Hacker News Reader

There are serious attempts to estimate the externalities that arise from automobile use in general: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.661...

Adjusting those estimates (only) for inflation, they work out to about $2.14 per gallon in driving externalities.

Note that these estimates are attempts in general, not in the context of Manhattan, or downtown in any other city.

I'm not aware of any studies that attempted to estimate them in such an environment, I think that local factors (basically density of people) change the calculation substantially.

So.....even if you add it to the price of US fuel, it's still cheaper than fuel in the UK? And British cities have just as much trouble with congestion as American cities do. I guess what I mean is that if this extra money is meant to be a deterrent to driving it's just not enough. I'm not sure what the price would need to be to be honest. I've been to Portugal recently and the price of petrol is just insane - in the order of $8-9/US gallon. And yet people sitll drive, larger cities still have loads of congestion. And Portugal is very poor by European standards.

If this is a reasonable calculation, then I don’t think the costs are so insanely high.

While it looks like a reasonable attempt, it’s only looking at directly related effects. For instance, it notes that pedestrian fatalities have decreased since 1960 due to, among other things, discouraging people from walking. The effect of cars on population health would be much harder to quantify but might end up dwarfing many of the other factors.

It's circular logic to assume the externalities are ridiculous, therefore we must price cars into the realm of ridiculousness. If you want your argument to be more than cheering against cars, you should actually calculate the externalities and you should do so for competing modes of transportation. Heck if you're talking about noise pollution, someone dribbling a basketball is far more distracting to me than the white noise of cars.

Also note that you are only calculating the costs of cars to others, not the benefits. Cars allow workers to spend less time commuting and be more productive at work, which benefits strangers. They allow people to travel farther to make purchases. They encourage the purchase of larger objects that can't be carried by hand for long distances. Their gas taxes help pay for the roads that people walk and ride bicycles on. Their tolls pay for the maintenance of bridges that bicycles and trains travel on.

I'm a fan of congestion pricing, but the purpose of congestion pricing is to reduce congestion, not to be a sin tax on driving. If people time shift their driving, that's fine.

> It's circular logic to assume the externalities are ridiculous, therefore we must price cars into the realm of ridiculousness.

You're missunderstanding me, I'm just stating the possible conclusion that makes this an interesting argument (that it's ridiculously high and no one would pay it except on rare occasion) before the logic (look at all these externalities).

> you should actually calculate the externalities

That sounds like inviting lots of nitpicking debate on the internet. I invite you to though. If you do so in good faith and come up with a number that is reasonably low, well congrats, you shouldn't find the argument convincing.

> you should do so for competing modes of transportation.

True - I haven't personally because I think it's intuitive that they are far lower, if you disagree feel free to include them in your version of the calculation.

> you are only calculating the costs of cars to others, not the benefits.

I struggle to think of any positive externalities for personal cars. Everything you list are benefits to the owner except tolls. Tolls naturally don't count because that's the mechanism by which we ask them to pay for their externalities, to count them as a positive would be double counting.

If you think there are some compelling ones, again, feel free to include them in your version of the calculation.

> a sin tax on driving

Pricing in externalities is not a "sin tax". A sin tax would be "being lazy (driving) is a sin so we're going to tax them for driving".

Instead of nitpicking, we can just look up the actual calculations.[1] They're on the order of 20 cents per mile traveled. Half of that is crashes, which insurance already covers. Most of the rest is congestion costs. Air pollution and global warming costs are actually quite low compared to the harm caused by crashes and congestion. Noise is also a rounding error.

Of course this is aggregate data. Congestion costs are higher in cities and higher during peak driving hours, so it makes sense to have dynamic pricing for that.

1. Quantifying the External Costs of Vehicle Use: Evidence from America’s Top Selling Light-Duty Models (http://www.ce.utexas.edu/prof/kockelman/public_html/trb08veh...)

If you can find any studies that take the context of Manhattan (or anywhere else in the first world with similar population density really) I'd be happy to look at them. The study you're looking at doesn't do so.

So while I assume that it's a perfectly valid study it really doesn't apply.

I could use that same logic to refute a study showing public transportation is cost-effective as long as that study wasn't done in San Francisco. It would be an obviously incorrect argument in that case, and that is also true in your case regarding cars.

The study's numbers are far more connected to reality than your made up numbers. You are making an isolated demand for rigor.[1]

1. https://slatestarcodex.com/2014/08/14/beware-isolated-demand...

I'm not making an isolated demand for rigor, because I'm demanding everything we do does not include clear systematic biases in one direction. I'm making a global demand for the most minimal amount of rigor you could possibly use and not necessarily arrive at incorrect answers.

Your study has a clear systematic bias (edit: Actually it doesn't even say what you claimed it did, but if it did it would have a clear systematic bias), that most costs scale with population density, and it was done studying an area (the US) with a much lower population density than the area at hand (manhattan). Assuming your study is entirely correct, its answer is incorrect when applied to the problem at hand. You would have to be an idiot to rely on it because the only way it is "right" is by being incorrect by the right amount in the right direction by sheer chance.

Likewise, if you tried to apply a study about public transportation done in the middle of the Sahara Desert to San Francisco, I would call you an idiot.

The study your citing doesn't say what you are claiming it does.

> It is also important to recognize that the five external costs calculated in this analysis do not encompass all possible external costs associated with vehicle ownership and use. For example, noise pollution imposes significant costs.

Not only does it not even attempt to put a total cost per mile - which I read as the main reason why you cited it. It directly contradicts you when you say "Noise is also a rounding error".

It specifically excludes some of the biggest costs

> Based on his review of the transport-cost literature, Litman (2002) believes that the largest external costs of automobile ownership and use relate to land use impacts [...] However, these costs are, to a large extent, indirect. In other words, they are a result of auto-oriented travel patterns and will not vary much by vehicle type, make and model. Thus, they are not examined here.

Another quote to make it extra clear that this study really doesn't say what you are claiming it does

> Many other external costs associated with driving exist, and many of these have been characterized at an aggregate scale, rather than by make and model. For instance, [...] disposal costs. Water contamination [...]. However, like the noise and land use costs described above, these other costs are difficult to distinguish by make and model. Thus, they are not covered here

A quote explaining what the study is about

> While certain studies have analyzed other types of external cost, beyond the five quantified here, these studies have been aggregate in nature, focusing on passenger vehicles as a whole, or cars versus light-duty trucks. In contrast, this work looks at likely variations across specific makes and models.

Your study is based on a land value of $2 million per acre.

My first Google result shows that Manhattan is, at its most expensive, $90 million per acre: https://cityroom.blogs.nytimes.com/2008/05/13/2100-a-square-...

This matches my intuition that cars are much more massively subsidized in city centers (where they should be heavily discouraged) than in suburbs (where they're mostly fine).

I also think roads have way higher value than other land, because they connect spaces. Think of it this way: you can build 99% of a wall, and I can still cross from one side to the other, but if you build 100% of a wall, I can't. So that last 1% of a wall has significantly higher externality cost than the first 99% of a wall.

Similarly, the existence of a road makes it impossible to cross a street without waiting. It's the last 1% of the land necessary for going from one part of a city to another, and so it's way more valuable than the price per acre would suggest.

Your study also doesn't include infrastructure costs, and also costs in inconvenience to pedestrians.

Your study also estimates congestion cost based on the assumption that 10% of travel is under congested conditions, but in Manhattan this is closer to 100%.

All your study's costs are conservative estimates.

In conclusion, I completely agree with the overall conclusion that car usage in most of the US has relatively low externality costs. I just massively disagree that car usage in a superdense city center like Manhattan also has relatively low externality costs.

Interesting point. We don’t even need to talk about basketballs. I find buses in New York to be a lot noisier than personal vehicles.

Even if you take into account the multiplied factor of personal vehicles that would otherwise (possibly) be on the road, the sounds would be much more spread out.

When I’m having an al fresco meal in the city, I never notice the cars. The snarling and snorting buses interrupt conversations.

You only need to visit a busy Manhattan street corner during commuting times to immediately realize how inapplicable those calculations are. They’ll be hundreds, if not thousands, of people waiting to cross but they only get the light 25% of the time. The rest of the time and most of the space is devoted to a tiny number of people in cars each taking up a huge amount of space and time per person.

What would the charge per car have to be to cross that street corner to possibly equal out all the time being wasted by all those pedestrians (times their hourly rate) waiting for those cars? I have to imagine it would be in the thousands of dollars per intersection or more.

There aren’t very many people in the world whose time is so valuable that it outweighs hundreds of other people’s, especially given that a random person trying to get to work in Manhattan him or herself is likely to be decently productive on average.

> Noise pollution.

How about we start taxing loud talkers?

> Pedestrian safety. Congestion.

And people who take up the whole sidewalk.

I don’t see people arguing for extreme measures, to the point of “ridiculousness”, to improve the state of walking around cities. So I don’t take this too seriously, it doesn’t seem to be done in good faith so much as, “I don’t like cars and there are some negative externalities, let’s use this as an excuse to really screw over anyone who thinks to use one.”

And of course, this is how a coastal elite most likely feels meanwhile they think traveling by airplane is okay or should even be celebrated.

It’s not in good faith.

Lots of people advocate for “superblocks” in lower Manhattan, where thru-traffic would be entirely banned and otherwise traffic slowed to something like 5-10mph. That is drastic, but very much in line with improving the state of walking around.

Admittedly a big focus on reducing cars in major cities likely is coming from the fact that the wealthy started moving to major cities rather than commuting from suburbia. Other people car commuting really does hurt their quality of life. But that seems more an argument that we should bring these policies even to poorer dense urban areas rather than limiting their roll-out only to Manhattan and other wealthy enclaves.

When every subway stop has elevators, then we cannot talk. Until then, try transporting kids or groceries using the subway. When the sidewalks are cleared of construction scaffolding, bike share racks blocking sidewalks or crazy grown adults dressing like Spider-Man, or psychotic homeless harassing people, then maybe we can talk.

As far as climate change; if NYC went to zero CO2, it wouldn’t make a single bit of difference compared to a single day’s worth of pollution compared to a small Chinese or Indian city. Perhaps a tariff on all goods and services from China or India would be more effective than banning cars in NYC. Banning cars from NYC would have a cosmetic effect on climate change and nothing more. And, in case you weren’t aware, cars already pay huge costs in NYC. It costs over $10 just to cross a bridge or tunnel into NYC. They don’t pay a toll to leave NYC: so someone is getting money just for a car to enter the city.

> try transporting kids or groceries using the subway

Err. I was transported plenty as a kid on the subway without using elevators, so were many other kids.

Groceries... generally you just walk. I live in a less dense city and it's never been a problem walking to the grocery store. Including for my parents when I was a kid.

> sidewalks are cleared of...

I mean, getting rid of cars means more room for sidewalks mean they are defacto cleared of this stuff. The reason it's a problem right now is they are packed into the same 5 feet as the pedestrians.

> As far as climate change; if NYC went to zero CO2...

I suspect you are right. I put what I consider the least compelling reasons at the bottom.

> cars already pay huge costs in NYC. It costs over $10

$10 really isn't that much money... as demonstrated by the number of people who pay it. Understand that when I say the right number might be "ridiculous" I do mean "ridiculous". I'm avoiding saying specific numbers because then I have to justify them, but think thousands, possibly large numbers of thousands.

> Until then, try transporting kids or groceries using the subway.

I live in Atlanta, a place the NYTimes often writes about our lack of income mobility because of a lack of mobility[1]. I also live 2 blocks from the southernmost transit (midtown) connected grocery store that gets a lot of business from the poorer south side of the city.

So, when you say "try transporting kids or groceries on the subway" I (and transitively, the nytimes) ask, which kids? because it's not the 20% of Atlantans that are low income and that have kids and multiple hourly wage jobs that can readily afford the costs of a car -- and they're already doing it, in a city more sprawled than NYC and other cities.

[1]https://www.nytimes.com/2013/07/22/business/in-climbing-inco...

try transporting kids or groceries using the subway.

When I lived in an area where I could commute by subway, I had a full service grocery store and several smaller specialty markets (a produce market, a butcher and a bakery) on my way home (well I had to get out one stop early to pass the butcher). I see plenty of parents with kids on the subway -- though none of them have the full-size strollers you see coming out of the backs of SUVs.

So instead of making one large shopping trip every week or two, I'd stop several times a week on my way home.

Though now I just buy groceries online and have them delivered, in the rare cases where I do want to go to the supermarket, I take my bike.

  Though now I just buy groceries online and have them delivered

... which requires more road use than people stopping by the store on the way home. (Round trip vs. one way)

Does it? A truck uses around the same road space as a car, so if it removes one person in the neighborhood's round trip to the store (not everyone has a grocery store on their way home from work), then it's a net win. Even if everyone makes a small side trip to go to the store on the way home from work, that one truck can replace ~50 side-trips to the store.

Though in my case, I usually alternate between bus and bike to work, so if I did drive to the store, it'd be a round trip. If I bike to the store, it means a round-trip from home since I usually have one or 2 commuter panniers on the bike when I go to work, and don't want to haul the empty trailer all the way to work and back.

  Does it? A truck uses around the same road space as a car

Well, if you ignore fuel consumption, pollution, and parking impact. Anyway...

  so if it removes one person in the neighborhood's round trip to the store

No, I was referring to the specific "on my way home" use case you referred to... which is the typical case. If you stop by the store on the way home, you aren't adding a trip in either direction. Having delivery adds two trips, from and to. Two trips vs. zero.

Depends on zoning. If each customer has to go a mile out of their way, but each customer's home is within a mile of the next, the traveling salesman route (start at the store, visit each home, finally return to the store) is shorter and doesn't need to happen during rush hour.

"No, I was referring to the specific "on my way home" use case you referred to"

I don't think you read my entire comment: "When I lived in an area where I could commute by subway, I had a full service grocery store and several smaller specialty markets (a produce market, a butcher and a bakery) on my way home"

I was referring to the comment asking how to shop by subway -- the answer is to shop in smaller increments, but now that I bus/bike to work and no longer walk past a grocery store on my way home, I just have groceries delivered.

Your answer seems to be saying that the answer to shopping on the subway is to not take the subway and drive to work and stop at the store on the way home.