Denser Cities are Smarter Cities

Excellent piece by Ryan Avent in the NYTimes:

The American economy’s famous upward mobility rested in part on middle-class access to rich, entrepreneurial cities. This machinery is breaking down, however, mostly because upward mobility strikes too many residents of rich places as too messy a pursuit to accommodate. During the Industrial Revolution, for instance, millions of workers flooded into fast-growing cities. This produced slums, but it also allowed poor workers to take advantage of opportunities in new industries, a process that helped create the middle class.

Rapid urban growth would mean denser neighborhoods, which makes many Americans uncomfortable. Preventing this density, however, denies workers access to the best opportunities, constraining the mechanism that helps support a strong middle class.


From the article:"The average Phoenix home is worth about 30 percent of the price of a house in San Jose." I'd say the Phoenix home is worth about 50% more than the San Jose house, it just costs 30% of the San Jose house. I wonder why someone would move.

Desert wasteland vs. Silicon Valley bay?

It is well known that all humans have exactly the same priorities.

Preventing this density, however, denies workers access to the best opportunities, constraining the mechanism that helps support a strong middle class.

Top five in population density in descending order: Bombay, Calcutta, Karachi, Lagos, Shenzhen. Yep, that the model future of the new USA.

There's no doubt some sweet spot in the middle. After all, the Alaskan tundra and the Australian outback aren't exactly ideal places to live either.

He specifically addresses that in the article. Check the 9th paragraph.

All of which have growth rates several fold that of the American economy.

Increased density isn't necessary if transport can be improved. For the hypothetical Vietnamese restaurant, what matters is the number of people (customers, employees, suppliers) within a 20-minute journey; not the number of people within a 1 mile radius. If roads or public transit are fast, uncongested, and reliable, then many more people can be within a 20-minute radius without having to live closer to the restaurant.

Furthermore, density isn't always good for business. Innovation can be hampered if start-up costs are too high. Apple started up in Steve Jobs' parents garage; same for Microsoft, Amazon, and many more. Without garages they would have had to rent office space, eating into precious start-up capital.

Many European cities, particularly in Germany, show how to combine well-planned transport networks (roads and public transit) with moderate-density living.

this is why automatic cars are transformative. Less congestion = everything gets a larger radius of usability.

But being able to drive a few feet closer together isn't transformative...its a mild improvement at best. AI driven cars have benefits far more on the side of safety than they do on the side of congestion improvement. The absolute best we could muster from AI driven cars would be if there were a massive mode switch to taxis, which foster higher utilization rates for cars. But that is just a different form of public transportation.

The estimates I have seen suggest that automated vehicles could at least double the capacity of existing roads. It's not simply that they would allow much smaller distances between vehicles. They would also eliminate the kind of sudden maneuvers that human drivers make -- sudden acceleration and braking, unexpected lane changes -- that disrupt traffic flow and cause accidents. This would create a huge benefit in reduced congestion.

And automated taxis would not be merely a different form of public transportation. They would be a radically faster and more convenient form of public transportation, allowing their users to travel much greater distances in a given amount of time than today's bus and train users can.

They could double the capacity of roads, but given the tipping-point style dynamics of traffic congestion, they could only do so if every car were automated. It only takes one car to cause a traffic jam. In any other condition, the benefits are drastically less favorable.

Considering the psychological barriers to having a faceless computer driving you in the worlds most dangerous device ever invented, you might as well go around telling people that teleportation will be transformative...because we aren't getting to 100% automated car penetration rates for another 100 years at least.

The benefits of automation are incremental. The more automated cars, the greater the increase in road capacity. And cars don't have to be fully autonomous to produce this benefit. Partial automation technologies that are already available, such as adaptive cruise control and lane departure warning, help to reduce accidents and traffic flow disruption. So do GPS navigation and traffic condition information systems. They can reroute cars in real time to avoid congested areas. Your prediction that it will take 100 years for full automation seems highly pessimistic to me, but we don't need full automation to produce dramatic benefits anyway.

Exactly. My own city (Ann Arbor) enjoys about 5 times as many restaurants as could be supported by 100K residents. That's possible because a significant fraction of the ~5M people in SE Michigan can drive here in 20-30 minutes.

In dense cities it takes much longer to get around even though the physical distances are shorter. So New Yorkers spend significantly more time commuting to work than the average Americans (who live in a lower-density areas):

Of the 231 counties with populations of 250,000 or more covered by the ACS, Queens (41.7 minutes), Richmond (41.3 minutes), Bronx (40.8 minutes) and Kings (39.7 minutes) – four of the five counties that comprise New York City – experienced the longest average commute-to-work times. Additionally, workers living in Prince William County, Va. (36.4 minutes), and Prince George’s County, Md. (35.5 minutes), – suburban counties located within the Washington, D.C. metro area – also faced some of the longest commutes.

Well, the large amount of young people to provide good, cheap labor helps.

Your high restaurant supply has far more to do with the age of residents than it does with transportation.

Having plenty of young people around to wait and clear tables doesn't help if you don't have the customers -- the point is that easy transportation by car is what enables the restaurants to serve a customer base that spreads over a much greater area (and population) than the city itself. I contrast our relative embarrassment of riches with Megan McArdle's lament over the apparent death of the one-only handy sit-down restaurant in her corner of DC. Somehow, density doesn't seem to be doing the trick in that neighborhood, and lack of high urban density isn't a big problem here.

Young people don't just wait and clear tables, they actually eat at restaurants...And by eat at restaurants, I mean they do it at much higher rates than other demographic groups. Young, childless, and single demographics are among the most desirable in the restaurant industry, because they eat out so much more than older, married families.

Your restaurant scene success levels can be found in other places that are extremely dense, congested, and low on parking. They just have to have the right demographics. Take for example Cambridge MA or Westwood in Los Angeles for example.

Young people don’t just wait and clear tables, they actually eat at restaurants…

Sure. But the student-oriented restaurants here are cheap and near campus, not higher-end and located downtown. The latter cater to older, wealthier customers (many of whom drive in from outside the city).

Maybe you haven't noticed, but young people go to nice restaurants too.

Regardless, I know what you are trying to say and I still disagree with it. I used to do demographic travel range analyses as a consulting service. You are correct that it doesn't matter how many people live within a certain distance. You are incorrect when you say it matters how many people live within a certain time range. What real restaurants want to know when they open a new location is how many people of their target market live within a time range.

If the only thing that mattered was the number of people within a time range, your favorite restaurants would be located in Livonia...not Ann Arbor.

Yuppies. Me. I've 27 and although I've recently added a girlfriend with a kid to my life, we still manage to eat at the nicest restaurants available, because food matters to me. I'm not the only one and I can guarantee that other yuppies are like me. We are not eating the lobster and steak with a bottle of Dom Perignon, but we are there having a salad and tapas every week. Last summer I was there every other day with friends, co-workers, or on a date. Yes my food budget is disproportionately high. No it's not a problem for me.

And lots of young people with money from home and institutions riding the crest of a student loan bubble. That helps too.

Well, yes, of course, the university buffers the local economy, but that has little to do with the point I'm making is about mobility/density. 'Effective' density shouldn't be measured by people-per-square-mile, but rather a combination of population and mobility. So if you want to know about the potential customer base for a restaurant, the important question is not, "How many people live within an x mile radius of the restaurant?", it is "How many people could travel to the restaurant in less than x minutes?" (And you'd also want to know about the cost, convenience, comfort, and perceived safety of that travel. Private auto travel in non-congested areas wins on all those measures as well -- though it loses on the need for a designated driver).

Yup, you can never find a place top eat in NYC.....

I mean, being in a residential neighborhood in northern Manhattant recently, I had the measily choice of 1 Dominican, 1 Salvadorean, One Chinese, One Indian, one Pizza joint, one Greek Dinner, and two contemporary American restaurants within a 5 minute walk. If I had extended my walk to ten minutes, my choices would have increased only by two more Chinese takout places, One Japanese, One Thai, an Italian restaurant, A wine and beer bar, a couple of lounges, Another two pizza joints, One McDonalds, another dinner, and a cafe.

If only I could live in a restaurant haven like Ann Arbor,. I mean, I go to Midtown, and you know what? Not a single place to eat! Damn those horrible commutes!

Apple, Micorsoft, amazon and many more businesses started up in Steve Jobs' parents' garage?

So the sprawling estates of the mega-rich have a negative impact on their access to entrepreneurial activities? Of course once at the top there's nowhere to go but down.

Population density is a significant explanatory variable for just about every social ill we have. The concentration of low-skilled people for cheap labor is a breeding ground for crime, pollution, gangs, political exploitation, labor unions, socialists and other forms of depravity.

I agree with Andrew's comment on transportation obviating the need for concentration. The US, China, Brazil, and Russia still have vast expanses of land through which we can spread our populations. Through expansion and better transportation and communication, we can improve quality of life, expand the middle class, and eliminate most social ills.

The studies Avent cites suggest that the effect of density on productivity is small. A doubling of density is associated with a rise in productivity of only around 6%. Higher density is also associated with many costs -- more noise, more crowding, more congestion, more pollution, less privacy, and others. Why should we believe the small increase in productivity is worth these costs?

And in fact, it appears that the supposed higher productivity of denser cities may be an illusion anyway. The studies measure differences in productivity as differences in wages and incomes. They do not appear to take into account differences in the cost of living. A $70,000 income will go a lot further in Houston than New York.

Cities are dense because they are centers of productivity. People that choose to live there put up with the downside of density because they highly prioritize the opportunity to produce, or to consume what is produced. Sparsely populated places are so because there isn't a lot of production going on there. The few who choose to live in these places do so because they have priorities other than being productive.

In other words, is density a cause or an effect of productivity? How can a study fully control for this?

Yeah, we definitely want to pay $2,000 for studio apartment like in NYC. That is the ticket to the middle class.

Right- but the point of the article is that the reason why apartments in NYC are so expensive is because there have been severe restrictions on new construction in New York and many other older cities which limits housing supply.

A greater supply of housing in NYC would mean even higher density. Higher density is associated with higher housing prices, not lower housing prices. So there must be other factors that cause housing in dense cities to be so expensive, not just restrictions on supply.

What you said makes no sense. How does higher density by itself raise housing prices? Housing prices go up because demand exceeds supply. Dense places are expensive because of high demand relative to supply - your logic is running backwards.

How does higher density by itself raise housing prices?

It raises the cost of supply. Land prices are higher, because there are more people competing for each square foot. Construction costs are higher, because it costs more to build vertically than horizontally.

Construction costs don't determine prices, supply and demand do! If you don't believe me, just do a little googling on Miami condo construction profits.

Construction costs, at best, have a temporary price floor effect on new construction. Temporary is the key word.

The cost of producing a good or service always puts a floor under the price of that good since if that cost is not met the good will not be supplied by the market at all (except temporarily until it is understood by the producer that money will be lost not made in production).

But the housing crisis shows that people quite readily sell for below cost. This is because prices are determined by supply and demand, not cost. Costs will affect the quantity of new construction supplied at a given price, but it doesn't determine the price.

Costs will affect the quantity of new construction supplied at a given price, but it doesn’t determine the price.

Cost of supply sets a floor on the price. If builders can't sell for at least the cost of supply they'll cut the supply until they can. The point is that price comparisons alone don't tell you anything about demand. Just because high-density housing costs more than low-density housing doesn't mean there is more demand for high-density housing or that it is undersupplied.

How exactly do builders reduce the supply of housing that is already built? They can' least rationally.

Are you seriously saying that new construction wouldn't happen in Manhattan, where prices are around $500-600/sf and construction costs are $300-400/sf? Even with the typical land/improvement ratio of ~20%, that is still a shitload of profit to be made.

Sure, whatever you say. Nobody wants to live in New York anymore. Its too crowded. Those high prices don't mean anything.

How exactly do builders reduce the supply of housing that is already built?

They reduce future supply until the price rises to at least the cost of supply. We're talking about the long-term prices here, the equilibrium price, not instantaneous prices. You're claiming that the higher price of high-density housing compared to low-density housing means that there's a shortage of high-density housing, that it's undersupplied. That's just wrong.

Excuse me, but I never said that. Are you on crack? High density housing is undersupplied because prices far exceed costs and new construction is stifled through regulation.

Excuse me, but I never said that.

You wrote: "Construction costs don’t determine prices, supply and demand do! ... Construction costs, at best, have a temporary price floor effect on new construction." That claim is false. The equilibrium price is always at or above the cost of supply. The cost of supplying housing rises with density; therefore, housing prices rise with density.

High density housing is undersupplied because prices far exceed costs and new construction is stifled through regulation.

Unless you believe that the housing market should be completely unregulated, this claim is also false. There are all sorts of regulations that "stifle" housing construction, from health and safety regulations to laws that set aside land for parks and other public spaces. Do you oppose all of them?

Thanks for letting me know that I don't need to respond to you anymore.

This is absurd. Housing density does not cause high housing prices. If you built a city for a million inhabitants and nobody wanted to live there (Hello China), housing would be pretty damn cheap.

It is demand that causes the high housing prices, and the only way to remedy those high housing prices is to respond with greater supply.

Housing density does not cause high housing prices

See my reply to CBBB above.

Major, the differences you note are completely trivial compared to the cost imposed by restrictions on land use. Land prices are higher, but only because the buildings you build on the land are more valuable because they have more units. That does not limit housing supply at all. Constructions costs may be somewhat higher, but at the same time the cost of land per unit goes down.

If land-use restrictions were eased, housing prices would go down, not up.

Major, the differences you note are completely trivial compared to the cost imposed by restrictions on land use.

How do you know? According to Ed Glaeser, the typical construction cost of adding a floor to a New York hi-rise is about $400 per square foot. The typical contruction cost of suburban housing is about $80 per square foot. That looks like more than a "trivial" difference to me. Ditto for land costs. Manhattan has 20 times as many people competing for each square foot of land as a typical suburb. 20 times as much demand per unit of supply means a much higher unit price.

Major - you are not very bright

But the infrastructure in NYC is already incredibly expensive to maintain (let alone construct) because of the existing density. Even if you removed all barriers to new construction and increased density in NYC, it would still be a massively expensive place to live.

And thoroughly unpleasant, if you have any or all of your five senses working properly.

No one would force you to live there--- millions of others would be lining up

I heard there's this new thing called the internet. It's meant to allow people to do business with each other more easily without actually having to live near them. Personally I'm skeptical, but that's what people are saying.

Well a little over 10 years since the tech bubble the productivity advantages of the big cities doesn't seem to have been affected so I guess it hasn't changed things all that much at all.

The problem is the Internet doesn't replace connections and relationships developed over time. A few people manage to achieve the "live in North Dakota and work on Wall Street" lifestyle, but they're extremely senior people who typically spent long careers building up the sort of relationships that allow a "remote career" to work. And they spend lots of time on planes going to the city to maintain their relationships.

For younger people, an "internet career" simply won't get off the ground, unless they're very lucky or are insanely gifted in hustling, which few people are - certainly not enough are to affect migration and settlement patterns.

This is why Silicon Valley itself still exists; the "critical mass" situation discussed in the article about Vietnamese chefs applies in spades to "startup specialist" employees. Having spent my career in startups, I'd never move to a town where there was only one startup, however wonderful the setting or cheap the cost of living. (Very few startups can support Internet remote-working, at least until the company is at least somewhat well-established.) And since most startups fail, you want to be where there are lots of other companies starting up so you can find your next gig without having to expensively relocate.

"A few people manage to achieve the “live in North Dakota and work on Wall Street” lifestyle, but they’re extremely senior people who typically spent long careers building up the sort of relationships that allow a “remote career” to work"

But the tone of the article is this Dick Florida-style pitch for high population density, as if that's the sine qua non. What the smart set in high-density areas are actually doing is hustling to make enough money so they can move to a lower population-dense area. Then they can lecture the rest of us on how badly we need more immigration.

But Silicon Valley is not very dense. Around 6,000 people per square mile. New York City has about 27,000 people per square mile (in Manhattan, 70,000). You don't need New York-like density to achieve this startup-synergy effect you describe.

Correct. Average density doesn't matter at all. It is merely the density of the relevant crowd that matters.

In the case of Silicon Valley, the extreme density inherent in packing tens of thousands of smart-as-hell students into a few classrooms at Stanford University was what mattered.

If average density doesn't matter at all, there's no advantage to building at the average density of New York City rather than the average density of, say, Houston. I don't think your claim about classroom density makes any sense, but we could provide dense classrooms anywhere.

There would be no advantage to building at any density if it were done in a top-down dictatorial fashion, because top-down dictators don't determine the interactions that people make.

But while tech entrepreneurs tend to find each other on the campuses of Stanford, Carnegie Mellon, and MIT, finance entrepreneurs tend to find each other in south Manhattan, marketing and advertising entrepreneurs tend to find each other in Midtown Manhattan, film entrepreneurs tend to find each other in Hollywood, energy entrepreneurs tend to find each other in downtown Houston, fashion entrepreneurs tend to find each other in Beverley Hills or Fifth avenue, etc.

Ryan Avent (and Matt Yglesias and Edward Glaeser for that matter) aren't going around telling people that everybody needs to be uprooted from their suburban shitholes and transplanted into Manhattan. There is no benefit in that. What they are saying is that we need to lower the barriers to entry for people that want to be a part of these scenes, and want to interact on a daily basis with the best-of-the-best and the up-and-coming in their industries. If Brooklyn could hold an extra hundred thousand musicians and artists, we would be much better off.

What they are saying is that we need to lower the barriers to entry for people that want to be a part of these scenes

They're saying that people should ease the regulations that limit the supply of housing in their communities in order to increase population and density. But they're ignoring the costs of increasing population and density -- more noise, more crowding, more congestion, more pollution, less privacy, and in general a loss of quality of life. These are external costs. Every additional person increases them for everyone else. Since there is no mechanism for limiting density by internalizing these costs, we limit it by regulation.

I don't agree that density "lowers the quality of life," as a previous commenter suggested--quality is in the eye of the beholder. Living in a densely-populated city gives me better quality of life--more and better cultural offerings and restaurants, less time spent driving and more spent walking or cycling (reducing the need to spend a lot of time at the gym), and so on. It's also an oversimplification to state that there is "more pollution" in cities--per capita, urban living is much "greener."

I like urban living far more than the suburburban environment in which I grew up (and in which my family still live, so I still get to experience it frequently). Urban life isn't for everyone, but neither is suburban life.

I don’t agree that density “lowers the quality of life,”

Up to a point, I don't either. But beyond a certain density, I think quality of life deteriorates from excessive noise, crowding, congestion, etc. And I think most people feel the same way. Different people have different thresholds, but I doubt there are many people -- even in a really dense city like New York -- who think density should be limited by market forces alone.

It’s also an oversimplification to state that there is “more pollution” in cities–per capita, urban living is much “greener.”

Pollution is measured as the concentration of a contaminant in the environment. In general, the more pollution-emitters -- people, cars, buses, trucks, stores, restaurants, housing units, construction equipment, etc. -- per unit area of land, the higher the level of pollution in that environment.

No, the internet hasn't replaced real life yet.

On the other hand, don't ignore the fact that it has lessened the need for physical proximity in many real ways. There are hundreds of tasks for which companies used to need local businesses or employees, but can now be done anywhere. Supplies used to be ordered from the local office supply store. Tech support used to be handled by employees in your building. Many products could only be effectively sold in stores close to where people live.

It is sort of surprising that the article didn't mention this. It also quite conspicuously used something that is immune to these changes as its primary example.

Yeah man...last I heard, Sergey Brin and Larry Page met on 4chan!

I always found it ironic that Sillicon Valley venture capitalists sold businesses to wall street on that promise, but required almost their investment firms to be within an hours drive of Sillicon Valley.

I'm skeptical, too. Manufacturing, mining and construction all require that you concentrate people in one location. In the financial industry, few companies want the risk of having computers with sensitive client information infected with viruses or keylogger programs so they still insist everyone come to the office and work from there. People who work in sales still know there is no substitute for a face-to-face meeting with a prospect.

And in the industries where it is possible to have people work remotely or from home, those industries seem just as interested in outsourcing all that work to India and China. On the other hand, look at Silicon Valley: why are so many software companies concentrated in one particular, very expensive part of California? Why not just shift the entire software industry to Plano, TX if location no longer matters?

The issue is density, not location. The point about Silicon Valley is that it's very innovative and productive, but not very dense. Your other examples don't support the idea that urban density is very important any more, either. The financial industry has become massively decentralized and globalized. Manufacturing is a declining share of our economy, and is increasingly automated. You might need your sales staff to be physically present in a downtown sales office for face-to-face meetings with prospects, but that doesn't mean you need anyone else there. And I'm not sure what construction or mining have to do with urban density.

I am a contract engineer. I travel all over the country from job to job. When I am at a job I spend all day on a computer. I keep trying to talk potential employers into this whole internet thing you talk about. I have even offered pay reductions. No dice.

The ticket to prosperity is Vietnamese restaurants selling delicious, piping-hot ethnic food! Like in the reporter's own hip, densely populated neighborhood!

Vietnamese food is extremely high profit margin. It could survive anywhere that people aren't particularly fearful of trying something different. A bowl of Pho, a plate of 5 Spice Chicken or Bahn Mi would be acceptable anywhere.

If anything, the high costs of renting and regulation in big cities would make it harder, not easier, to thrive with a Vietnamese restaurant.

I've eaten good VN food in both big cities and small towns. They've got VN food in the suburbs of Birmingham, AL now!

Pho is like the soup grandma made.

During the Industrial Revolution, for instance, millions of workers flooded into fast-growing cities. This produced slums, but it also allowed poor workers to take advantage of opportunities in new industries, a process that helped create the middle class.

Right, it was the Industrial Revolution. Large factories engaged in large scale industrial production required large masses of labor to concentrate. How is this related to contemporary America? Are there huge "media factories" requiring large masses of labor to fill all those "cool" "media jobs" in America's cities or something?

Preventing this density, however, denies workers access to the best opportunities, constraining the mechanism that helps support a strong middle class.

Is it constraining the mechanism that helps support a strong middle class, or a strong landlord and rentier class?

How is "middle class" being defined?

I thought "the middle class" was just a euphemism for stable nuclear families capable of reproducing themselves from generation to generation.

I guess it now can refer to anyone who lives in a big city that works a job that other people consider to be one of "the best opportunities" even though he or she might be in debt and have negative net worth their entire life and never be able to afford a piece of land and children, as long as he or she wears the latest fashions and has the latest gadgets.

Resentment much?

"The American economy’s famous upward mobility": it may be famous but is it exceptional?

I think it's wrong to equate raw density with "hell being other people" as the article does. (Although obviously the general thrust of the piece is pro-density.) There's more to the annoyance of "other people" than just quantity. Cities are anonymous. There are so many people that it's intractable for people to expect to know their neighbors so they just pass by. There are more people per square mile, but they end up just being background noise. As such, a person in a large city can separate themselves from the hell of other people far more effectively than someone in a small town. Cities are actually quite a lot like the Internet in the sense that there are so many people that you can pick and choose which ones you want to interact with.

So in that sense, the Internet is complementary to urban density rather than encouraging people to move out to the middle of Montana.

"Cities are actually quite a lot like the Internet..... So in that sense, the Internet is complementary to urban density rather than encouraging people to move out to the middle of Montana."

You are correct until the last sentence. I would have concluded the internet makes for a good replacement, rather than a complement.

Your internet connection in the middle of Montana is crap. My internet connection in the largest city in Montana is good, but we don't have fios or 4G. Large cities have creative people who want to use the latest gadgets for productive and creative reasons. They don't work in Montana, leaving Montana out of touch with those services. This country is dividing more and more.

Since the wealthy and their sycophants refuse to recognize that the subsidy of wealth inherent in civilization's refusal to tax wealth (ie: charge a use fee for property rights that would not exist in the absence of government) is resulting in (largely childless) engineers being thrown out on the street during middle age in favor of H-1b coolies who work at lower wages to further enhance the wealth of the wealthy, and sent to compete with illegal immigrants for minimum wage jobs (or beg from hostile civil servant ethnic minorities for welfare "entitlements" to which they aren't "entitled" since they are mostly white male heterosexuals) to further enhance the wealth of the wealthy, it's pretty obvious that the cities are doomed for at least two reasons:

1) Personal bioengineering labs.

2) Thermobaric IEDs (turns the world's supply of gasoline into weapons of mass destruction).

Not a lot of engineers are being thrown out on the street (relatively speaking)., Joblessness is heavly concentrated in the low-skilled and low-educated population. You can argue that eventually these people will begin to riot, and maybe you're right, but if the past (as in several millenia) is any guide their rioting will mostly be destructive of their own neighborhoods.

There are only so many people in the US with expertise to operate a personal bioengineering lab, and they have left a paper trail. Buying personal bioengineering equipment leaves a further paper trail. Won't take the FBI long to narrow the list of possible suspects to a small enough number that each of these suspects can be searched for evidence. Now if you are talking suicide bombers, that's a different story. Yes, someone unconcerned about being punished can do some damage. Society will respond with all sorts of enhanced surveillance mechanisms: anyone with bioengineering expertise or access to bioengineering equipment must install surveillance cameras in his/her workshop, and be subject to search by the FBI at any time, and perhaps must carry an GPS locator device on his/her person at all times, etc. Society will do what it takes to ensure society's self-perservation. Mark my words.

Thermobaric IEDs are bulky and no real improvement over sniping with a hunting rifle, which people have been able to do for a century now. Hasn't been a problem in the past, won't be a problem in the future. Lots of people talk big talk about going on a shooting spree, but when it comes time to throwing away their life away, apathy tends to set in.

Society will do what it takes to ensure society’s self-perservation. Mark my words.

This is part of my point. There is an evolutionary arms race and there will be many casualties, not just in terms of human life but in terms of quality and type of life, liberty, etc. "Society's self-preservation" won't necessarily be pretty and look like some utopia.

I think we ought to build a ton of high-rise apartments in Fairfax VA and hand out plane tickets to everyone in Haiti and Guatemala. Then sit back and watch the smartness blossom.

Sounds funny, but many Haitians and Guatemalans would probably jump at the chance. I think they would be better off to. Better government and clean water at the very least but I expect financially to. (It doesn't take much to improve your life when you start from almost nothing.)

I am not so sure Americans would be better off. I don't know what employers would do with so many people in one place whose education suited them for little more then manual labor. As another poster said the days of the industrial revolution are over.

This misses the point. High-density areas are only smarter at equilibrium. I would generally be more interested in the political opinion of an upstate New Yorker in 1850 rather than that of a NYC factory worker. Now, not so much.

Is the claimed reported increased productivity of denser cities a treatment or selection effect? Specifically do all of the barriers to new construction in higher density cities limit the supply of housing to price out the least productive people in the city? Is the purported productivity of San Francisco and New York really because people are living so close to enable knowledge spillovers or is it that the low productivity /poor people have been priced out the communities and as the less productive people in these communities flee, the average productivity mathematically grows?

The reason I suspect that its a selection and not a treatment process is that in the bay area, the county with the highest average income (Marin County) is the bay area county with the most restrictive growth policies and yet the one with the lowest population density. Most of Marin County is either owned by the government or the Marin Agricultural Land Trust. In the bay area,restrictions on new housing seem to be a better explanatory predictor of income than density. Whereas the county with the highest density (the County of San Francisco) is only ranked 9th in the state for income.

I agree that there probably are knowledge spillovers when people work in close proximity to each other. I think this is why many people still work with fellow employees in large office buildings instead of working from home. But I am confident that these spillovers occur whether one is working in an suburban office park or a downtown high rise. In the bay area, the San Francisco Chronicle publishes a list of the 200 largest companies in the bay area. For the past 40 years the trend has been that the number of those companies listed in urban San Francisco County keeps falling and the number of those companies listed in the very suburban Santa Clara County keeps going up.

Since the 1950's in this country the cities that had the highest density housing infrastructure are also the cities that have been shrinking the most and still have some of the least productive populations. Look at St Louis, Detroit, Cleveland, Buffalo, Milwaukee etc.

If the benefits of density were as high as Ryan Avert purports, I suspect that these cities large legacy of high density infrastructure should have made them much more competitive and resilent the the historical record has demonstrated.

Watch Indianapolis, a city with no natural boundaries, built when gas was cheap and white flight was preferred. It is hollowing out by spreading out. the only vibrant parts of it are the relatively densely populated Hispanic areas in the old, old parts of the city along the older National Road.

"the only vibrant parts of it are the relatively densely populated Hispanic areas in the old, old parts of the city along the older National Road."

I look forward to seeing what Next Great Thing comes from all this concentrated vibrancy. Guatemala and Mexico have been rocking along with this vibrancy for centuries. But the vibrancy seems to be voting with its feet to get away from all the other vibrancy. It's a mystery, a paradox, a conundrum ... I guess it will just remain an eternal, unanswerable question.

Do people on this blog really believe humans are fungible? That, like money, if you just scrape enough of them together you have numerous, wealth-creating options?

I would argue density is part of the problem. Many of the major innovations of the past century took place in what were at the time low-density (the Great Lakes region, northern California). Density seems to make areas somewhat sclerotic. You obviously need a density higher than a farm town, of course, in order for an exchange of ideas.

I'm very surprised nobody has pointed it out, but the biggest benefit of having very dense cities are extremely efficient labor markets: employers are sure to find the right workers, and workers are sure to find the right job that matches their skills. But you don't need everyone to pack into dense inner cities to accomplish this -- you just need to be in the same metro area.

If anything, this calls for the continued development of megaregions more than anything else. For example, integrating Seattle and Portland (and maybe Vancouver) into one giant metro area would yield enormous efficiencies. There is lower-hanging fruit that nobody is even talking about -- The Omaha/Lincoln-KC-Des Moines Triangle formed by I-80, I-29, and I-35 contains over 4 million people and none of them are more than 200 miles from one another...yet they're not that economically integrated. You could add interstate highway lanes in rural areas for almost nothing, up the speed limits, and maybe add a little light rail and you're within a 90 minute to 2-hr commute from everywhere. With autocars, this is great.

Just because the internet hasn't revolutionized work doesn't mean it won't. It took decades for engineers to fully transition from steam power to electric power and that is a small task compared to reorganizing the entire basis of human relations, as is required for the telecommuting transformation.

Imagine a flu pandemic that wipes out 10% of the world's population and creates panic about being around crowds of other people. Won't take long for employers to accept dispersal of the workforce. Even for things like manufacturing. Workers in one part of the world can prepare plans for a factory in a site they have never visited, send those plans to another group of workers to build the factory, have another group of workers operate the factory with supervision by top management performed using videos, such that top management may never once visit the factory in question in person. What is needed is a powerful incentive to make this happen. The possibly of top management being infected by some deady flu strain is such an incentive.

All white collar jobs are candidates for video-conferencing. No reason for personally appearing at the bank to get a loan, as opposed to using video-conferencing and transmission of relevant information by email. No reason for bringing a bunch of people into a room to conduct a law trail, even for criminal cases. Just use video-conferencing. Right now, video-conferencing is seen as a threat, because it would eventually lead to elimination of jobs. But faced with a flu pandemic, judges and the rest of the legal community would quickly get their priorities straight.

For all those saying that high-density living is disagreeable, look at this list:

I haven't been to Mumbai, though people say it's an exciting place to live. But a lot of the cities on the list are places I'd be thrilled to live in. Paris, Barcelona, New York. Montreal is on the list and I live in one of the densest (and very nice - I live on a tree-lined street) neighborhoods. I keep thinking that making it denser would make it even nicer and keep musing about how it could be done.

I dunno about crime. Where I live, crime and gangs are associated with suburbs and neighborhoods with inadequate public transportation. Kids are bored and get into trouble. The rate of robberies is higher in the denser areas because they are crimes of opportunity. I've been robbed. I got over it. It's upsetting, but ultimately it's just stuff. Fathers killing their families, or isolated kids stewing and collecting guns and waging war on a school - that kind of thing happens in suburbs where people have more privacy/ are more isolated.

"I keep thinking that making it denser would make it even nicer and keep musing about how it could be done."

Offer to take in a couple of families in your apartment, and have everybody else do the same thing. You could invite families from inner-city Detroit, the Bronx, New Orleans, Atlanta et al. because they're used to high-density living (hence law-abiding, unlike dangerous suburbanites).

Also, uproot those space-wasting sidewalks and trees and expand the buildings out to fit more people.

But a lot of the cities on the list are places I’d be thrilled to live in.

Whatever you personally may want, the trend in the developed world favors less dense cities. The fastest-growing cities tend to be lower-density ones. And large cities in general are becoming less dense. Between 1950 and 2000, the average density of central cities in the United States (a central city is the largest city in a metropolitan area) declined from about 7,500 people per square mile to about 2,700 people per square mile.

Paris has a density of 54,300 per square mile, according to Wikipedia. It's extremely liveable.

Comments for this post are closed