I feel like this article disingenuous. Uber has reduced drunk driving accidents by 62%!! (in Portland, but likely similar for other cities) That's huge.

A 10% increase in heavy drinking seems small by comparison. Why not focus on that fact? Well, because it makes Uber look good, for one thing. Can't have that.

> A 10% increase in heavy drinking seems small by comparison. Why not focus on that fact?

10% seeming smaller than 62% is not a fact that anyone should be focusing on.

There are actual numbers that can be compared in context; the fact that a percentage of one thing is expressed as a larger number than the percentage of another thing doesn't mean it's more important. Very few people die of drunk driving accidents (as compared to the number of people who drink), so if 20 fewer people in an area per year die from drunk driving, but incidents of binge drinking in that area increase from 1 million to 1.1 million, it's not simple to say that the outcome has been positive. Maybe 21 more people get esophageal cancer. Maybe 21 more cases of domestic violence end in a homicide.

That being said, this is a wealth effect and inevitable if you're making progress. People are drinking (out) more because their perception is that with rideshares it has become safer (easier to get home.) A cure for HIV would cause an increase in the spread of other sexually transmitted diseases, for example, but that doesn't mean that you shouldn't cure HIV.

One material difference I can see between rideshares and a cab is that rideshares are often cheaper because 1) they are subsidized by investors and 2) they can be shared. This doesn't say anything interesting about Uber; if you wanted to make cabs cheaper, you can tax them less or subsidize them. The other, which I think might be more important, is that many rideshare users don't feel the need to buy a car, and that the rideshare demographic is largely similar to the binge drinking at bars/clubs demographic. Of course, taxing cars or the parking around bars/clubs could have the same effect.

I agree with you on all but pointing out the cancer rates.

If we're looking at it purely from a statistical point of view, then yes, the numbers might not look that good in favour of "more drinking + ridesharing is better than less drinking + drink driving".

Ethically, I consider people drinking themselves to an early grave of their own free will to be "less bad" than a drunk driver killing someone else on the road or a drunk person committing a violent crime against another person.

I make this point in the context of the rest of what you've said, though, which raises good discussion points about "should we fix X really bad thing if it leads to people doing more of Y maybe-less-bad thing?"

Ethically weighting deaths is a dangerous road. What would you say to "Innocent deaths from drunk driving go to 0, but suicides increase 1000-fold"?

This is a question I don't feel equipped to answer, despite having asked myself many times.

I think it's widely accepted (and I believe) that "drinking + driving + killing someone else = very bad", and that "drinking + assaulting = very bad", but I don't think I know enough about the health implications of drinking.

If a person really was drinking themselves to death of their own free will (i.e. they are aware of the physical health effects, and they do not experience any addictive effects from alcohol), then I don't see an ethical issue with them choosing to do that.

But if suicide rates go up, why could that be?

My opinion is that if someone who chose to drink (i.e. knew the risks and was not addicted) accidentally fell off a bridge or accidentally shot themselves with their own gun, it is their own fault.

When someone is suffering in terms of mental health such that they are addicted to alcohol, or are depressed, or feeling suicidal, then I just don't know. It feels to me like that although the alcohol is not the root cause in these situations it definitely could've pushed them over the edge. I think this is analogous to "someone who felt suicidal committed suicide after another person told them to jump off a building"... I really don't know where I stand on the issue.

> One material difference I can see between rideshares and a cab is that rideshares are often cheaper

Taxis (at least in Melbourne) used to be uniquely awful for getting home after a night out. They'd make you "prepay" a ridiculous amount like $50 cash for a trip that should have cost $15 and then refuse to give any change at the end of the trip. For the specific set of customers I was in, taxis were more expensive than themselves because they were basically stealing from us.

The paper does open with that, so at least the authors acknowledge that right up front:

"Recent studies show ridesharing services, such as Uber and Lyft, reduce intoxicated driving.However, ridesharing may also have negative health effects by increasing alcohol consumption."

"When we focus on areas with relatively weaker public transit options, we estimate UberX is associated with a 17.5% to 21.8% increase in instances of binge drinking"

Considering that public transport in the US is generally weak, the underlying driving force seems to not be Uber itself but alternative means of transport for everybody.

Even in non-US, late night service can be spotty.

Especially if/when the metro closes.

Like late night taxis who realize you're a bit drunk and just desperate to get home to bed.

Uber has changed everything here. I can leave the bar at 3am and know I can get home without being sized up for extortion by a professional taxista who is colluding with the other late-nighters on the shift.

Is an increase in binge drinking necessarily bad? Americans dont drink much. Shows my historical roots.

You can't compare apples and oranges percentages. If drunk driving accidents went from 13 to 8 incidents (62% drop), but the 10% increase in heavy drinking is applied to 10,000,000 people, then the negative health effects could outweigh the accidents by a huge amount. I think your key words are, "seems small by comparison". You'd have to know the actual numbers.

But this is America, and self-harm is a fiercely protected liberty. Harming others is not.

If we could put half of the energy we put into policing other people's vices into policing Corporate vices, think of what we could achieve.

Should we not be studying the trade-offs?

Yes we should study the trade-offs. But that calculus should also include the fact that most people genuinely enjoy going out and getting drunk with their friends.

Yet rarely if ever does that term ever into the equation. I'm not saying it necessarily outweighs the public health costs. But a truly fair accounting of all the trade-offs should take the hedonic benefits into account.

Serial killers enjoy murder, but that doesn't mean we necessarily want to encourage it.

I'm all for increasing the safety of nights out, but if it's got a corresponding increase in harmful or dangerous drinking behavior, that's a significant finding too.

The post you were responding to was trying to correct an over-correction. You've now introduced a superfluous over-correction.

To put it another way - you've disputed a point that nobody was making.

Normally we make a clear distinction between people harming themselves and people harming others, also known as "your freedom to swing your fist ends at my nose".

Binge drinking regularly harms others.

Same effect in Austin. Very interesting to see how the metrics changed when ride-sharing apps were banned as well.

That one could attribute %62 of alcohol related collisions (and their associated injuries and fatalities) to the price fixing by cab licensing bodies is very interesting indeed.

Taxi companies and driver unions insist they provide safety and reliability. Appears there was an externality they need to account for.

Uber is priced the way it is because rich financiers are subsidizing rides. If Uber were running a profitable business, the rides would be as expensive or more so than taxis. Another problem with taxis is they don't show up when you call them, and you don't know what it will cost until the trip is over.

With Uber/Lyft, I never have to worry about being stuck outside in the cold, drunk, waiting 60 minutes for a $30 taxi.

> If Uber were running a profitable business, the rides would be as expensive or more so than taxis.

Might be off topic, but in downtown Chicago before/after work cabs (including tip) are often cheaper than Uber/Lyft (with no additional tip,) although they offer other conveniences (payment, tracking, can be "hailed" to anywhere, etc.) At less busy times ride sharing can still be a bit cheaper than a cab.

Was once a compelling theory, but in the last earnings (Q3 2019), Uber's ride hailing business was "adjusted EBITDA" profitable, with an increase in revenue growth. It's not hard to imagine them being GAAP profitable if they decide to sacrifice a little bit of revenue growth.

EBITDA is nonsense disguised to hide costs.

If you need to adjust it to appear profitable, then you are not profitable.

(Yes, I am aware that there are valid use-cases for EBITDA, but in general it is used as a way to avoid accounting for actual costs).

You can use fancy terms and fancy accounting to make it seem like Uber isn't subsidizing riders, but it doesn't change the facts.

What are you talking about? This has nothing to do with cabs.

You look at the rate of drunk driving accidents before availability of Uber in Portland. You look at the rate of drunk driving accidents after. The second number is 62% less than the first.

America was already prone to binge drinking.

Also, public transit also reduces drunk driving, so it’s hardly unique to taxis.

Anglo-Saxon drinking is the worst kind of drinking. In places where being fall-down drunk is not socially acceptable we drink differently and nobody is racing to puke on the sidewalk. It's not that difficult, just use moderation.

Binge drinking is defined as 4 drinks for women or 5 drinks for men in a 2 hour period [0]. That's hardly falling down, puking on the sidewalk levels of drunk.

[0] https://www.cdc.gov/alcohol/fact-sheets/binge-drinking.htm

That's the minimum amount i.e 4.99 drinks within a 2 hour period is not binge drinking. Also, I have a drink just about every day or every other day, and I haven't had 5 drinks within a 2 hour period in at least the last decade. It would cause me to fall down and puke on the sidewalk.

That's fair. I suppose I'm just used to my own (probably unhealthy) drinking habits. I only drink one day a week, usually at some event like a football game or wedding, but I drink a lot when I do.

I suppose that varies by person and by what qualifies as a drink, but if I had five pints of microbrew in two hours I'd be drunk, and done for the day.

Certainly varies by person. But there is a standard definition of a "drink": 1.5oz of 40% liquor, 12oz of 5% beer, 5oz of 12% wine [0].

[0] https://www.niaaa.nih.gov/what-standard-drink

You'd also have to piss like a race horse.

People over-attribute the impact of culture on drinking habits. In reality genetics plays the largest factor in problematic drinking. Just look at the epidemic of alcoholism in American Indians, despite that culture having no prior traditions of drinking.

The thing is Northern Europeans are probably more genetically prone to alcoholism than other European populations. For example British Jews, even the non-religious subset, binge drink at much lower rates than the rest of Britain. Predictably the population has a much concentration of genes associated with resistance to alcoholism.

[1] https://www.telegraph.co.uk/news/science/science-news/329933...

I admire your confidence.

Yea the tipping point for self driving cars will be when it can drive you home while drunk legally. Probably will be made legal first someplace like South Korea — there you can easily pay somebody to show up with their buddy on a scooter to drive your car home for you with you in the passenger seat. The guy on the scooter follows you and picks up his buddy. I forget what the pricing was but it was something like 2x the cost of a taxi. So wasn’t actually too crazy.

There was an affordable service that would drive your car home for you. The problem leading to its failure, as I understand it, is that drunk people that needed it were too wasted to remember that it existed.

That's funny in a sad way. I do remember that being a thing ~10 or 15 years ago - they were at least demo'ing it at one of the blocks in my city where a large concentration of bars could be found. I was very curious on the technology of their fold-able scooters (how quaint given Lime at al)

I could also imagine having to drive a drunk person's car home would be somehow risky, given that you have no idea how said person will act with their property.

Years ago Top Gear did a bit on a service with small, foldable motorbikes that could be folded and put in the trunk. However, it didn't really save money compared to just taking a taxi both ways.

I always found fascinating the "scoot" or "scooterman" (or similar) service, in the UK some cities have it.

You call them and someone will arrive on a foldable scooter, fold it and put in your car, drive you home with your car, then get his/her scooter and go away.

There are probably a lot of factors that have contributed to the increase in high risk drinking, but even if we ignore that, I feel like the trade-off being made here is acceptable.

People who maybe prior couldn't imbibe due to no public transport options etc, have decided now that with ride-hailing apps that this is an available option. Whether or not you agree with the choice they are making, the fact that they even have the choice seems worthwhile.

I'm very happy with a tradeoff where people damage their own health instead of innocent people on the road.

I also wonder how much of this increase in drinking at bars is drinking that would have happened at home previously.

Maybe now that there are viable alternatives to getting to and from home/work the law here in the States will catch up to saner legislation elsewhere and properly punish drunk driving instead of a slap on the wrist for the first few offenses “don’t do it again but here is your license, you can keep driving.”

Allow me to give the standard response about how driving is necessary here. DUIs are not cheap, not convenient, but someone shouldn't become unable to provide for their family based on one.

No, just kill the sole provider for another family instead, right? It’s only luck that stopped a dui from being a vehicular manslaughter charge.

I live in Ireland. Driving is essential here, unless you live in a major city.

Nonetheless, we clamp down hard on drink-driving, to such an extent that I was massively shocked by how acceptable it was in the US.

Sounds like Arizona here. Penalties are extremely harsh there.

found the paper: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3484845

I wonder about other changes as various things change.

Could self driving cars mean more driving? / car related demand?

I drive my family around all the time. I'm not a fan of it, but if my car drove itself I'd probabbly be quite happy to plan additional / longer driving trips. I'd certainly like to setup my minivan as a series of mini desks everyone could sit at .. I could do lots of things during trips...

My kids have a lot of after school activities which my wife spends hours a day bringing them to and fro, even when they are more than capable of doing the activity entirely independently and don’t expect to be watched by a parent during the activity.

For some reason it’s deemed “not appropriate” for them to Uber to and fro. Occasionally there will be parents who we can carpool with, but almost always it ends up being a huge time suck to drive them, wait around, and then bring them back.

I think it’s not so much an issue of them being in a car alone, it’s the random stranger driving them that makes it hard for a, e.g. 10-16 year old to get around on their own if they can’t walk, bike, or take public transit.

So my guess is that a self-driving car will be more acceptable to bring kids around the suburbs to all their after-school activities. Which is a massive productivity boost for the parents, basically freeing up an entire human for 4 hours a day.

Man I would love that.

I will note that at least in urban areas, I see kids ubered to school on odd days ... mostly at poorer communities where parents have to work on a random day and still need to get their kids to school. But yeah in the better off suburbs I don't see it happen ever... then again there is more carpooling for those.

There are actually some non uber specialty services out there that claim to do background checks, have extra training, and etc that focus on moving kids or even vulnerable adults around. I don't know how widespread those are.

If my kids could just show up at the door that would save me hours every week.

It's not uncommon for wealthier families here in Guadalajara to have their 13-year-old daughter Uber around.

We sure do like our coddling in the USA, though, where it's even safer.

An example of someone writing about the problem: https://www.thecoddling.com/ -- Makes a good case how instilling these ideas in our youth that everyone is out to get them comes back to bite them (and us all) in the ass as they enter adulthood.

An example of an org trying to fix it: https://letgrow.org/

Of course, a self-driving car is even better than an Uber driver for various reasons, not the least because ape-driven vehicles aren't particularly safe.

I feel like choosing to uber or not is a big distance away from "coddling".

I think SDC's will definitely increase VMT (vehicle miles traveled). Think of the talking points...you can just lay back, have a snooze, and in 8 hours you'll be at your destination. This means a LOT less barrier to long-distance travel, a LOT less barrier for long commutes, and increased attractiveness of cheaper housing/communities that are farther out. If the main barrier to more driving is time/effort, then SDC's will greatly increase VMT.

I commuted by train for a year and it was 2 hours each way. Even being able to do whatever I wanted, the commute was absolutely soul crushing.

I wouldn't want that for my normal commute. However when visiting family it would be nice. "Hey car: start driving towards my in-laws" [sleep] "hey car, find me breakfast someplace" [eat] "hey car, find me something interesting to do" [learn about the swamp spirits] "Hey car, find me something else to do" [learn about railroad that used to run though this little town]. "hey car, find lunch" And so on throughout the day, before letting the car drive another 8 hours while I sleep and waking up in time for breakfast at my in-laws.

I agree. Removing the other people (i.e doing it in a car) helps make it bearable but spending X hr/day in a box commuting is not good for quality of life no matter how you cut it.

Yeah my commute is kinda long now. I work from home a lot but boy a desk in my car... would be pretty nice / get a lot done.

According to the some preliminary studies, yes, people will use self-driving cars for more road hours than existing models. Apparently a similar effect was visible as ridesharing services became popular.

https://jalopnik.com/zombie-miles-and-napa-weekends-how-a-we...

Self driving cars that see my kid waiting to cross the street and stop to let him cross would mean my kid rides the bus home. He is 6 - I don't trust his judgement crossing the busy street the bus runs on. I hate crossing it myself - I often get impatient after waiting 3 minutes with no break in the traffic and run for the left turn lane I(middle lane) to wait for the other direction...

Jokes' on them, the Bay Area responded to this new challenge by having no place to go drink or have fun. Problem solved!

wait what? I live in Bay Area and can confirm there are a lot of places people drink and have fun.

This adds absolutely nothing to the conversation and is pretty provably false.

I will note one cheerful conclusion: so many people care so much about not drinking and driving. (Consider that 30% of the people don't drink, and another 30% drink rarely, these statistics are a credit to the species.)

I look forward to people pointing out the flaws in this reasoning.

I feel like there is a huge demographic issue underlying these numbers. The cohort of young adults in the US probably peaked around 2012, so more binge drinking is likely caused by that alone. (I didn't read the paper the article is based on so maybe my point is covered there)

Whilst Uber and the like have improved the cab situation in UK cities, it's hardly been revolutionary. It's become a bit more convenient and (sometimes) a bit cheaper.

What was the situation like in places like Portland before Uber? Were old-school cabs just unavailable or just really awful?

Not Portland but I lived/drank in Eugene for awhile when Uber/Lyft were still banned (IIRC they've since been let in). Outside of metro areas there is only one answer to that question in my opinion, lots of drunk driving.

> Outside of metro areas

I guess this is key. Most places in the UK aren't that far from a town.

Does "metro area" have a precise definition? Would an small to medium UK town count?

In this context it's about mass public transit options and walking distance between a safe residence and wherever "downtown" is. Eugene and adjacent Springfield combine for a population of ~220,000 over ~88 square kilometers with limited late night transportation options other than sketchy and unreliable taxis.

Imagine what fully self driving cars will do, uber without the cost

Why would you assume there is no cost? Either you hve to buy it, or rent it as a service. You are still paying either way.

through scaling the cost to the provider should be much lower, which means either the cost to the consumer will either be less, or there will be a great opportunity for competition.

Yeah it'll be great to be able to enjoy a nice bottle of wine on those long road trips.

Good thing there is no reasonable public transit in US and no plans to improve it. Imagine the amount of binge drinking that would ensue!

I always thought that once self-driving cars are fully autonomous, pub/bar business will go through the roof.

Concur in part. When I am on vacation under the right circumstances, such as multi-day Disney visit or cruise, I do drink more. Some people would probably drink more just because it would be safe, but I think a lot of people would dial it back again due to hangovers interfering with work.

Maybe not "through the roof" but definitely a sustained increase.

I should have said - I'm talking about the UK here. More of a drinking culture than, say, the US :-)

It's odd how each graph shows a dip at -1 year.

Correlation is not causation.

Alcohol consumption is horrible for public health though :(

Correlation does not mean causation

Can confirm, definitely see many more drunk assholes loitering outside the bars on my street.