The Opportunity Cost of Streets

Here from Alain Bertaud and the Urbanization Project is another way of thinking not just about the high cost of free parking but also the opportunity cost of streets. In New York City, a place with some of the most valuable real estate in the world, 26.6% of the land is devoted to unpriced streets (and an even larger percentage once we include parking). In Manhattan we go to great expense and effort to make it possible for hundreds of people to use the same 10*10 square feet of land, we build skyscrapers, and yet at the same time similar quantities of land are being taken up by a few people and their cars.

NYC land use 1
Hat tip: Brandon Fuller.


I wonder what percentage of the total is on street parking? I'd guess something around 1/3 to 1/2.

In any event, that's a massive amount of value devoted to cars and parking and a massive subsidy to drivers.

You do realize it isn't pleasant when the next building is five or ten feet away, dreary, dark, and not even smidgen of privacy. There is no way people would prefer the buildings get boxed in on all four sides just to reclaim the street space for buildings. There is no subsidy to parking here. The value is dedicated to making density livable by requiring clear views through and between tall buildings.

We've seen densities implied by using that 26% for buildings, we usually call them slums.

Very well said Jon. If people really want to live in a big, dense city without "wasted" space on parking and streets, they are welcome to live here in São Paulo and enjoy our overall, overwhelming congestion.
Here you don't have to worry about wide streets and sidewalks keeping you apart from up to 8 different neighboring buildings...

Higher urban density is always a nice thing to recommend to other people...


Did Manhattan have tiny streets in the pre-car era? Alex's reasoning seems silly.

People just like wide streets. If not to drive, then to walk, bike, let the sunshine and wind in or whatever else they do.

Yes, last time I checked "people just like" in a one person-one vote system leads to absolutely brilliant outcomes.

How does it follow that the inter-building space has to be used for parking? It seems like you are making a major non-sequitur. That space could be parks, bike lanes, bike racks, bigger sidewalks with more space for hot dog carts, urban farms, low-story extensions of buildings for retail stores, or a million other things.

Curt, I agree there is a trade off between streets and parks. Alex framed the issue as valuing the land used for either of these things by the value of more building density. That is wrong. Having non-building spaces accrues value to the building spaces themselves.

So now we agree to have open spaces, you have a choice between open space used by cars and open space used for parks. That's a very different value trade-off, one which arguably makes the opportunity cost of streets relatively low compared to Alex's insinuation.

Thanks for the reply Jon; I definitely agree that open spaces are good and that the vast majority of people would like to keep them. If Alex is arguing against that then it sure is a funny view.

Agree. In another form: It all depends what you consider opportunities.

Where is this free parking in NYC? Wherever it is, there can't be that much of it, given the high prices that private parking garages are able to charge.

I assume that one counts among the "few people and their cars" that benefit from "subsidized" streets: people that ride buses and taxis on those streets, people that receive trucked shipments from online purchases, customers of businesses that receive trucked shipments even if those customers walk or bicycle to those businesses, people that use car-sharing services like ZipCar, people that want police, fire, and ambulance services available in case needed, etc. I am all for privatization where possible and practical. One doesn't need to own a car, however, to use streets. Also, we should keep in mind that, at least at the federal level, autos are the one transportation mode that is not subsidized [].

When vehicles pay for the congesting they cause and street parking is dynamically priced to clear the market, talk to me about what people "like" about streets.

How is it cemeteries are not considered a waste of space ? Requires too much of a cultural shift?

or parks, or sidewalks, or trees, or bike paths, or baseball fields.....

Non-rival, non-exclusive? In the case of parks at least. To an extent.

Even cemeteries are more pleasant to behold and less rival/exclusive than a highway.

How is a park non-rival and non-exclusive? One could, if one wanted to, limit access to a park, hence exclusive. Also, if one is using space in a park for a picnic, for example, then no one else can use that same space at the same time, hence rivalrous. Doesn't the fact that sometimes one can or needs to get a permit to use a large part of a park, say for a concert in Central Park, demonstrate that the park is inherently neither non-rival nor non-exclusive?

Sounds like a free lunch, then. Just convert park space to streets.

As a driver who likes to drive fast and unimpeded, and who likes to park wherever I please, what is the opportunity cost of all those buildings in NYC, not to mention the stoplights and yielding to pedestrians and those pesky bicyclists (especially the middle aged and older ones in spandex).

Do it for the children, the future drivers!

As a cyclist who like to ride fast and unimpeded, I say get rid of the cars!

Cyclists paying for the 80% of the user fees and convincing the elected officials to kick in the other 20% is the beginning of a case to get rid of cars. Getting rid of the cars with no skin in the game is cheap talk.

"In any event, that’s a massive amount of value devoted to cars and parking and a massive subsidy to drivers."

Alternatively, it is a massive subsidy on windows.

Do you think that if we did not have cars that cities would be just one massive building?

Alex, if you truly think this: "In Manhattan we go to great expense and effort to make it possible for hundreds of people to use the same 10*10 square feet of land, we build skyscrapers, and yet at the same time similar quantities of land are being taken up by a few people and their cars." then perhaps you should advocate multi-tiered roads.

Cars exist because they are the most efficient method of moving random people in random directions.

Not in a city.

Walking, or on foot, is far superior. Followed by bikes. Then public transportation.

Foot, bike, car, then public transportation.

If this is true, then why are the streets of NY packed with cars?

Sometimes, walking is superior to bikes or buses or cars, oftentimes not. At any given time and depending on your needs, the order of superiority will change.

Would not the success of New York City rather suggest a revealed preference of people for a city that devotes 26.6% of the land to "unpriced streets"? Correlation does not imply causation of course and maybe they all want to live there despite the streets, but I can still not see how this statistic on its own could show anything. Wouldn't it be better to point to one of the many successful cities without streets? Oh wait...

Revealed preferences mean nothing in this case. Many of these streets were built ages ago by people who many would agree were misguided (Robert Moses, I mean).

I guess what I mean to say is: If the existence of something was proof of the desire for it, the High Line would still be a railroad track.

I personally like NYC with streets, but think they are used inefficiently. I think streets in NYC will get much smaller once we get automated cars.

I don't mean a revealed preference of the people who built the streets but of the people moving to NYC today. Now obviously nobody moves to a city for the sweet streets, but you do move to a city that comes with a package that contains streets and the lack of cities without that feature suggests to me that they probably add value in a way that makes the city attractive. Is 26.6% street the optimal level for NYC? I have no clue whatsoever, but if you want to make the case for less streets then it would certainly help your case if you could point to any growing, modern city that isn't a holiday resort of some sort that featured markedly less streets than NYC.

*it would certainly help your case if you could point to any growing, modern city that isn’t a holiday resort of some sort that featured markedly less streets than NYC. - See more at:*

Mexico City
Hong Kong

*it would certainly help your case if you could point to any growing, modern city that isn’t a holiday resort of some sort that featured markedly less streets than NYC.*





Mexico City

Hong Kong


(Edit: reformatted)

Brian's list is extremely dubious as anyone who has been to Paris (see Haussman's boulevards), Barcelona (check out the outer suburbs let alone Las Ramblas), Mexico City etc.

Randomly naming world cities in an attempt to win the blogging comment contest is lame.

I have a feeling Brian is one of those Americans whose contempt for his fellow citizens has caused him to make wildly incorrect assumptions about the rest of the world.

I'm glad you didn't include Madrid.
I've just been to Spain. My experience, including Barcelona, is that the streets are positively midwestern-America in their width. The Eixample (the HUGE late 19th C extension of the city) has ginormous streets with wide sidewalks - and the intersections are mainly (giant) octagons because of cuts back from the 90degree.

Madrid streets seemed even bigger. Paris - well, depends on the neighborhood. London streets are generally narrow.

I think the subways have a lot more with why people move to NYC than the streets.

Insert "to do" where appropriate.

There's an article in today's Washington Post about Northern Virginia officials exploring the use of air rights over highways and subway stations to build elevated office space. Notably, Northern Virginia is currently glutted with office space.

Does that mean cheaper contractors, or simply more contractors?

The first thought that comes to mind whenever I read some urban planning thing is the demolition of the chicago public housing. The externality that is the most costly is that the idiots coming up with these brilliant ideas don't have to live there, pay for it, build it our maintain it.

I'll throw in a simple question. How do you propose to get say a fire truck to close proximity to the building? Underground tunnels? How about building the whole thing, getting materials in? Helicopter? Are you willing to take personal liability for the costs when your brilliant planning forgot something and the lack of any redundancy or flexibility makes some simple thing very expensive because you can't get near the thing?

Maybe, just maybe the density its possible only due to a series of unusual factors, factors that grew organically and were responded to by free economic actors, and that a change in those factors have an even chance of resulting in a Detroit rather than a Manhattan.

I agree. I don't think there's really any value in being able to see the sun or the sky. How good are they really?

Then again, the real estate wouldn't be very valuable without good communications, i.e., streets and parking. (Snark aside, how do you put this intuitive idea in the language of economics?)

People - we have a system to decide when streets are valuable compared to offices or other uses. The price system. The problem is its lack of use for streets.

Why not buy up some land yourself and start building this streetless utopia on it, if it is superior then surely market forces will cause people and organizations to flock to it

So, in your world, mike, pricing streets means streetless? Weird how so many pizzas get made considering they are priced, isn't it?

Yes... the comments so far have been remarkably low quality. I was expecting someone to mention applying a land value tax in the city, including the land managed by the DOT. I guess the market failures will continue....

Offer us a counter-factual in form of land use percentage of a city you like.

How much is the street percent of say Berlin or Munich or Copenhagen?

Exactly – unfortunately the price system is dominated by the almost entitlement attitude we have towards parking, and roads in general.

What is up with Indians labeling every government program that benefits middle class white americans as an "entitlement"?

AAM syndrome, I imagine.

What's up with middle class white americans labeling every government program that benefits poor people an "entitlement?"

Everyone I know calls that "welfare"

It is a term of art:

The streets are priced by the owners of the streets, but the streets are not generating a profit in cash for the owners, either in dividends or Wall Street pump and dump.

You just object to the shoe shine guy on the corner having the same ownership share as the billionaire Bloomberg.

You just object to the shoeshine guy getting to walk and bike his streets while impeding your car driving through his streets. And that is after you paid the troll $10 to cross the bridge to drive his streets.

I would note that Wall Street loves to sell US roads, but they don't have much luck finding US buyers. If NYC streets were privately owned, I'm pretty sure they would end up being owned by Asians. That is because Asians have long timelines on their investments instead of the US pump and dump focus on short term capital gains, and Wall Street's desire for asset churn to generate fees.

Sigh. It would make these debates so much less tiresome if instead of falling back on the boring old moralistic trope of the "price mechanism," we could just cut to the effective chase: this is yet another plea to give more of the world to the rich and less to everyone else.

Let's follow the reasoning. The people on the streets of Manhattan are on average much poorer (and non-resident) than the people in the unaffordable apartments next to the streets. So they are being "subsidized" with underpriced access to semi-breathable semi-fresh air and a sometimes-view of the sky. If we get rid of the streets, the rich will fill them with more apartments that only they will be able to afford. Due to their floor-to-ceiling windows, skylights, and penthouse gardens, the rich in the new apartments won't miss the air-access of the old streets. Because they have so much money, we have arrived at the economic optimum. What a moral victory.

I'm now extremely hopeful that distributional analyses like this will feature prominently in all of Alex's future pleas for greater use of the free market.

In most areas it is more effective to associate the initial (and sometimes on going) street costs with the private asset. The land my home is built on is made more valuable by having a road, sidewalk, water, sewer, cable tv, electrical power and gas in the street. The developer paid to have these installed and then charged me as part of the price of my lot. Note that the street contains a significant amount of priced amenities. Also note that on a low volume street that the cost of collecting road use fees exceeds the benefits.

The big city is highly similar. The utilities are still in place and priced. The streets are fundamental to the current valuations of the buildings. Note also that on net that people pay the full price of congestion with there own time, and it is only the differences in valuation of time that causes congestion externalities.

I believe it is too strong of a statement to state that streets are un-priced. I think it is more effective to consider street as not perfectly priced. Improvements may be possible but there may be many more rent seekers than governments can overcome to implement congestion pricing effectively.

Anyone who lives in Northern Virginia and can afford it pays a high premium to live in a suburb with trees, roads, and lawns, uncrowded. The politicians and wonks have fantasies about wonderful new urban centers with towering highrises and lots of trains. It's a planning sickness. I can remember paying much more per square foot to live in Oakton rather than, say, Falls Church (circa 1981). Even in cities a "low city" is prized. High rises are nice just as long as you can see over the next buildings. The price system really only expresses the opinion of the relatively rich. Look what they are willing to pay for, i.e. 50 million to look at Central Park rather than some looming apartment houses. Economists themselves tent (in my limited experience) to prefer nice suburban or exurban digs.

Alex - we don't price air or sunlight either....

Streets are like blood vessels - you can be unhappy about how much space they take up, but you will be even less happy without them and their function.

And the REAL question - why do we have to cram so many people into such a small space? We don't/can't we develop ways to get the "benefits" of cities without require high density?

Because the "benefit" of cities is precisely the density and the ensuing high real estate prices that exclude undesirables in a way that residents can plausibly deny intending.

Streets on the surface would be better if green space.

NYC has turned an abandoned elevated railroad line into a park - the Highline. The rail line served industry that has left the city, built using air rights over the streets. The "market" solution was to tear it down for the scrap metal, freeing up the air rights for resale. (Built as a utility, the rail line did not buy the air rights, but was granted them as long as it was a utility - when abandoned, the property reverted to the city.)

Activists got the Highline converted into a park, which had a significant price, but also with a significant and growing benefit and revenue yield to the city. On a ROIC basis, providing a park will probably have a 10-20 year time to breakeven.

But in terms of improving the city for people, only a few still object to the Highline.

You should see the street on a swimming and dinner club near our house in Chennai. Not only do they park on the street, but they park the car [i]orthogonal[/i] to the side, eating about 80% of the space.

Mind you, this is in a country 75% of the people walk, and building roads is not socially optimal to begin with.

Errrp. Also, something I've always wondered: how does one italicize text! Thanks.

I love your gems of wisdom.

In India, "building roads is not socially optimal to begin with"

Really?! You have an argument for this assertion? I'd love to hear.

Sidewalks would increase throughput much more. You should open your mind and think more. I understand there are some cases for building roads, but it looks like you've already jumped to that conclusion. Surprise.

But as it happens I've spoken with quite a few urban architects and developers over the past few months, all of whom are surprised at the vapid lust to build more roads, when a mix of cycle and walking lanes would actually improve speed and throughput; not to mention cost.

Maintaining roads is far more expensive and since you're the one advocating an activity (I'm just saying between borrowing more to build roads and do nothing; choose nothing) I have the null. It's for you to tell me why it's socially optimal, and yet unthinkingly you've just assumed a conclusion.

Maybe the streets could all be made indoor with proper ventilation (or a ban on non-electric cars) and suitable optics to collect and distribute sunlight?

And then ultimately you'd have a single city-wide building, solving the problem.

How about a little something like this:

(Link goes to a photo of Buckminster Fuller's plan for a geodesic dome for Manhattan.)

What is the unaccounted for 12.4% at the top of the page made up of ?

Streets are clearly priced in NYC, because they are paid for in many ways. Gas taxes, registration fees, parking fees, tolls, property taxes, income taxes.

What streets are not is owned by a Wall Street corporation which would be virtuously charging higher and higher prices for using the streets which would then drive the price of the streets higher and higher. And one way to drive the price higher and higher is to spend less on the streets based on the idea that the monopoly of owning the streets means the quality does not matter, and it is better to fix just a few streets and set pricing on those that is ten times the price of the rest, concierge street pricing. And if the streets can be reallocated for other things, then limiting the supply of streets will mean the price charged can be higher, so the Wall Street price goes higher and higher as the profit motive liberates the value of the streets.

After all, the market is always the best determinant of the optimal.

Limiting NYC to only the rich would be the ideal because they would pay the highest price for the streets and that would maximize the market cap of NYC Streets Inc. It would ensure that NYC was limited to 10 million millionaires and billionaires - no poor or working poor or middle class allowed by pricing them out of NYC.

Well, the opportunity costs of streets and parkings seems to skyrocket, when the values of homes and buildings rise. But what would the values of homes and buildings be if we had no streets and parkings?

Interesting debate.
The streets in Singapure are within the building and shopping malls. A nice way to get from place to place in an air conditionning environnement.
Same in Montreal and Atlanta. But I prefere the streets in the Quartier Latin in Paris.

If you put everything close enough so that only pedestrians could wend through, you'd be getting close to 19th century fire-spread risks (not to mention, how do fire trucks or ambulances get to all those skyscrapers?)...

Part of what makes that Manhattan real estate valuable is that you can get there. By street.

Alex, what is the point of this post? So what that not all assets, like streets, are priced. Streets are a paradigm example of a "public good", so of course we should expect a non-optimal level of streets.

By the way, one of the things I love about some of my favorite cities, places like Havana, Tangier, or Rome, is the mix of wide avenues and narrow lanes

Call me crazy, but I betcha a lot of that valuable land would become lots and lots less valuable if it became harder to get to and from it...

Exactly. The WTC was always a tough sell before 9/11 because you couldn't easily get there.

Positivism is a congenital disease most frequently contacted in macro economics courses. :)

Density is "bad". There are diminishing returns on density.
Diversity is "bad". Homogeneity increases trust, and trust investment in commons. Diversity decreases trust and investment in commons.
Trust and spatial proximity are correlated. "Good fences make good neighbors."
Good commons require good private property.

All of these linguistic expressions are usually considered irrelevant externalities, simply because they are hard to measure and invalidate ideological assumptions. They are however "facts". ;)

If we produce grad students and PhD's who produce the above cited work, then efficiency would dictate that we produce too many grad students and PhD's, not too many parking spaces.

At least that is the logical deduction, no?


There is a revealed preference for dense cities. Land prices there are high. Losing population density has not made Detroit more attractive.

Detroit proper lost population because there is not a revealed preference for dense cities.

One guy implemented a streetless "utopia" in Sim City. Maximum population, no air pollution. Watch it to see what kind of person builds that kind of city.

How much of this "wasted space" is located on Staten Island or in Queens? NYC is not just Manhattan.

How do we imagine the allegedly-superior streetless city is going to come into existence? Because all the answers that don't involve either magic or giant zeppelins involve, as a first step, lots of big trucks full of construction materials. And while giant zeppelins are of course ineffably cool, they are a bit on the fragile side and never have been terribly economical as a means of bulk transportation.

So here in the real world, it is an absolute requirement of urban infrastructure that it include a pervasive network of streets suitable for large trucks carrying the raw materials to build that urban infrastructure. Given that requirement, the most cost-effective way to deal with the subsequent logistical requirements for ongoing delivery and waste disposal is with smaller trucks using the same street. And at that point, the marginal cost of enabling the streets to also handle cars, bicycles, pedestrians, and open sightlines, is small.

People who think the streets are for the cars, that the streets can go away if we can get rid of the cars, are missing the point.

So what exactly is the proposed solution? If you sell half a street who would buy it? They can't exactly build and block other entrances. Would current building owners tear down and rebuild out 20 feet onto a narrower street or pedestrian mall? That seems impractical. And if you do this for one block and not the rest you create unsightly lines and bottlenecks. It seems that when the original grid was laid down it fixed things for a long time. Maybe when wholesale re-building developments come up in the future narrowing or discarding streets should be considered. But somehow, despite the large quantity of streets NYC just happens to be extremely successful. Perhaps the 26.9% street quantity is part of the recipe for that success?

Comments for this post are closed