If Uber works, you might be less happy with your transport options

Let’s say you are optimistic about the long-run prospects for Uber to transform society, as I am.  Counterintuitively, that could mean the perceived quality of an Uber ride goes down.

The better Uber gets, the more people can do without cars and the more taxis will go out of business.  In contrast, we are currently living in a world where there is excess capacity of automobiles, especially for short-term rental.  The stocks of taxis, rental cars, and personal automobiles mostly have been determined by past considerations which did not take the existence of Uber into account.  Those stocks will decline over time as idle vehicles are used more efficiently.  There will be fewer circulating vehicles and perhaps the price of a short-term ride will be higher, once the excess capacity disappears.  Waiting time might be higher too.

(Admittedly these results could go in either direction.  Think of this as being a race between two adjustments: cutting back on the stock of vehicles, and using the extant stock of vehicles more effectively.  Arguably we already have reaped the latter economies, at least short of driverless vehicles, but with time the former adjustment will kick in more and more, and so the flow of “available vehicle time” into the market already may have peaked.)

That is precisely the scenario where Uber is probably doing the world the most good.  Resources put into automobile production can be diverted to other purposes, with potential environmental benefits as well.  The beneficiaries would be consumers in non-ride industries, who will enjoy lower prices and better selection.

Yet the lower consumer surplus from the Uber experience may decrease the political popularity of the service.  More of the social benefits would be invisible to riders and instead dispersed across a wide range of consumers elsewhere.

Here are related comments from Izabella Kaminska.  It is an interesting piece but I do not agree with her conclusions, and you can think of this post as an attempt to state what I think she should have argued.


I can see Uber compete against overpriced Taxis very well but does Uber really compete with short-term rental?

A local agency rental car costs $25 for 24 hrs. Can Uber really compete with that segment?

Well, you know how to drive and you have a driver's license. What's the value of both?

Good point. In a generation, the only people who know how to drive might be people who do it professionally. And they'll band together and lobby the state to make the driving test harder. Maybe put a hefty fee on getting a license.

The public won't care, because they won't be driving. They might even support the tougher requirements as a safety feature.

I doubt that will happen based on human-driven cars from Uber and its competitors for 2 reasons:

(1) If you are a person who drives on a daily basis, the cost of using Uber will still be high relative to the cost of driving your own car. That will remain the case.

(2) There will still be lots of places outside major metro areas where obtaining an Uber ride will be borderline impossible due to time and expense.

Self-driving cars, on the other hand, could dramatically reduce the percent of the population who know how to drive.

Learning to drive is easy--takes on the order of 100 hours. It would be hard to decide to pay someone else for 1000s of hours when you could do it yourself for 100. If our current inequality continues to worsen, then the elites could justify it, but you'd have to be making 10 times as much as the driver, even discounting a bit for the fact that you could do something more worthwhile than driving during all those rides.

Well, sure, because in most cases of Uber employment, the person was not thinking of a short-term rental instead. For instance, in NYC, Uber is often used as an alternative, after a night out drinking, to:
* Waiting on the street corner for 30 minutes or more to get a yellow cab late at night; or
* Waiting at a subway station for 30 minutes or more for a late night train.
This bar goer is definitely not thinking of renting a car as an option!


That's why I'm surprised when Tyler mentioned short term rentals. I see Uber as competing with Taxis. Not car rentals.

It is competitive for some classes of Uber rides. For instance, I'll generally rent a car from the airport when I go somewhere if it's cheaper than a taxi/uber (including parking at my destination). But a taxi or uber could be competitive if the price were lower.

Unless the place I've to go to is real close to the airport I never find a taxi / Uber turning out cheaper than a rental car. That has been my experience.


Yeah, usually rental car wins. But I've had taxi win, e.g. in Las Vegas if you're going to the strip a taxi is usually a better bet (as long as you don't get scammed - which is why Uber is so good).

You guys! A rental car also costs you at least an hour, in and out, with a pain-in-the-ass shuttle to boot.


Depends on airport. I fly into YYZ a lot, rental cars are right next to terminal and I have an acct with Hertz where I just go direct to car.

It's easy to see the Uber vs short term rental relationship. I'm currently in San Diego for eight days with a Toyota Camry rental which spends over 90% of its time parked. This inefficient arrangement cost me $350. I did not do the numbers in detail, but using Uber might have saved me money. At some point the numbers will be widely disseminated and understood and a rule of thumb might emerge where on future such trips I Uber it.

It doesn't compete directly, but it competes indirectly. Remember, Lyft offers rental cars to drivers (see bottom), drawing on the stock of idle rental cars[1]. A greater supply of unused rental cars would thus increase the driver supply.

So that would support Tyler_Cowen's point that rental car slack affects the cost-effectiveness of Uber/Lyft.

[1] Personally, I think these programs are insane. By my rough calculations, driving for Uber/Lyft only makes sense if you had good enough reason to buy the car anyway and it would otherwise go idle. Having to rent the car, at high daily rates, in order to drive, would wipe out most of the profit.

You mean, cars can be driven a million miles per year for the same capital cost of driving it 12,000 miles per year?

The IRS mileage rate is 56 cents per miles which is the full capital and operating cost for a vehicle that would meet taxi standards. That rate is derived from data from the government's own fleets, as well as data on corporate fleets.

If you drive just to work, and occasionally drive to the shop and get a rid to work while your car is worked on, you look at that service as costing you only for the investment into the car for repairs and service, but if you use your car for business, you will need to forego work, or rent a replacement at a hefty out of pocket cost.

From reading Uber driver comments, it seems many drivers do not cover their capital costs with the fare they get after Uber takes its rent, and they work for free. Uber can provide a way to turn past savings stored in your car into cash - you had a great job and bought a great car with cash, but now you are unemployed and need to withdraw your savings deposited in your car.

And if you hunker down and spend as little as possible, meaning you drive your great car only 2000 miles a year for three years, the fall in price for your used car might make those 6000 miles turn into a capital cost of $2 per mile, with the low mileage becoming a virtue in 20 years or 50 years depending on how "great" is was.

"I can see Uber compete against overpriced Taxis very well but does Uber really compete with short-term rental?"

You have failed to read what Uber says it is!

Uber provides no transportation!
Uber provides no transportation capital!
Uber provides no transportation labor!

Uber is purely a rent seeker taking a 10% cut off your payment to labor and capital which is totally outside Uber's control or responsibility - the independent contractor over which Uber exercises zero supervisory control of any kind - that's why it is completely independent of Uber.

Uber is competing with the telephone. The street and your arm. The taxi stand. The doorman.

The taxi stand at the airport charges taxis to pick up fares, which is lower than what Uber charges except no one provides the same service in other places. But in those places, you can call a taxi service. The problem in many places is "which one"? In some cities, a taxi exchange exists, a co op of sorts that dispatch independent taxis to callers. But for frequent regular customers they call directly to eliminate the middlemen.

In reading comments by Uber drivers, those who do ok, use Uber and Lyft to get customers and fill time when otherwise idle because the 10% rent to Uber is too high for the value they provide.

>Uber is competing with the telephone. The street and your arm. The taxi stand. The doorman.

When I call a taxi on the telephone, they rarely show up. When they do, it is usually in a filthy, broken-down vehicle being driven by someone who is screaming over the phone in a foreign language. Uber vets its drivers reasonably well, something that taxi companies do not do.

Street, arm, taxi stand, doorman? Do you think everyone lives in New York?

Maybe that's why my perception is so different. I'm in a smaller city & even the phone taxis come clean, punctually and with none of the problems Larry refers to.

For me the only dimension Uber could compete with my existing taxi option is purely on cost.

Yes they do. I live in Manhattan and don't have a car. I do have a drivers license.

My options are basically ZipCar, Uber, or a Hertz/Enterprise/Budget rental car. Every trip yields a different decision depending on the circumstances. ZipCar is best for a few hours, but a key limiting factor is that the car needs to be returned back to where you picked it up from. Rental cars are great for a whole weekend. Uber is good for point to point.

This played out last weekend when we had to get from New York to Philly. We had a ride back so it was a one way trip. Thus it was actually cheaper for 3 people to pay an Uber to drive us from New York to Philly than to take a train or rent a car. It was about $150, which was roughly half the price of Amtrak tickets and about $100 less than a rental car.

Plus having to park

$25 is really cheap. I am not sure what kind of car you get for this. Also it's the rental time charge, typically you might get in addition:
Airport access fee ,Customer Facility Charge , rental tax ,State Tax, Vehicle License Fee . I often get the nasty surprise that various fees and taxes account for ~ 40% of the total charge

Nissan Versa 4 Door Automatic with unlimited mileage.

$25 includes 2-3 different fees surcharges and sales tax. Perhaps there are more hidden charges. Dunno.

Don't forget urban parking and just having to find your own way around...

Yeah, but think about all the illegal things one can do within their rental. Talk about opportunity costs. Great, classic Cowen, post!

Self-driving cars eliminate the problem of urban parking.

If they work out, which is still a big if, they'll be a massive force in favor of increased car ownership.

Actually Flying Cars would eliminate the urban traffic snarls too.

Sure, but they're even less likely.

Where? I would expect to pay nearly twice that with fees and need to make my way to the local airport for pickup.

I rent perhaps half a dozen or a dozen times a year, and usually pay between $25-40 a day (plus gas) all in, with price depending on if it's a busy day.

Yes, that sounds right. I end up paying in the same ballpark too.

$25 is really cheap. I am not sure what kind of car you get for this. Also it’s the rental time charge, typically you might get in addition: Airport access fee ,Customer Facility Charge , rental tax ,State Tax, Vehicle License Fee . I often get the nasty surprise that various fees and taxes account for ~ 40% of the total charge

If you shop around this is an achievable price for an ordinary car. I recently booked a 3 day weekend in Toronto for 79 CAD including all taxes and fees, which is about $60 USD or $20/day. Taxes and fees were a good chunk of the price.

This did involve a coupon offer, but it was one advertised on Hertz' homepage, so it was fairly available to the general public.

I find that week-end rentals are significantly cheaper then week-day rental.

In the future, we'll all look back and laugh at how American families used to own their own cars.

Yes, it is a waste of resources.

Just like eating meat instead of beans.

Archie Bunker, driving his cab on the weekends to earn a little extra cash. Depressing.

But wait, if there's a cool app for it, then it must be hip and cutting edge!

American (urban) families.

Do those exist?

Suburban, affluent, retired couple here. Would happily go from owning two cars to one self-driving. Can't see owning zero.

@Lord Action: yes they do, tons of them in San Francisco I can report.

Are people really going to want to wait even 10 minutes for an Uber car every time they want to go grocery shopping? Seriously?

If I use Uber to go grocery shopping I'm pretty sure my Uber bills will rival my grocery bills.

Excellent post!

You mean: totally unrealistic post.

No, he did not mean that. You might think the post is totally unrealistic. You might be right in thinking that the post is unrealistic. But that is /not/ what Gene meant.

Uber exploits the surplus of drivers, not the surplus of cars. Look at a typical taxi fare: 75% goes to the driver's paycheck, 15% on marginal or per-mile costs (fuel, tires, maintenance), and only 10% for fixed car-ownership costs (insurance, cost of capital).

Car-ownership is even more attractive if you count insurance as a marginal cost, but that would require insurers to offer a pay-per-mile insurance product. I imagine such insurance products either already exist or will be launched very soon.

In short, until such time as driverless cars eliminate the largest part of that cost, car-ownership will remain attractive; and there will be a natural surplus of cars.

Uber exploits the surplus of investment money. Right now, it's pretty much pure wealth transfer of billions of dollars of investment money to drivers and riders.


Its like how the last two years have been a boon for online buying in India. With all the Private Equity in the sector essentially you get a subsidy on everything from potatoes to cellphones only so long as you buy them online.

"The better Uber gets, the more people can do without cars and the more taxis will go out of business."

Won't there be a large income effect from this? In other words, the beneficiaries of Uber will not only enjoy lower prices and better selection, they'll actually have more money to spend on those other things, now that they don't have to buy a car.

Good point. I take public transport (and occasionally taxis) all over the place without regard to cost because I know not having a car more than cancels that out.

This really only works in dense urban areas. Lower population density areas will have fewer uber cabs, longer response times, longer distances to get places and thus higher fares. At some point, the convenience of having a car in your garage and the length and cost of cab rides outwieighs the cost of owning a vehicle. Uber merely slightly lowers the population density at which this threshold occurs.

Yes! I don't get all the uber-exuberance over Uber. It disrupts the taxi market, sure. But taxi rides are a tiny fraction of the automobile usage in the US. Maybe Uber disrupts the rental car market (but personally I have had almost no car rentals where a Uber ride would've been an adequate substitute). That's still just a tiny fraction of what Americans do with automobiles.

Most of us own our own car because we can drive it on demand and for a very low marginal cost. When I lived in Boston I was usually glad to not own a car. But I lived and worked near the central core. Most Americans live in less dense and less urbanized settings, including me now. Imagine living in the suburbs and having to depend on Uber everytime you wanted to go anywhere or take the kids to soccer practice.

No matter how efficient Uber gets, you're still paying for the time of a human being to drive you, rather than using that time to drive yourself, so unless you're doing some highly valuable work during your Uber ride, that's a huge additional cost that Uber has compared to driving yourself.

You are missing the point. I am not always in a position to drive myself, nor do I always want to (if, for example, I am trying to catch a plane). Tonight I locked myself out of my car. I took an Uber home, 10 miles for about $15 (there was a 3 minute wait for the Uber), got the key, and took another Uber (1 minute wait, another $15) back to the car. If I had to take a taxi it would have been an absolute nightmare. Taxis generally do not come to the part of Chicago where i was stranded, and they do not cross city lines (while they're allowed to, most drivers refuse by saying something like GET OUT). Without Uber, the only realistic alternative would have been to call a limousine, which could take an hour to show up and which would have cost about $150 for the round trip.


Less wasteful drives. (You never Uber 500 meter, nor "drive around")
Vastly less wasteful cars. Uber drivers will buy the most efficient cars. Just driving comfort. No boasting nor signaling of wealth

" Uber drivers will buy the most efficient cars" Yeah, first time I used Uber, the driver had a freaking Hummer looool (ok it was in Hollywood, but still).

According to some, we are 5 years or less away from fully autonomous passenger vehicles. Since this technology looks like it will be quite cheap to implement and run compared to a driver's wages in developed nations, it should result in a significant decrease in the cost of non-privately owned vehicle transportation services. I presume these vehicles would be electric with low fuel and maintenance costs and also low insurance costs because they would rarely run into things. Because of their low operating costs it would be worthwhile to have many sitting idle outside of periods of peak demand, and so waiting times and prices outside of peak periods should be very low. This development seems likely to eliminate both taxi driving and uber driving positions.

Everyone on this blog believes that autonomous vehicles are just 'round the corner.

I find it deliciously ironic that the same cohort of commenters [including @ least one of 'les proprietors'] that gnash their teeth at central planning think that some centralised database of maps; adds/moves/changes/detours; these changing on an hourly basis @ the local level across an entire continent will confer the ability for the world's largest car fleet to drive it/themselves autonomously.

It ain't gonna happen.

Not like that.

Not only are robot cars just around the corner, but they'll be fusion-powered.

Oh no, that's forty years away. How do I know? Because fusion power is always forty years away.

Quantum computing will come first then, its only perpetually 5 years away.

Australia has robot trucks in its mines because it is cheaper than having a human drive them. That's cheaper in terms of safety, fuel efficieny, mechancial wear and tear, and wages. I don't know how long it will take to get fully autonomous vehicles ready for roads, but the same incentives exist. There are places where a self driving taxi, by operating multiple shifts, might be able to save $100,000 dolars a year in driver pay alone. That kind of money makes me think autonomous passenger vehicles is a nut that will be cracked at some point.

Ports have robotized trucks too. But that's a totally different ball game. Well mapped routes, predictable environments etc. Lot of these trucks etc. depend on tags / markers and other cues embedded in the environment.

Self driving cars dealing with pop up road construction should be amusing.

My Waze app seems to know where every cop is, so I think it can figure out where pop-up construction is too.

Most of commenters believe that if there's a marginal benefit in automation, automation is developed.

I don't know why commenters have these crazy ideas: http://www.abc.net.au/news/2015-10-18/rio-tinto-opens-worlds-first-automated-mine/6863814

I know a roboticist who works for one of the Google ABC companies, and is of the opinion that automatic cars in an urban environment are decades away. Interstate driving will happen much much sooner.

MR gets in these tizzies every now and again. Tyler went on a tear about 3-D printing presses once. Of course, people pointed out they are already used extensively by manufacturing professionals, and nobody's wife is going to care about the plastic dishes you're pressing out in the garage. Then the posts about 3-D printing dried up.

The obvious application is driver-assisting technology for OTR truckers rather than creeping through your transient hipster neighborhood in Brooklyn so you can stay glued to your iPhone, but that's probably too mundane for MR.

What kind of camera and computer system will be needed to deal with a kid darting into traffic or a deer bounding out of a yard or someone suddenly backing out of a driveway?

360 degree camera and radar/sensor coverage with instantaneous reaction is not going to be easy or cheap.

Are you under the misapprehension that present camera and computer systems are in some way inferior to humans in a deer/kid test? Because they are not.

360 degree cameras and radar/sensor coverage with "instantaneous" reaction is on test cars now, when it's not just in the showroom. Multiple makers (BMW is one) will sell you a sensor package that is specifically designed not to hit kids running into the street.

As an option, these aren't cheap, but they're low thousands of dollars to retail buyers.

The new Teslas come with a sensor suite and can do things such as automatic lane changes and park itself. It's a long way from self driving, but production vehicles are now starting to come with the type of sensors that will be necessary for autonomous driving. So it's not really the sensors, but getting the programming right that is the hard part.

It should result in significantly decreased cost of private vehicle ownership too.

Now, it's still not obvious that this is technically plausible.

>According to some, we are 5 years or less away from fully autonomous passenger vehicles.

I hate being the pessimist on this particular topic because I actually wish that it were true. But I fear that the success of self-driving tests in specific very well-mapped areas in good weather mostly on limited access roads leads a lot of people to extrapolate to universal self-driving without a human even in the car. At least some of the researchers working on these things think we won't get full self-driving for many decades and I'm inclined to agree. The transition will be interesting though because I do expect we'll have systems that work pretty reliably on highways relatively soon. And once people don't need to pay attention most of the time, they won't. Heck, they don't today.

The other big question is how does Uber impact total miles driven.

It increases it, obviously. Any more efficient utilization of cars increases total miles driven, because more efficient utilization of cars is by definition taking a car that was parked near its owner, driving it over to near someone else who needs a ride, and then driving that car back or some other car over to near where the first car's owner was when he eventually needs a ride.

Maybe. On the other hand, to the degree that rentals (in whatever form, including Uber) replace car ownership, the marginal cost of driving somewhere goes up so you plan your trips better. Forget Uber. Think about if you use a Zipcar rather than owning a vehicle.

Who the hell would use taxi/uber every day to go to work? Thats insane... Much more expensive than public transportation (unless you share it with other 4 or 5 people, which is a big if). Also why not use bycicle instead? Maybe less glamorous but for distances up to 10k is great. My point is "can-do without cars" needs much more than just Uber... Ok, they might drive taxi out of business, but you have to assume they will do something to lower their cost and compete with Uber... I mean they are not 100% stupid right?

Time is money. If a taxi saves enough time over other methods of transportation it can be well worth it for many people. Of course just how fast a taxi is compared to public transport will depend on location. And having time free to work or sleep while riding in a taxi can be a reason why people take them instead of driving their own car. That's also a reason why people take public transport. It appears that smart phones have lowered the opportunity cost of using public transport for many people and contributed to decreased rates of driving and car ownership among young people.

Sitting on public transport is fantastic "me time". Maybe 80% of my reading is done on the bus.

You can do that in an Uber or your future self-driving car without having to jostle in a dirty environment full of poor people.

I get sick reading in cars :(

You get used to it really quickly.

It's interesting how so much technology these days (Amazon, home entertainment, "self-driving" vehicles) is to enable affluent whites to avoid public spaces which they perceive as increasingly hostile.

@The A-G: yeah, all that whites-only tech. What do darker people use instead of Amazon and home entertainment?

@msgkings They still go to WalMart because they don't want to pay the extra shipping.

If you can afford it. I bet you most Americans can't! But Tyler speaks for people with IQ and/or wealth in top 5% of the distribution only...

Clear America needs higher median incomes or cheaper taxis. Or both.

I actually know someone who does this. I think he goes to the office 3-4 times a week and always takes Uber. He does it because, while he lives in Arlington, VA, a city with Metro access, he is in between stops such that it would mean walking like mile to a bus and then taking it to the Metro. The total distance to his office is really not far, so the cost is not that bad compared what he would pay for parking. He also often drinks with clients after work--not sure if that is part of the Uber calculation.

If he would walk/bike to the metro station, not only he would save money, he would also have better health and less stress ;) (Ofc I know distances in DC area are quite big, but still a point)

I know people who do this. It's cheap, comfortable, and reliable.

Bikes, for obvious reasons, don't work in the weather or if you need to be presentable in an office environment. At least without putting a lot of effort into basing your wardrobe in the office, which I realize some guys do. Women basically can't. Nor do bikes work for anyone not young and fit.

Bikes work, even if it is snowing and is very cold. I have done it for years...

I'm guessing you're young, male, single, live in European "cold" and not American "cold," or European "hot" and not American "hot," don't need to look nice at work or be bothered if you smell, and aren't really thinking about safety.

Bikes are for poor countries where there are no better options.

Such narrow-mindedness... so typical... just come and live in Europe and you will see ;) ... You can even carry bike on the train/metro/bus (if you buy a foldable bike), so just use it in the parcours where there is no public transportation or where the connections are not good.

Also if you are a libertarian, the bycicle is the only true "free" transportation. no tax, no license plate, no regulation, etc.

Europe is both poor and childless Moreno. It's not a good point to introduce in the argument.

Really lool? Well coming from someone who thinks women can not bike (facepalm).

"Well coming from someone who thinks women can not bike"

Biking on weekends in a residential neighborhood or a park, sure.

You see a few people using a bike to commute in NE Ohio. Males, under 40. All the ones I know personally have access to

I have never seen a woman commuting to work. Its not that they cannot, they just do not.

Biking is dangerous, time consuming and subject to weather conditions. There are hills on all routes going to the main city in NE Ohio, biking UP the hills is arduous as well.


All the ones I know personally have access to changing facilities and showers and have flexible job starting times.

Bob, your experience is similar to mine.

I know guys who bike to work, where they shower. And they work in the kind of place where something that was rolled up in a backpack is acceptable office attire. Uniformly they have a car that they use in inclement weather. I just don't know any women who do this - I can't imagine the clothing options would be acceptable to most of them, even if they had no other objections.

Its really strange. Here its so normal to see a woman bike (from 5-95 age group) that your comments almost sound bizarre to me. But ok, i admit that in the country where i was born for instance, the culture is also very automobile centric and its rare to see women biking. (But the country where i was born is kind of backwards so...). Also geography and climate can be a big obstacle, although in most cities there is always some way to avoid climbs.

You're German originally, right? I don't think Germany is backward, I just think they're so concerned with appearance, what with the reputation they need to live down, that they favor appearance over reality reflexively.

I am not German and i dont live in Germany :)

Also your theories kind of fall down with electric bikes. No more problem going up the climb or sweating. You can even buy an electric folding bike, which basically means you can go anywhere, taking train/light rail for the bigger distances. No need for driverless cars...

They still don't work in weather, hot or cold.

Have you bicycle commuted in the American South where mass transit just means a bigger pick-up truck?

I have been in the American South, and i have noticed that virtually no one uses public transportation. I was the only person in my flight that used public transportation to get to my destination loool

Public transportation is at an inherent disadvantage in USA because of the low population density compared to Europe.

A lot of people would. It costs me $6 to get to work by Uber, as compared to $2.75 by bus. If I save even one minute, I'm ahead because I make more than $3.25 a minute (at least I will if the market ends up this year). And I think a lot of people who post here are in similar situations.

Take a look at Uber's flop in HK (it was a flop even before the police cracked down) for some evidence in this regard. Not a car to spare in this city unless it's owned by someone who very much does not need extra income from driving Uber.

Of course there are other reasons: 1) Taxis are already very cheap, 2) Taxis are already everywhere - in 90% of cases you can get a taxi in less time than you would wait for an Uber in SF, and 3) Cash, not credit, is king

De-regulating taxis and even buses would be an obvious move to make transportation cheaper and more accessible. Uber is a response to regulation and wouldn't even exist if cities didn't feel compelled to control every aspect of their residents' lives.

At least in the U.S., Uber does not have the capacity to transform society. Most Americans don't live in dense urban environments (where roads are congested and parking is scarce and expensive). Before Uber, what was the percentage of Americans who used either cabs or public transit on a regular basis? At least for commuting, the numbers are VERY low (5% commute by public transit, 0.1% by taxi vs 85% by private automobile):


Why do Americans living in dense urban environments persist in believing their experiences are representative? Uber has undoubtedly been wonderful for city dwellers who were previously stuck with the taxi cartel, but it's had no real effect for the vast majority in the country.

Yep, nailed it.

1% of America deals with urban traffic congestion.

9% think they drive in urban congestion.

90% think urban congestion is waiting for a red light.

1% of America is 3,300,000 people. There are that many people in San Diego.

20% is more like it. If you've ever been in a suburban traffic jam where it takes two hours to go 10 miles, you might know what I am talking about.


It is single oriented as well and ignores suburban families with children. Sending Uber to take them or pick them up after school?

Don't you know affluent people, especially Europeans, are not supposed to have children anymore, or more than one? You can always import your replacement population! I just saw a news report on TV about micro-homes and hipster types who think they are amazing and oh-so-sustainable.

Why not? I walked to and from school by myself from about 10 years of age.

What did you do at 9 years of age?

Another in a long line of posts opposing the public option (whether conventional public transit or the less efficient public fleet of driverless cars). Uber, driverless cars, and the other private options Cowen favors for moving people from place to place are like re-arranging the deck chairs on the Titanic. People who rely on the inefficient automobile for daily commutes know its inconvenience: traffic congestion and wasted time and money, not to mention the air pollution. But have you driven I-95 during a holiday or the summer vacation season? It's like a parking lot, from New England to Florida. People continue with their irrational relationship with the automobile for the same reason they eat at McDonald's: because they are. I have a home on an island, where the distance from one place to another is no more than a few miles and the speed limit is no more than 45 mph. Yet, the transportation of choice is the SUV, behemoths circumnavigating the roads like dinosaurs. It's an ideal place for a small electric car, and there are a few, driven by oddballs with Obama/Biden bumper stickers on their way to the Universalist Church - this is the Bible Belt. This being home to many very wealthy folks, there are the automobile excesses, including the Rolls Royce Phantom I witnessed being delivered a few Saturdays ago for use in making short trips around the island. No, you won't see it on I-95: the owner has a private jet for trips away from the island.

In your hypothesized scenario, transport would become a powerful economic driver for the low-end of society, maybe on a similar scale to what pluming and sewerage was to the masses in the late 19th century. With 40% less traffic, the future would be unrecognizable. You worry about the knowns when clearly, an amazing surplus exists that will propel a new class of people to a world which is as unimaginable as now to people in 1870.

This is pure fantasy... low-end of society can not afford taxi or uber....

How low end are you talking about? These are not big luxury expenses.

In America, most of the poor already own cars. Of course they can afford Uber. Uber is basically car + driver, less the cost of parking. Today, parking is more expensive than drivers in the city. It's less expensive outside the city, which is why Uber has no presence there.

Do you know how the poor get to own those cars? They get outrageous finance deals through shady "Everyone's approved!" used car dealers and then the car gets repossessed a good portion of the time. That's good for the car dealer because he gets to finance the same car to another person with no other options. Does Uber finance? I guess if the poor could get credit cards (They can't) they could use Uber until their card was canceled.

Somebody's driving those old Civics. And even in poor neighborhoods around here, there's no shortage of cars.

Maybe you have a more restrictive definition of "poor" and I'm being more expansive. In fairness to you, the American poverty line is set at a level that few would actually consider "poor" by any global standard.

If you use Uber to cumute every day, it will be much more expensive than public transportation, on average, unless you can ride share, or we get the driverless vehicles...

I used to commute via public transit in a major American city with heavy transit subsidies. I saved money by converting to using my car.

I don't think Uber's quite there for everyday use - obviously, it's a markup over having a car - but it's not that far off, and the convenience and ability to work while commuting might be worth it for some.

Really, plumbing? How, exactly, Colin?

What cities real need is an uber helicopter service.

It exists and is called blade.

Doesn't getting in some stranger's helicopter seem kind of sketchy? They could be a serial killer, there's no protection against them taking the i by route to drive up the cost, AND you don't even know what kind of insurance they have. No thanks, I'll take my own chopper.

This is exactly why I always fly in my own airliner.

Not if the alternative is getting kidnapped for ransom while you're driving your BMW around the favela looking for authentic tapas. That's why some of us don't appreciate people like the GMU Econ Department trying to turn the US into Brazil or Mexico.

+1 of course, parts of the US are already like the some of the worst parts of the third world, and I don't mean the WASP parts.

So this suggests the system could be in a state of flux as, over time, it seeks Uber equilibrium?

"Resources put into automobile production can be diverted to other purposes, with potential environmental benefits as well."

Since auto workers are the highest paid manufacturing workers, what new purpose will pay as much? I suppose the money will be spread among some higher paying jobs and many lower paying jobs. I'm not sure that is what most non-economists would call progress, regardless of the environmental benefits.

Since uber is doing the world good and becoming demanded more, wouldn´t his increase the the market for automobiles? I understand your reason for the demand of cars to decrease because more people we depend on uber's for transportation but in order for this service to be provided cars are needed. As the demand of uber increases wouldn't the company need more employers who will need to have cars to provide their rides to buyers?

No, it will just crowd-out taxis...or vehicle ownership will decrease (if Uber is so good)

But it Uber being demanded more then they will a greater supply of Uber drivers. This includes those workers owning their own cars that meet the requirements made my the Uber company. So if Uber is good then wouldn't the ownership of cars could possibly be affected positively too?

Seems like this is what happened to air travel after deregulation (over a long period of time). Airlines became more efficient: flights are crowded, flights to small Midwest markets (e.g. anywhere in Iowa) are few and far between or non-existent, everything is priced (bags, peanuts, etc.).

Efficiency is a good thing, but not necessarily for airline passengers when they are flying.

People have the option *today* of flying in the luxury of the 1970s, it is called business class. Weirdly, very few people do it unless they are spending OPM.

The trend has been, for as long as I can remember, that people move the hell out of the big city and into the suburbs as soon as they can afford to do so. While this isn't guaranteed to continue forever, there is still a lot of empty space in North America, and besides that, telecommuting is becoming ever-more-prevalent.

So my inclination - which is probably wrong, because I suck at predictions - is that there will be more and more Uber drivers competing for fewer and fewer customers. Total number of taxis will be reduced unless and until Taxis start performing more like Uber drivers. Total number of cars on the road will not change much.

Not anymore. Crime is down, hipsterization, etc.

Not really. A lot of the reporting has been misleading nonsense -- this for example:

"More Americans Moving to Cities, Reversing the Suburban Exodus"

Notice that the headline and data in the article are contradictory -- the increase is for people living in metro areas (which includes suburbs and exurbs), not central cities. For example, between 2000 and 2010 the Chicago metro area gained 360,000 people. At the same time, the city of Chicago lost 200,000 -- meaning the suburbs and exurbs outside the city gained more than half a million in total. Sure seems like continuing suburban exodus to me.

Explain to me why, in a post-taxi world, Uber would not face downward pressure on price from new, more nimble competitors.

It will, but the current guys have already made their money. Not that there's anything wrong with that.

Such a NYC/DC/SF conceit about this post.

Nobody but poor people use buses. They are massively time consuming and only ares used by people with no other options.

Light rail, sure, its more convenient but there are only a few lines at best in most cities. A niche use at best.

If you have children, especially multiples, using public transportation is useless for anyone with any resources.

I have seen cities (not in US) where children, even in primary school, use light rail/metro by themselves, without adult supervision.

But you are right, i mean, Uber (with driver) is not the solution for the majority of americans.

I really don't see Uber replacing personal car ownership much more than Taxis already have. The places where it is more efficient to take a cab than to own a car already have people taking cabs rather than owning cars. Uber's slightly lower prices merely shift that into a slightly larger area. It doesn't really change that out in the suburbs people are still going to want to own a car and commute to work. I doubt Uber's prices or response are really going to drop to the point where it's cheaper and more convenient to call an Uber cab to commute to work every day.

Now that you say it, it seems quite obvious.

But perhaps with self driving cars, costs will get low enough that it may become practical.

Self-driving cars will be quite expensive.

No, it's mostly just electronics -- expensive to invent but cheap to manufacture. If self-driving cars actually work and make it to market at all, after a few years they won't be any more expensive than non-self-driving cars.


And by being able to park remotely, they'll solve the major issue of cars in urban areas: the cost of parking them. They'll greatly increase car ownership in cities.

Assuming they work out technically, which is still a big assumption.

"And by being able to park remotely, they’ll solve the major issue of cars in urban areas: the cost of parking"

They'll solve the hassle of you having to park them; car owners will still be charged for parking space.

Parking in my office building is nearly $600 a month. Parking two miles west is $100 a month. The issue is trying to park in the very-high value real estate where people want to be.

A self-driving car could economically drive back to my home, park for free all day, and come back and pick me up.

I always feel the need to hedge that I'm pretty skeptical of the technology, but the economics are very attractive.

You are buying a car plus very sophisticated robotics, CPU and a sensory array plus a certain amount of redundancies . There is no way it won't be any more expensive than a car without all of that.

All cars will have to have that the instant it shows improved safety. And it's electronics, not sheet metal. It will eventually be cheap. It's basically just the sensors that are unique to self-driving anyway.

Inventing it will be expensive, but building it will be cheap.

That's unlikely, the costs of manufacturing the electronics themselves will likely be a small component of the total cost of a self-driving car. There are lots of other costs to be paid upfront, in systems integration, and safety testing and other regulatory compliance. Consumer electronics are probably a poor model for the costs involved, I'm guessing civilian aviation or medical devices are a better model, since the manufacturer will be liable for any safety issues that come up.

"the manufacturer will be liable for any safety issues that come up."

You mean that the manufacturer will capture the profits of the car insurance industry. That's not a cost, it's a benefit.

I meant the cost compared to Uber or taxis, because you don't have to pay the driver. The amortized cost of the car itself is a small fraction of the cost for having a driver.

But the discussion on the cost of self-driving cars themselves is interesting. I agree with Slocum: "expensive to invent but cheap to manufacture"

All of your analysis of the impact of uber is based on the time a car is driven.

But rather than time if you used miles driven you might come up with very different conclusions. For example the uber driver would have to replace his car based on miles driven so he would have to replace the car more often.

Does uber make any difference in miles driven?

So why should you evaluate auto usage in term of time used rather than of miles driven?

Doesn't the argument here hinge on monopolistic rent-seeking?

Uber is a taxi service, and self-driving doesn't change that fact. Most people don't take taxis everywhere because it's expensive and owning a car is cheaper and more convenient. Why wouldn't car companies try to sell consumers their very own self-driving cars that were cheaper than or cost competitive with taking Uber self-driving taxis all the time?

The argument then proceeds to say that this won't happen because parking will be too expensive. But parking in most places in the US outside of downtowns and city centers are very cheap or free, so this would presume that people would monopolize location rents and raise rents so high as to make parking too expensive for most people, hence car ownership and driving too expensive for most, even if the self-driving feature would mitigate this by driving to parking lots further away from the destination. This then would allow Uber taxis to monopolize transportation and avoid competition from car firms that would sell you your own self-driving car for cheaper, and of course the price for Uber taxis would be higher due to its monopoly.

Why is this sort of monopoly and rent-seeking regarded as a positive thing by the free market, libertarian economists and readers here? This monopoly would be very centralized and be a kind of central planning.

I don't think it would result in a monopoly. The only barrier to entry in a given city is having enough money to buy a decent fleet, which some dozens or hundreds of American firms could walk into if there was extra profits in it. And on the software side, I don't doubt that hundreds of firms in SF could get something working in the space of a year.

"Why wouldn’t car companies try to sell consumers their very own self-driving cars that were cheaper than or cost competitive with taking Uber self-driving taxis all the time?"

It doesn't make economic sense to own a car, because the car will sit unused 20+ hours per day. It makes more sense for a fleet owner to own the car, and have the car on the road as close to 24 hours per day as possible.

I was going to say something like that. But perhaps the cost of idle capital is not such a huge deal, and cars driving 12 or 24 hours a day are going to wear out a lot faster.

The intuitive appeal that it means using idle capital more efficiently can be easy overstated, I think.

Electric cars are mechanically much simpler

"But perhaps the cost of idle capital is not such a huge deal, and cars driving 12 or 24 hours a day are going to wear out a lot faster."

But the cost of idle capital is a huge deal. And so are other fixed costs that must be born regardless of whether the care is driven or not.

A person pays insurance on a car, the rate of which is not terribly closely related to the miles traveled. So the cost/mile for insurance in a car driven 5,000 miles per year is about twice the cost per mile for insurance in a car driven 10,000 miles per year.

Property taxes are also fixed costs, so the cost/mile is cut in half if the miles driven are doubled.

Further, a car driven 12 or 24 hours per day can be driven many more miles before the end of its life than a car driven only 2 hours per day. For example, engines suffer a lot more wear at cold startup than if the engine oil is kept flowing.

And it makes a lot more sense to do major repairs and more careful maintenance on a car driven many miles per year. For example, suppose you need a new engine in a car that's 10 years old and 150,000 miles. You probably don't do it. But if that same car has 150,000 miles after only 3 years, you'd probably go ahead and put in the new engine.

P.S. I think electric cars will really be favored when computer-driven cars rule the road in <3 decades. So I don't think many cars will even have engines.

In contrast, we are currently living in a world where there is excess capacity of automobiles, especially for short-term rental.

I wonder if a large number of people will shift vehicle preferences from owning cars to renting cars and owning some combination of bikes (I now have a cheap single-speed bike), electric bikes, and scooters, all of which are a tiny fraction of the cost and hassle of a full-sized car. Think of it as the New-York-ification of more of the world.

New York-ification???? This trend certainly didn't start in New York.... In fact i dont see many bikers in New York, at least in Manhattan. In some European Countries bikes are massively used already for years.

Most people in NYC don't ride bikes or scooters, and taxis are too expensive to take regularly. Most people in NYC have to walk and take the subway, which is pretty dirty and has bums in it and often smells like piss.

Kids need to ride in car seats, by law and according to official safety recommendations. How are they supposed to do this if their families use Uber to get around? (A search for "car seats" on this page turns up nothing. How did nobody think of that after almost 200 comments??)

Aside from car seats, kids just come with a lot of stuff. Families with cars typically store some of this stuff in the car full-time.

In general, people keep a lot of their stuff in their cars even when they're not driving. You can't do that with an Uber cab. So, cars also serve as portable storage.

Thus, cars are not just point-A-to-point-B machines. They are customized to match our particular needs and lifestyles, even to express our personality. We customize them further beyond that. But with Uber you get only what shows up.

Finally, cars are an integral part of the American culture any of us can remember. It's the culture we grew up with, and it's one of the main things that distinguishes us as Americans. Few motifs are more American than the individual, free on the open road in a privately owned car. Classic car shows are cornucopias of Americana.

It's something many of us hold dear and don't want to see go away, and the cosmopolitan efficiency of Uber and driverless cars (apparently designed for a world without children or personal effects) are nowhere near compelling enough as a counter argument to change our minds.

Just checked. The following terms do not exist on Uber.com:

- car seat
- kids
- child
- children



Thanks for the link.

It looks like UberFamily is a pilot program in certain large cities (NYC, DC, Philly). It means that Uber drivers are carrying a certain car seat with them, and it accommodates only children between certain age/weight ranges. It comes with a $10 surcharge.

For people willing to spend the extra money who have kids within that range, it solves the car seat problem. That still leaves all the other problems.

That would definitely make for a strong case that people are more likely to own their own self-driving cars rather than use Uber.

This is significant though.

People keep saying driverless cars will eventually lead to the end of car ownership. If they're wrong about that, what else are they wrong about?

What else about having humans behind the wheel are the driverless car bravoists forgetting?

Uber with human drivers is a transient phenomenon, of no lasting importance.

Uber founder Travis Kalanick knows that the real revolution is computer-driven vehicles. Those will transform the world:

1) Freeway speeds in excess of 100 mph.
2) Personal vehicle ownership eliminated.
3) Cost per mile cut from 60 cents per mile to under 15 cents per mile.
4) Parking lots and garages (both commercial and residential) eliminated.
5) Brick and mortar stores of every type eliminated (Walmart, Target, Costco, Lowes, Home Depot, Kroger, Walgreens, etc. etc.).

This will all happen in less than 3 decades.

The first paragraph got me thinking...

"Price is No Obstacle"

"We argued...that the economics of self-driving taxis don’t necessarily make sense.Which is to say, we’re not entirely convinced (at this stage) that self-driving taxis will be any more or less affordable than those driven by humans. "

That is irrelevant. The economic value is that people who are currently driving taxis will do something else productive... Likely something that is either not being done now or is under-served based on the desire of the public. We should hope that self-driving taxis WILL (without general inflation or legal mandates) increase in cost because that means that the value (productivity) of those trips is increasing.

Comments for this post are closed