American railroading?
How about the massive land grants given to railroads? The financing of Transcontinental lines via the issuance of US Government Bonds?
by Alex B. on Dec 7, 2011 2:11 pm
stay classy
by ed on Dec 7, 2011 2:12 pm
I suggest reading this:
http://www.amazon.com/What-Hath-God-Wrought-Transformation/dp/0195078942
by oyhewont listen on Dec 7, 2011 2:17 pm
1. given at a time when the US was already giving land to settlers outright, because it was so abundant and therefore cheap
2. land that was of extremely low value without the presence of the railroad. The railroad is what made that land desirable and therefore valueable
3. miniscule in comparison to the private investment which built the vast majority of the US railroad network
The fact is, US railroads were overwhelmingly privately funded, (as were US pipelines.)
But we digress. The topic is self driving cars, and the reality is that they are just not that big of a deal. The next revolutionary innovation will involve transportation, but self-driving cars is not it.
by Robert on Dec 7, 2011 2:27 pm
http://www.amazon.com/What-Hath-God-Wrought-Transformation/dp/0195078942
by Robert on Dec 7, 2011 2:30 pm
I didn't say it did. There is obviously SOME movement to some parts of DC. But no one has produced any evidence of even a small net population increase in DC from domestic migration (that is, more people moving in than moving out). What we do know for sure from the 2010 Census is that DC has grown much more slowly than its suburbs. And this is the general pattern across the country. Cities are either growing more slowly than their suburbs, or are actually losing population.
by Bertie on Dec 7, 2011 2:33 pm
Chicago, for example, experienced significant growth in its downtown between 2000 and 2010, but the city overall lost 200,000 people.
Again, the number to look at is number of households and median household income, not raw population. When urbanists use the shorthand "the return to the suburbs" they mean the gradual replacement of poor households with middle-class households.
As an example, I was talking with our mailman--who has served our neighborhood for about 25 years now. He had long since moved to suburban PG county, but had stories to tell of all the 800 sq foot rowhouses that once had 10 children living in them. Now they've got a couple. Or one person. Or a family of three. And the median household income numbers reflect that:
[Chicago] Estimated median household income in 2009: $45,734 (it was $38,625 in 2000)
(http://www.city-data.com/city/Chicago-Illinois.html#ixzz1fscuxZ3T)
As far as the population numbers go, one could argue that's a function of Chicago being in the Midwest. The population of Grosse Point fell by 14% over the 00s. I wouldn't presume to argue that's proof of a national exodus from the suburbs.
by oboe on Dec 7, 2011 3:14 pm
[C]ities have natural advantages but the failure to develop honest and efficient institutions has meant that cities have not leveraged those advantages well.
I think political dysfunction is a function of poverty. For all the drama about individual actors in DC politics, the functioning of its government has mapped pretty well to its median household income growth.
(Now I'm going to duck and cover.)
by oboe on Dec 7, 2011 3:38 pm
and the land was NOT valueless. It was saleable and was sold. The decision to give it away free was a political decision, to subsidize settlement, and was controversial at the time.
Of course the presence of RRs added to the value. One could say something similar for contemporary infrastructure projects
@oboe and falls church re urban governances. yes, there have been rural political machines in the past. The progressive rebellion against them was largely a function of income and education (except in some places like NYC where an ideological segment of the working class (educated in its own way) joined that rebellion.
by AWalkerInTheCity on Dec 7, 2011 3:45 pm
This decade should see the first comercial autonomous car. I assume the first generation will be an expensive, high-end add-on. The next generation will be more practical, the next generation commonplace in all but the cheapest models.
If you figure ~10 year replacement time for the average car, that puts autonomous vehicles commonplace by 2040 or 2050.
The biggest implications for planning are going to be land use, and I would argue today AGAINST expending capital to expand freeway or parking capacity, given that the demand for this capacity will likely be drastically reduced in the medium-term, certainly during the lifetime of the infrastructure in question.
by Daniel Howard on Dec 7, 2011 10:20 pm
by Robert on Dec 8, 2011 8:36 am
Again, income data is NOT evidence of middle-class migration from suburb to city. The entire DC region has boomed over the past decade, and incomes have risen dramatically in both the District and the suburbs. The highest median household incomes are in the suburbs, Loudon County and Fairfax County, not in DC.
If you have migration data, then show it to us. If you don't have it, stop claiming there's a trend of migration into the city.
by Bertie on Dec 8, 2011 11:58 am
The discussion in it is quite thorough. State involvement in infrastructure finance, including RRs, was quite common and led, IIUC, to some state bankruptcies in the panin of 1837.
Are you accusing me of lying?
by AWalkerInTheCity on Dec 8, 2011 12:04 pm
I believe the largest INCREASES in median income were in Loudoun and in DC.
some surbuburan jurisdiction had declines or slower increases.
by AWalkerInTheCity on Dec 8, 2011 12:05 pm
http://en.wikipedia.org/wiki/Indiana_Mammoth_Internal_Improvement_Act
http://books.google.com/books?id=8VaqAAAAIAAJ&pg=PA77&lpg=PA77&dq=ohio+internal+improvements&source=bl&ots=ZiMoPZsUCe&sig=KoNIeOjlTNhXCxNkh9-ytXS69sg&hl=en#v=onepage&q=ohio%20internal%20improvements&f=false (see page 80)
by AWalkerInTheCity on Dec 8, 2011 12:13 pm
Unlike most people here, I'm skeptical that the self-driving cars will take off any time soon. I know some people involved with Spirit and Curiosity, the Mars rovers and they talk about how difficult the autonomy is to rely on. But I recognize all the advantages of them.
I absolutely disagree that it means that people will want to live in the suburbs, because it ignores the fact that time=money. And it will still take time to get to places. In fact, if we assume that public transit disappears as do biking and walking, that means everyone is driving. For two employees who work in the city, one who lives in the city and one who doesn't, who gets to work first?
Also, it's not like transportation will be free. So there will be another connection between distance and price. That will be especially true of a car you don't own, since you'll have to pay for maintenance and depreciation by the mile.
So living close to work - and close to the things you like to do - will carry advantages until we invent teleportation.
by David C on Dec 8, 2011 12:56 pm
robocars.com/urban-plan.html
by Brad Templeton on Dec 8, 2011 1:13 pm
Not true either. Over the past 10 years, median household income increased by about $18,000 in DC and about $30,000 in Loudon. The cost of living index is higher in DC, too, so the effective difference is even greater. If we are supposed to assume, as oboe wants us to, that income increases are a reliable proxy for middle-class migration, then this data means that the middle-class is migrating from DC to Loudon, not the reverse. But I don't think income changes are a reliable measure of migration, anyway.
by Bertie on Dec 9, 2011 3:28 pm
Huh? Self-driving cars mean that more people will be able to travel greater distances in a given period of time. This is particularly true for transit users, who will be able to switch to self-driving taxis. Transit is almost always slower than cars. This means the time cost of travel will be reduced. The financial cost will be reduced too. Faster transportation has always meant lower urban densities.
by Bertie on Dec 9, 2011 3:36 pm
It seems pretty simple to me Bertie. Even if travel speeds increase, places that are close together will still be closer together and places that are far away will still be far away. So if the choice was between a 10 min commute and a 40 commute and, now speeds increase by 10%, it will still be a choice between a 9 min commute and 36 minute commute. Not to mention that if you're paying by the mile, it's more expensive to live far away.
There are two contradictory forces at play here - 1) People don't want to spend less time and money on transportation 2) people want to spend less money on housing. One pushes them away and the other pulls them in closer. Even if (1) becomes cheaper, it doesn't equal zero.
by David C on Dec 9, 2011 3:47 pm
you seem to assume loudoun and DC are the only two jurisdictions. what I am saying is that both had income increases faster than other jurisdictions (like FFX, MoCo, and PG) and perhaps they are both getting migration from those counties. Whether they actuall are, whether its simply that both are getting better at drawing high income transplants from elsewhere, I dont think matters much.
I very much doubt its that LC and DC people are getting raises while MoCo people are not
by AWalkerInTheCity on Dec 9, 2011 3:59 pm
Obviously, travel time will still increase with distance. But the point is that people will be able to travel greater distances in a given period of time than they can today. A given cost in time will allow people to travel further, which will reduce the (already low) incentive for dense urban development with short distances between destinations. It'll be the same effect as the transition from horse-drawn vehicles to motorized streetcars, then to buses, then to private cars. Each transition produced an increase in average urban travel speed and led to a reduction in urban density.
In addition, since self-driving cars will allow their occupants to use their travel time more productively than today's cars, the effective cost of travel time will be reduced.
by Bertie on Dec 9, 2011 4:24 pm
IF,
the cars can truly reduce congestion - I'm skeptical
induced demand, longer driving, more people and a higher percentage of people using cars (as you suppose) doesn't suck up all of the added benefits - the arc of congestion bends towards heavier, so it will take something truly Herculean to get us back to even 2005 levels. Whatever benefits you suppose will almost surely be dwarfed by the natural increase in congestion
they can find a place to park all of these cars
there is universal adoption of this technology
Then, some marginal group of people who want to live in the suburbs now, but feel that they can't (if there are such people) and feel like they could if the suburbs were just a few minutes closer to work will likely do so.
But that is a lot of ifs, and a very specific slice of the population.
led to a reduction in urban density
I believe most cities have gotten denser, not less dense. Things have slowed down a bit in the late 20th century, but the population density of NYC in 2000 was at an all time high. And it was in 2010 again.
by David C on Dec 9, 2011 4:48 pm
No, as I already explained, self-driving cars will reduce travel times through other means besides congestion reduction. At off-peak times, there is plenty of unused road capacity in almost all parts of almost all urban areas. Therefore, current transit users would be able to switch to self-driving taxis for off-peak travel without any increase in congestion, even if self-driving vehicles had no congestion-relieving effects. These taxi trips will generally be MUCH faster than the equivalent bus or train trip, because they will eliminate all the walking and waiting involved in using transit.
I believe most cities have gotten denser, not less dense.
Between 1950 and 2000, the average population density of American central cities declined enormously, from about 7,500 people per square mile in 1950 to about 2,700 in 2000. See Demographic Trends in the 20th Century
As for New York City, although the total population has increased, the population has become much more evenly distributed throughout the five boroughs, rather than being overwhelmingly concentrated in Manhattan. Manhattan's population density peaked in 1910, when it had almost a million more people than it does today. None of the other boroughs has ever come close to this density.
[if] there is universal adoption of this technology
Again, no. Self-driving cars will start to produce the benefits I have described as soon as they start to appear. The benefits are not all-or-nothing. They're gradual and cumulative, although I suspect there will be a series of "tipping points" from non-linear effects.
The decline of transit systems is also likely to be non-linear, They'll be able to survive some defection of passengers to self-driving taxis and still keep operating. But there will come a point at which there aren't enough remaining transit users to justify the huge fixed costs of a large transit system, and the system will have to shrink dramatically, if it can survive at all.
by Bertie on Dec 9, 2011 9:30 pm
That may be true of the smaller cities and more sprawling cities in America, the Little Rocks and Phoenixes and Raleigh-Durhams. However, because of politics, density, the desire of some people to live in dense urban areas (and, yes, Bertie, there IS a preference among some American people to live in dense urban areas, even if others prefer the suburbs), and habit, transit in the traditional sense will survive in the larger and denser urban areas. I can see anywhere with hourly bus service and one or two passengers per bus switching to robocars or robo-vans, while the lines with well-used systems will keep their large vehicles and use smaller vehicles in the off hours. The systems will adapt, rather than shrink dramatically; they will buy robotaxis themselves and use them on their old lines, if they have to. No dismantling of transit stations needed - just the oversized buses in places that didn't need them anyhow.
by Rickyrab on Dec 10, 2011 3:50 am
Manhattan in 1910 was overpopulated. As for the outerboroughs, they obviously gained in density themselves at numerous times in the past century. New York's density patterns have fluctuated over the years: Manhattan gained and peaked in 1910 due to the immigrant wave that was funneled into Manhattan through Ellis Island from less dense shtetls and towns in the old country. Then there was a decline to more sensible levels through transit provision. After that, not only did the automobile kick in, so did racism. Old neighborhoods were "redlined", making insurance and mortgages impossible to get in those neighborhoods, and whites got spooked by black rioting and crime (or white police rioting against blacks, take your pick), leading to white flight. "Busing" didn't help. The Bronx flatlined and abandoned buildings could be seen in places like Brooklyn and Harlem, as well as elsewhere. After this bad period came the revival. Giuliani put police on the streets. Immigration was flowing into the city and other places from other countries again. Gentrification set in, and New York City was suddenly seen as desirable. A cleaner mass transit system helped, and people started flowing in again. So NYC saw a population increase and a density increase. Lower Manhattan gained population after 9/11, thanks to aid intended to help that district recover from the attacks. (In fact, due to construction of Battery Park City, lower Manhattan had been gaining population in at least one location even before the attacks.)
In places where sprawl is the custom, people will find more and more reasons to save open space, as the farm and countryside they drive (or robocar, for that matter) through diminishes. This is an issue that is more evident to the motorist than to the transit taker, because transit is more of an urban mode of transport. Thus, one can see more desire for growth boundaries and other limits to sprawl, ultimately limiting its extent. Some jurisdictions will reach full build-out and infill land will become more common. Densities will thus rise in some areas.
by Rickyrab on Dec 10, 2011 4:07 am
Robocars might help increase density in some areas by freeing up some parking spaces (if people are able to rent out their cars in massive carshare operations, anyhow). In such a case, salable land might become free, and people might build on that land - and that might increase density. That, in turn, might lead to clutter and traffic among robocars, even with the higher robot reaction times. (Robocars might also jerk more because of those higher reaction times - starting more quickly, stopping more quickly, etc.)
by Rickyrab on Dec 10, 2011 4:15 am
The vast majority of urban space in the United States is much more like Little Rock, Phoenix and Raleigh-Durham than like lower Manhattan. That is, most urban space is already low-density and car-oriented. Self-driving taxis are therefore a natural fit in most urban areas and will completely, or almost completely, displace buses and trains.
But even in Manhattan and other high-density areas, there is plenty of unused road capacity at off-peak travel times. Robocabs will be able to exploit this unused capacity to poach off-peak transit users. And even at peak travel times, robocabs will be able to expand capacity through the mechanisms that have been described. I suspect that eventually, when almost all cars in Manhattan are autonomous, and the current absurdly over-engineered NYC taxi fleet (full size sedans and SUVs) has been replaced with Smart Car-sized robocabs, Manhattan's roads will have more than enough capacity for all current road traffic plus all current rail users, and there simply won't be any subway or commuter trains left at all.
by Bertie on Dec 10, 2011 5:50 pm
"I suspect that eventually, when almost all cars in Manhattan are autonomous, and the current absurdly over-engineered NYC taxi fleet (full size sedans and SUVs) has been replaced with Smart Car-sized robocabs, Manhattan's roads will have more than enough capacity for all current road traffic plus all current rail users, and there simply won't be any subway or commuter trains left at all."
I suspect that you're wrong on this. Autonomous cars or not, it isn't possible to match a subway's throughput using separate cars on a network of roads.
by Gray on Dec 10, 2011 6:04 pm
Why? Americans have been suburbanizing and sprawling for 70 years. Why should we expect this to stop in the foreseeable future? Robocars will make car travel even easier and available to even more people. So they're likely to promote sprawl, not inhibit it. It's not like we're running out of land. 97% of the land area of the United States isn't urbanized at all. It's still rural.
by Bertie on Dec 10, 2011 6:07 pm
There are numerous places where land is running out, mostly in the more urbanized areas. Yes, 97% of US land may still be rural (or wilderness), but how much of that wilderness is easily accessible? How much is in people's backyards? There is a reason why open-space preservation measures have passed in the past - because open space within earshot of people has been disappearing in some places. And robocars might decrease the need for parking spots, making it easier to group land uses together and thus produce denser areas.
Prediction: Sprawl inhibited in some areas, promoted in other areas.
I suspect that eventually, when almost all cars in Manhattan are autonomous, and the current absurdly over-engineered NYC taxi fleet (full size sedans and SUVs) has been replaced with Smart Car-sized robocabs, Manhattan's roads will have more than enough capacity for all current road traffic plus all current rail users, and there simply won't be any subway or commuter trains left at all.
I'm not so sure. Overengineering is still a problem, for one, because people like legroom and having space. Furthermore, Manhattan is an incredibly crowded place. A robotaxi, if it is to serve the people in Manhattan, will need to wait for other robotaxis to get out of the way in some areas. Subway lines are within easy walking distance of the people. There'll need to be places for taxis to stay when not being used by passengers. And people are used to the communal vehicles anyhow. The NYC subway will remain.
by Rickyrab on Dec 10, 2011 6:23 pm
So where are the Segways? Plus, smart car sized robocabs have the same failing that present taxis do: they aren't wheelchair accessible and they're cramped.
by Rickyrab on Dec 10, 2011 6:26 pm
Of course it is, as long as you have enough roads. The comparison isn't between a subway line and a single one-lane road. It's between the subway network and the road network. The entire New York subway system has only a few hundred miles of track, and much of that is outside Manhattan. Manhattan has thousands of lane-miles of roads.
But perhaps you're right about capacity. Perhaps even with self-driving vehicles we'll run into a hard physical limit on road capacity in Manhattan that means some amount of rail service will still be required. It'll still be much less than there is now.
by Bertie on Dec 10, 2011 6:30 pm
Where are these numerous places where land is running out? Even the New York urban area has vast areas of land that are mostly empty. That's why New York's urban density is actually less than that of the Los Angeles urban area, despite its reputation for being dense.
And if people "run out" of land in one part of the country, they can move to another. There's been a massive migration of people from the northeast to the south and west. That's almost certainly going to continue.
Yes, 97% of US land may still be rural (or wilderness), but how much of that wilderness is easily accessible?
Vast amounts of it. Haven't you ever been west of the Mississippi? There are vast areas of land suitable for development.
And robocars might decrease the need for parking spots, making it easier to group land uses together and thus produce denser areas.
Robocars will definitely reduce the need for parking space, but that isn't likely to lead to higher density. Robocars will allow more people to travel greater distances in a given period of time. That reduces the incentive to build things close together.
by Bertie on Dec 10, 2011 6:48 pm
And a lot of that "empty land" is in the form of parks or farmland that people already want to preserve.
Robocars will definitely reduce the need for parking space, but that isn't likely to lead to higher density. Robocars will allow more people to travel greater distances in a given period of time. That reduces the incentive to build things close together.
If urban planners let it reduce such incentives, it might. However, the greater distance traveled can reduce the impact of the higher fuel efficiency of hybrids, and that, in turn, would limit the effect of travel cost reduction. Because there is a market for living in dense areas, there will continue to be dense areas and those dense areas will get denser. Look at gentrification around downtowns. Sure, inner city slums might continue to get less dense, because they're slums, but the downtowns will continue to hold and some might grow. Look at Manhattan (which has bounced back from a lower point), Miami (which has Manhattanized), and other downtown areas which have gotten denser, even as their suburban outskirts have sprawled. There is a loser here - older residential inner city areas - but it's not city centers.
Robocars will allow more people to travel greater distances in a given period of time.
The same can be said about airplanes and transit. And they have spread out urban areas, true, particularly rail transit in the 19th and early 20th century. However, if, as you posit, robocars will mostly be owned by their users, then the marginal impact in suburbia (where most of the auto ownership is anyhow) won't be all that great. The impact on denser areas will probably be greater, and transit will be made more useful because of the "last mile" problem being solved.
The "last mile" refers to the need to get from a suburban transit station to a suburban destination when your workplace is at the other end. Right now, some people use park-and-ride in places with strong urban centers. A lot of those users live near transit stations and wouldn't mind taking transit from other stations to their workplaces if they could navigate the "last mile". Now, perhaps robocars might be more convenient in going door to door, but those cars (and other roboservices) might find it convenient to drop people off at the stations, have the transit line deliver the user to another station, and have another robocar pick them up. This will particularly be true of higher density corridors which a lot of people fan into from, say, a bunch of exurbs, or for jobs which are far from where the users live. Infrastructure is expensive, and minimizing it is a good idea. That is an incentive to build things close together.
by Rickyrab on Dec 10, 2011 7:14 pm
And a lot of it isn't. It's just land that hasn't been developed yet. We're nowhere close to running out of land suitable for low-density urban development.
If urban planners let it reduce such incentives, it might.
The incentive has nothing to do with urban planners. It's a consequence of the fact that robocars will allow more people to travel greater distances in a given period of time. That will promote lower-density development. Just as the appearance of the original motorized streetcars and subways promoted the development of the first inner-ring suburbs, then commuter rail and motor buses promoted expansion into more distant suburbs, then the mass adoption of private cars promoted even further expansion into outer suburbs and exurbs.
Because there is a market for living in dense areas, there will continue to be dense areas and those dense areas will get denser
The market for dense development will be smaller, because fewer people will be dependent on mass transit for urban travel. With no transit, or with very little transit, there'll be little or no "transit-oriented development."
Infrastructure is expensive, and minimizing it is a good idea. That is an incentive to build things close together.
But that's irrelevant. The question is what effect robocars will have on incentives. Robocars won't make building infrastructure more expensive. In fact, they will reduce the cost of building low-density, car-oriented infrastructure by increasing road and parking capacity. With robocars, we'll need fewer lane-miles of road and fewer acres of parking to support a given amount of car travel. They will also reduce the cost of longer trips by shortening travel times and increasing travel time productivity. Each of these effects -- cheaper infrastructure, shorter travel times, and more productive use of travel time -- will reduce the incentive to build things close together.
by Bertie on Dec 10, 2011 10:48 pm
And shorter distances will remain cheaper than longer ones. You still need material to maintain infrastructure, and shorter distances still use less fuel than longer ones, lead to less wear and tear, etc. That's not going to change. So there'll still be some incentive to build close together, even if there is a reduction in such an incentive from that factor.
Some might even say there'll be a growth in such incentive, because the increased road capacity will lead to an ability to pack more people into an area with existing infrastructure. So land costs might rise in desirable areas, leading to capitalists building apartment buildings, leading to density growth. So the robocar, for all its grandness, might not change very much of anything, except parking lots will be smaller, dense areas denser, suburbs less dense, and inner cities still in a questionable status.
With no transit, or with very little transit, there'll be little or no "transit-oriented development."
Transit isn't just about saving fuel those days, Bertie, or about "getting people out of their cars". It's about a sense of place, about real estate, about creating a perk for people to live nearby. There'll be transit oriented development all right. Why do you think they build streetcar lines in a world where buses are more flexible?
Suppose transit kicks the bucket, as you forecast. Then someone will find a different rationale for building dense neighborhoods. "Walkable neighborhoods" and "bicycle villages" come to mind. So long as a market is there for that real estate, that would continue. And planners and developers being who they are, transit's likely to come into the equation, even if "nobody" rides it. People understand it and they see it, and some even expect it to be there. So some might ride it for fun. Perhaps someone's going to make a store or a funfair ride out of it. Who knows?
And let's face it, how many BTU would you burn per passenger mile on a bicycle?
by Rickyrab on Dec 11, 2011 3:59 pm
You keep repeating this irrelevant observation and ignoring the point. It doesn't matter that there will still be a cheaper option. The point is that LONGER DISTANCES WILL BECOME CHEAPER THAN THEY ARE NOW. If the price of apples goes down, apple sales will probably rise, even if oranges remain cheaper. If the price of longer trips goes down, people are likely to take longer trips. Self-driving cars will make the price of longer trips go down, especially for people who currently depend on buses and trains. They'll be able to travel two or three times as far in a given amount of time than they can today, and in greater comfort, and with greater convenience, and with a greater ability to carry shopping or other cargo with them, and so on.
Transit isn't just about saving fuel those days, Bertie, or about "getting people out of their cars". It's about a sense of place, about real estate, about creating a perk for people to live nearby. There'll be transit oriented development all right.
"Transit-oriented development" without transit is an oxymoron. You can't orient development to a form of transportation that doesn't exist where you're developing. Even if there are a few places where some form of mass transit survives, and some form of TOD therefore survives, it will be even rarer than it is today.
by Bertie on Dec 11, 2011 6:01 pm
Perhaps. First, the cars actually have to get to them, and then they have to tell the cars where to go. And then the cars have to find places to pick them up and drop them off. That takes up space. And when they have stuff they want to carry and want legroom and elbow room, that'll also take up space. Now, imagine an area with compact bus stops where people currently wait for the bus. If several people want the robocars at the same time, to, say, go to a movie, a bunch of appointments, or a game, that'd cause overcrowding and the need to scramble for taxis, which would be better solved with a bus or other compact vehicle (if people are going in the same direction).
There are some transit lines that use more direct infrastructure than highways, simply because the railroads got to use the land first. With robocars, those railroads get to extend their reach, true, but they run into the same compactness problem. So what to do? I suspect a transit vehicle type that is a combination of robocars might be the solution. Even more elegantly, such a vehicle might be able to take to the railway and run express to the other end of the line, get off, and deposit people directly at a compact destination.
While there may be more roadway than railway, there are still places rails reach better than roads, and so a dual-infrastructure solution is advisable in some areas. Furthermore, if several people are taking a long trip to a common destination or area, it'll benefit them to combine.
Self-driving cars will make the price of longer trips go down,
um, how would this be true for normal cars? Anything with a chauffeur might be cheaper, if the robot is cheaper than labor. However, in normal driving, we already do our own driving, so we're not really saving by automating. Sure, insurance costs will probably decline. But that is not a per-trip cost, that is a per-time-period cost, which exists regardless of whether or not you travel. One can turn it into a per-trip or per-mile cost, and so short trips would continue to be cheaper than long trips.
Yes, long trips may get cheaper, and so will transit in places that it survives in. Why? Because transit drops the routes nobody rides on and focuses on its strongest routes!
Taxis drop more in cost than buses
Yes, perhaps, but would costs be cut to lower than buses? They might, but if bus routes get dropped in places where they don't matter very much anyway, this is more of an iffy question. Ultimately, the van or the jitney will become much more popular than it is today, while the big bus fades from many areas, and all modes of transportation will continue to get more efficient, as new materials and more efficient hybrid motors come into play on all fronts (not just the automotive). Not all railroads are hybridized, and not all locomotives have converted to electricity.
Even if there are a few places where some form of mass transit survives, and some form of TOD therefore survives, it will be even rarer than it is today.
Oh, it'll survive all right. And people may well be calling it "bicycle-oriented development". :)
by Rickyrab on Dec 11, 2011 8:48 pm
Item 1 says asphalt is much cheaper than steel, and abandons a maglev PRT idea (which was rightly abandoned) because it couldn't find a niche where maglev PRT would fit. Population densities under 5000 people per square kilometer went best with robotic cars, while population densities above would see competitive transit and robot car use (and perhaps transit dominance).
Item 1: http://swiftprt.com/blog/2011/12/the-future-of-ground-based-transportation-systems
Item 2 suggests robot taxicabs might have a problem competing with buses or bus transit, but would be more energy efficient than heavy rail. It looked at the full lifecycle of transportation systems to come to this surprising conclusion, but, even then, it noticed that "dial-a-ride" (i.e., paratransit) was the least efficient mode of transportation. Since robocabs would essentially fill the niche of paratransit and taxis, except perhaps more efficiently, this is a concern. Carpools and vanpools were listed as being more efficient.
Source 2: http://www.cbo.gov/doc.cfm?index=5301&type=0
The inefficiencies of rail were due to grandiose train stations and park-and-ride use (the driving of automobiles to the train stations). Hybrid robot car use might cut down on one of those inefficiencies; simplifying train stations would cut down on the other. Of course, this being a 1977 report, the facts may be off a bit. But it DOES explain all the focusing on bus transit in the 1980s as part of transit spending.
by Rickyrab on Dec 12, 2011 12:39 am
by Rickyrab on Dec 12, 2011 1:02 am
by Rickyrab on Dec 12, 2011 1:06 am
by Allen on Dec 12, 2011 9:58 am
Regarding the cost of auto transportation being cheapest it's ever been...I won't debate the point. Going forward however, there's a broad realization that the gas tax hasn't kept pace with transportation needs. How does this technological innovation (self-driving cars) interact with some form of broad-based (and likely more expensive) distance-based VMT tax (another technological innovation)? Do the two in combination fundamentally change the economics of vehicle travel? Seems the free time/productivity alone might be the game changer, if we ever get there.
by AAmey on Dec 12, 2011 3:26 pm
by OriginalGeoff on Mar 6, 2012 6:17 pm
Add a Comment