The Effects of Uber and Lyft in U.S. Cities

Huh, I’m currently in the opposite boat. The cost of ride-sharing a few times a month is so much cheaper than the total cost of owning a car. Like hell it’s less than just the insurance.

So my wife and I decided to do the digital nomad thing earlier this year. Our first thought was that we would drive her Hyundai Tuscan and across the country and sell my old car.

Then we realized we don’t like road trips when we drove from Atlanta (our home) to Nashville. She gets carsick when she is not driving. Which is ironic because it doesn’t happen when she is driving and she has had a commercial driving license for decades and she was a city bus driver before we got married and a school bus driver afterwards - mostly for the benefits until two years ago.

We sold both of our cars saving $800 a month in car notes and insurance not counting maintenance - I was already working remotely - and we decided to fly everywhere and have a $900/month “Uber budget”.

Even before then, I hated everything about car ownership and for year we had one car when I gave my car to my older son who doesn’t live with us. I loved when I went on business trips and didn’t have to drive.

We don’t have to worry about car maintenance and repairs and statistically flying long distances is much safer than driving.

Another anecdote is that before my son could drive he had a job. We paid about $400 a month for him to take Uber everywhere which was much cheaper than a car note and car insurance for an under 21 year old male ($550).

No we couldn’t buy a big six foot three dude an old little beater and I hate dealing with car repairs.

Yep. Owning a car is about availability. You want to go to X place at any time, any place? You can walk to your car and do so immediately.

The same happens with private jets. Yes, some first class seats are insanely luxurious. But you will never have the same experience as owning your own jet that you can hop on at a moments notice.

If you live in an area with readily available rideshares, it's even better. It may be 5 minutes instead of "instant", but you'll save some of that 5 minutes by getting dropped off directly at your destination - no parking/walking required. And you don't have to worry about gas. And you can drink if you choose!

I couldn't remotely get by without a car where I live. But even if I lived in a city with good transit I'd [in most cases] still own a car so I could head out of the city for a weekend without a lot of planning, have the car setup the way I like it, etc. I know a couple who live in a city without a car and I hear of their hassles with Zipcar, regular rentals, Uber, etc.--and it really doesn't sound worth it. Maybe right out of school where you do a lot of activities with friends some of whom have cars. (Basically how things worked when I was an undergrad.)

I'm unsure if anyone is seriously making arguments that dropping in Uber/Lyft into a city like Phoenix, AZ would be much more efficient. Rather, it would hopefully help as an intermediate transportation mode for cities that have a somewhat useful transit system. For those cities, mass transit is the highly available mode.

To state the obvious, people can want all the availability in the world, but they don't often get that because of the costs associated with it. Cars are expensive[1], and so is parking[2]. The costs of congestion are also typically only paid for by drivers with time (although many cities are toying with congestion pricing). Of course, the cost calculation for residents changes for different cities; the costs in Manhattan vs in Atlanta are very different.

So while you in particular may be able to afford the private cost burdens of car ownership, I believe the convenience of owning a car is a woefully inadequate metric to make policy decisions on. This is all mostly moot anyway as at least according to the original post, ridesharing does not positively impact car ownership.

[1]: https://www.hks.harvard.edu/publications/64-billion-massachu... [2]: The High Cost of Free Parking by Donald Shoup (ISBN: 1884829988)

On the flip side, one of the things I love the most about ridesharing is that you're always just dropped off at the front door and don't have to worry about availability and quality of parking.

Even in the best cases, any moderately popular place is going to involve a winding driving path and a moderate trek through a car park to get to where you actually want to be due to the inherent lack of density of cars.

In the worst case, there was a time when a group of 4 of us needed to go from the Mission to Alamo Square and one of us happened to have a car on us which involved a 15 minute drive and then a 30 minute circling of the block looking for parking which finally involved us finding a spot half the distance of just walking to the place we wanted to get to.

Sorry for late reply, but I meant the former. I find it plausible that even with robust public transit, the idle miles driven by rideshare do not incur any net benefit.

I heard of taxis, not sure I saw one anywhere other than the airport.

You could never get a taxi from a warehouse in LA to the south bay cities at 6AM before uber. You'd get laughed at. You pretty much had to have a place to crash or you're out $100 and you were waiting a while for the cab.

None competitive with Uber/Lyft. Too much sprawl to for hailing to be viable, so pre-Uber one would have had to call to request a cab, wait as indeterminate amount of time, risk a no show, all without feedback.

lol, taxis in the US outside of NYC and maybe the core downtown of a few other cities were a joke.

You’d call in to get a cab and they would say someone “should be there in 20-40 mins”. Then there was a 70% of them showing up if it wasn’t around bar closing time and maybe 25% if it was.

If you wanted to plan a night out you needed a car service if you wanted reliability and that cost a fortune.

That doesn’t even begin to scratch at the “my credit card reader is broken” and “should I take route X, it’s faster” scams nearly every cabbie would pull on you.

Inconvenient, unreliable, expensive.

I can’t speak for LA, but taking a cab in other cities sucks and sucked. Dispatch was rude and inaccurate and you had to actively watch and assess the meter if you didn’t want to get scammed in the cab itself. Also if the driver didn’t want to go the direction you were headed you were SOL. This can still happen with ride shares of course, but it happens when you are trying to find a ride instead of when the ride arrives.

> why didn’t people just take a cab before Uber

This is addressed in sibling comments, but basically it was impossible to get one, especially late at night, and if you did, it would cost a fortune.

Yes. Getting a cab into the city/nightlife is difficult but possible. But if you don't know when or where you're going home from. B/c you want to hang out with your friends and moving around, then coordinating a cab ride home was near impossible. Only exception was NYC where the density of cabs in nightlife locations kinda/sorta-of solved this.

Because they were absolutely terrible. No feedback to back-pressure terrible service and scams. No pre-quoted fares. No arrival estimates. No reliability.

Based on my experiences going out in the pre-Uber world: When taxis are in high demand they’re also every hard to find. Usually you could find one in ~15 minutes but not always. And back then they’d sometimes insist on cash so if you didn’t carry enough you’re out of luck.

So I’d say that while convenience is one factor, it’s more about reliability. Lyft/Uber just work. Surge pricing is frustrating but it does create the right incentive to get more cars on the road.

Well Uber and Lyft were subsidized by investors and naive drivers — cabbies negitiatee their worth and thus it was a much lower margin experience for consumer. Eventually when Uber needs to be profitable and drivers demand real wages, all that happened before will happen again. But we will have nice apps at least as we wait.

When calling for a cab (in most places you need to call a central dispatch to get one, hailing only works if you’re in the middle of a major city), you have to call into a dispatch service and they put out a request over the radio. Cabbies each decide if they want to take the fare, and if nobody does then too bad for you. There’s no feedback telling you if someone accepted and if you’re too far out of town, it’s very possible that no one will ever show up. I’ve definitely sat for hours waiting for a cab and there’s no way to know the status.

At least with these apps, you know if the ride was accepted and when they’ll be there. If nobody accepts, you can make other arrangements.

The problem is that drivers perceive removing the existing carrots as applying a stick. If private car owners paid the full cost, transit would be a no brainer. How much public land is devoted to cars? How many DOT employees have jobs only to maintain traffic signals and pave roads?

The problem is that, at least in the U.S., we have been giving carrot after carrot to car owners in every conceivable way. We've frozen gas taxes, built highways into downtowns, mandated parking on all new construction, zoned to create the ideal environment for cars, and created a trillion-dollar nationalized highway network.

Now, when people advocate for public transit carrots, those carrots have to compete with a century of driving carrots and the momentum that this status quo has.

maybe this is just my american myopia, but this seems like an unrealistic goal, at least outside of the densest urban centers.

I lived in northern virginia for a while, literally 100 feet away from a stop on one of the best public transit systems in the US. inside DC sure, the metro was about as fast as an uber, or even faster for some routes / times of day. but starting from across the river, it often took twice as long by metro. I honestly don't see how they could make transit better than uber without running trains so often as to negate the efficiency advantage of bulk transport.

as an engineer, it kinda seems like people are expecting the latency of a synchronous request/reply service, but combined with the throughput of a batch job. surprise! you can't (usually) have both.

imo both point-to-point and mass transit have their place. but people are always going to prefer point-to-point when it's economic for them. congestion pricing seems like a reasonable way to preserve the low latency characteristics of point-to-point, while nudging people to use the batch job when it's sufficient for their needs.

It's funny about that. Our transit system began an initiative just a bit before COVID-19, aimed at cleaning up the bums and criminals. So most people are not able to hang out on the train or bus, they're hustled off at the nearest stop.

What effect that did have is that the train and bus is more empty, but all the bus stops and train stations are clogged with people literally camping out. There is nowhere to sit and wait, because the stop owner either removed the seating or someone is sleeping/camping on the benches already. Every stop is full of skeezy people who have nowhere to go and aren't boarding the bus.

So perhaps the perception of clean/reliable/safe has increased aboard the bus, but it has plummeted at the locations where it really counts: the stops and stations where I spend the most time. I feel notably unsafe when waiting for a bus and there's nothing to be done about there, and there's no recourse because nobody's in charge there (the city will point the finger at transit and transit will point the finger back at the city. LEOs won't touch it because it's not a priority.)

I think the perception with crime on the subway has much more to do with the fact that poor people use the subway, moreso than a danger of subways for poor people.

Crime is certainly up, but still you're looking at 1 in a million or less odds to be victim of violent crime on the subway.

And anti-social behavior will problem actually go down as a proportion when there are more people on the subway.

They ought to just define the word the way everybody else does. What possible reason could they have not to?

In the policy brief, they're

- not investigating TNCs, by any reasonable definition, the two companies they target are publicly traded companies

- exclusively seeking to criticize ridesharing relative to public transit, or as-yet non-feasible means

- not framing things in any conceptual framework, ESG is not a conceptual framework it's a political viewpoint about role of government (in this case literally regulatory recommendations) masquerading as a conceptual framework

The structure is not investigative, it's persuasive. This is what market manipulation looks like.

CMU is situated in an extremely affluent district in Pittsburgh, and this type of behavior is to be expected. Totally beautiful place though, do recommend a visit if you have the time.

To repeat, outside the frame: places that didn't have a robust taxi system to begin with, internal operations and transformation capabilities of a tech company controlling taxi and traffic distribution.

A lot of people parroting this line, but I suspect you guys all did not use or were not in the market for Uber or Lyft. Their main offering pre-pandemic was in fact Uber Pool and Lyft Line, which was indeed shared rides. This was extremely popular and was the default option in the app.

NYC had quotas - "taxi medallions". Arizona had minimal regulations: vehicle safety inspections, meter accuracy, driver screening. The big taxi companies were mostly trusted to do their own regulatory compliance for their fleets and drivers. Anyone could start their own taxi company - I understood the state agency was quite helpful. I had no interest in making a career of transportation, so I just drove for the company enough to pay my bills, while I worked on my other endeavors.

Deregulation helped.

But the Uber dispatch was way better. Ubers would actually show up and you could see them coming on gps.

For a while in Chicago, uberX was banned but you could use Uber to hail a registered cab. It was still much better than trying to use dispatch.

What about a dense village or town, with nature nearby that you could walk to in 10 minutes?

I wonder if single family home communities might be nicer if the homes were clustered more closely together and surrounded by natural areas. But having only single family homes might not be dense enough to get the advantages that come with more density.

It's a real shame that zoning for single family homes, lot sizes, and parking requirements became so uniform across the country. It's not easy to do anything different. Making things worse the taxpayers in more dense areas end up subsidizing the infrastructure build-outs in less dense areas, through state taxes and grants [0], along with other car-related subsidies talked about in this thread paid for by all taxpayers.

[0] https://www.strongtowns.org/the-growth-ponzi-scheme/

In many (maybe most) cities, there are regulations that taxis have to follow. If you want to drive a cab in NYC, you have to buy a medallion.

Uber simply ignored those regulations, and ran unlicensed taxicab operations. While taxicab companies were paying for limited numbers of medallions, charging fixed rates, etc., Uber was just doing whatever it wanted.

Even if you think the regulations don't make sense, you can see why the taxicab companies are upset.

Taxi companies do not have venture capital funding that will eat a $2 billion loss per quarter. They have to follow labor laws.

The prices have drifted up to the point where there's often not much of a difference between taxis and Uber/Lyft. Uber can even be more. But, yeah, I never really understood why the Uber strategy forced them to undercut taxis by so much when they were often a better service--especially after it became obvious to just about anyone that large-scale autonomous driving in the dense cities where the economics work best was at least a ways off.

The answer is probably better Taxi industry regulation. I remember having to find a Taxi in Los Angeles before Uber, and it was impossible. Surely there is a way to ensure taxis provide a better service without destroying the entire industry.