New Cities

In 2009, the percentage of the planet’s population living in urban areas crossed the 50% threshold…this year the population of the world’s cities will grow by a further 65 million people, equal in size to the total population of France…

As recently as 1990 the United States had the highest number of one million plus inhabitant urban agglomerations globally with a total of 33….by the year 2020 China will lead the world with 121 followed by India with 58…

Remarkably, in 2009 China generated some 40.9% of GDP from just 16.6% of its population living in the 35 largest cities.

From an interesting Credit Suisse report, Opportunities in an urbanizing world (pdf).

I was surprised at how close the association is between state level GDP and the urbanization rate (Ryan Avent in The Gated City and Matt Yglesias in The Rent is Too Dammed High make similar points.)

Urban dwellers also have much lower levels of carbon dioxide production than rural dwellers [Addendum: this seems to be per unit of GDP]. Moreover, the half of the world’s population that lives in cities occupies only approximately 2.7% of the world’s land area.

Hat tip: Gulzar at Urbanomics.


On CO2, to what extent is this due to agriculture, without which cities could not exist?

Meaning a fair assessment of cities would count the whole supply chain and not just imagine you could turn off industry and still have cities?

Exactly, I'm very skeptical of this widely-cited data until we see a more integrated evaluation of supply chains. Not just agriculture, but how about power production, and other industrial uses? "Urban" is too convenient of a way to select out most heavy industry. Is anyone aware of any attempts to address this question?

Thank goodness that Krugman's New Economic Geography laid the foundation for people to think about cities like this.

It is interesting that they use CO2-per-GDP-per-capita in the second graph. I'm thinking that the normalization for GDP, in general, makes the developing-world appear much worse than it otherwise would.

"Urban dwellers also have much lower levels of carbon dioxide production than rural dwellers."

Although you can't tell this from the graph, I expect the opposite is true. It's probably only true when you look at CO2-per-GDP-per-capita as Rahul notes.

The more I think about it the normalization-per-unit-GDP seems really sneaky. It just detracts from the whole urbanization point.

All this is showing is that the high-GDP nations are more "CO2 efficient" i.e. There's an economy of scale to emissions efficiency.

I tend to agree. It is worth noting and desirable that city dwellers and richer countries produce more gdp per unit of CO2 emissions. But it is also necessary to note that they tend to produce a lot more CO2 in absolute terms. Climate dynamics are directly affected by the later, not the former. The Morton Arboretum outside Chicago build a nice large modern visitor center a few years ago. There is a display on how much more energy efficient the new center is than the old one when measured in energy use per square foot. What's not emphasized is that the new center must be at least three times larger than the old one, so total CO2 emissions are far greater.

The Developed world also produces a lot more CO2 on a pure per-capita basis. I guess the underlying point scratched by the "CO2-per-GDP-per-capita" metric is whether being more productive is an alleviating factor for CO2 emission.

First graph; cities sell services with ambiguous value and content, but which feel vital, preventing commoditization and allowing inflated prices to be charged, thus higher GDP relative to labour.

Second graph; more urbanized populations make relatively more services and less stuff (e.g. food, humans, machines). CO2 emissions are an index of making stuff and transporting stuff.

What is "CO2-per-GDP-per-capita", by the way?

My guess is they mean CO2 / GDP, which is equal to CO2 per capita / GDP per capita .

The graph above this one (in the PDF) is labelled in "7UDQVSRUW&2
HPLVVLRQV ,EVSHUSHRSOH " umm that's odd, anyway "Transport CO2 (lbs per 1000 people)". So I presume that is their unit "CO2", it is per population.

(And per year, I presume too. But you already lost all the points assigned for units.)

When cities are decent places for children and not just 25-year-old males I will listen to people who tell me this trend is a good thing. The paper didn't have much on birth rates, but it's my understanding that cities are still an abattoir, though these days the thing keeping the population down is orthotricyclen not smallpox.

If cities are so bad for your genetic prospects, why do people move there? I suppose many people move their temporarily to find a mate and then move out. Were they ever good for your prospects? Was it a local knowledge issue, as in "I know my farm is 100% doomed, so I'm better off taking the 50% gamble of a plague-infested city?" Was the prospect of higher genetic diversity worth the reduced number of offspring?

When I think of third world rural people, I picture a peasant scratching a living out of an acre or two. You actually can't work more land than that without machinery. Now suppose that you are the third son of one of these peasants. This third son will not inherit any land. Where can he go except to the big city to try to make it? Aren't these the circumstances that drove up the population of London in the eighteenth century, Chicago in the early 20th, and Mexico City, Kinsasha, and Kolkata in our time.
I doubt that anyone is leaving a bucolic paradise to move to Trenchtown, but what are the alternatives?
Once the people are in the big cities, what are the forces that turn the big cities into places of excitement, culture, innovation, learning, industry, and
I am reminded of Samuel Johnson, " Who loves not London, loves not life."

If I understand your comment correctly, your suggestion is "It's a local knowledge issue." Certain people know they are 100% doomed in rural life, so they take the crapshoot of city life as a lesser evil. On average that isn't what rural people should expect, but it only needs to be the expectation for a significant enough number of people to keep cities around and abounding in fodder.

"When I think of third world rural people, I picture a peasant scratching a living out of an acre or two. You actually can’t work more land than that without machinery. "

It doesn't really effect your point, but yes you can work significantly more than an acre or two without machinery. Average farm size in Iowa in 1890 was about 150 acres.

Some of that was fallow, some was pasture and some was wooded, but still acreage under till was almost certainly significantly more than an acre or two, even per capita.

I'm not sure what classifies as a city, I would assume that many of the people who 'live in cities' actually live in the suburbs, many of which are ddecent places for children and not just 25-year-old males'.

ddecent should read 'decent

That's certainly a fair point.

I should also add that I'm not completely crazy and do think cities are fun. It's their poor livability for families that I'm griping about.

The upper middle and upper income types raise their kids just fine in cities. Agreed it's pretty hard at middle income and below.

If by "just fine" you mean they choose to have fewer kids or move when they hit a threshold. My peers generally flee the city when they get the second kid. If they do not have more than one kid, they may stay in the city. Most people would call them upper income.

We're just trading anecdotes but many of my peers have 2+ kids in the city. Including myself. Some do leave.

Also, this isn't particularly new. Folks have been having kids and moving to the burbs ever since they were invented in the 1950s. Before the suburbs came about, you had kids on the farm or in the city. 'Livability' for families is exactly what suburbs were designed for.

Why just males? The countryside just doesn't compare to the city in terms of providing opprtunities for you and your girlfriends to prance around in heels and cocktail dresses on weekend nights to get the validation and attention you crave from a variety of men, then posting photographic documentary on Facebook after. Plus... shopping!

That may be my own bias... For various reasons I tend to assume everyone commenting here is a 25-year-old single male.

I think the women you describe are mate-seeking. They probably won't live in cities long-term.


All young adults are mate-seeking. Even the ones with SOs, whether it be girls trying to upgrade or guys looking to upgrade/diversify.

I don't dispute that - anyway I didn't mean to say anything that could be interpreted otherwise.

I know you're attempting to be facetious, but at least in China, most of the migration to the cities is from rural women who have much better opportunities in the factories of the big cities, not rural men who are more likely to stay in the provinces working the family farm.

That pattern was mirrored in industrial revolution era England and New England.

I don't know what definition you use for "cities" but I can think of many "cities" (?) in the US with a population of around 250k that are excellent school districts.

Also, once you go outside the US many of the biggest cities are quite great for kids.

I'm only really familiar with the US and Europe. Europe is significantly less family friendly than the US, particularly in urban areas. Hence its' low birth-rate. Try buying groceries in Paris, or getting an apartment suitable for two parents and three kids. Try getting the mom back in the work force after the kids are a little older.

Regarding the definition of city, the comments show it's unclear. I was thinking of dense urban areas. Boston, not Brookline. Manhattan, not Westchester.

White American total fertility rates are no higher than White French total fertility rates (which have increased recently, even excluding the immigrant population) - so I'm not sure labor flexibility and big box retailers are much of influence of the number of children people have.

You're right, I picked about the worst European example I could - White French TFR is only about 0.15 behind White American TFR. Thank you. I don't think that is a strong attack on my point, but I don't claim to fully understand what's going on so I appreciate the input.

I'm pretty sure the labor market flexibility - TFR connection is well studied, but I can't find the paper I normally cite to. Big box retailers seem obviously connected, although causality probably runs both ways...

"I was surprised at how close the association is between state level GDP and the urbanization rate "


What planet are you from?

It surprises you that rising economic activity leads to urbanization??

That's remarkably ignorant of the history of civilizations for someone who preens in being so well read and a bien pensant.

I think perhaps you were faux-surprised simply because you're thinking of this from your preferred opinion/policy advocacy perspective -- that urbanization leads to GDP.

When the causation is actually reverse -- it's GDP that leads to urbanization.

Are there spillovers within urban areas? Sure. Folks like Lazear argue convincingly that those spillovers exist, and that they are positive. Let's also be clear that there are also negative spillovers though.

Isn't it amazing that in 1990 the US had 33 cities with over a million people, and today we have, well, what --9 or 10? That Obama really can destroy America, I say.

I realize you're joking, but for others, in economic geography terms, "city" refers to a wider metropolitan area rather than the legal boundaries of a municipality. America has only nine of the latter with over a million people, but either 44 or 51 of the former with over a million people (depending on whether you go by MSA or CSA).

You wouldn't use CSA's to count urban centers because not every urban center is part of a CSA but every urban center is part of an MSA. Using CSA's you'd miss (according to wiki) urban centers like Miami, Phoenix, San Diego, Tampa, Portland, and San Antonio.

And you see CO2 savings by living in Loundon county, va? Or westchester?

There are so many mixd up datasets here I have to think somebody slept through April fools.

Urban living in density. That means, in the US, you've got maybe 6 big urban areas: New York, Boston, Chicago, Philly, San Fran, LA and maybe DC. 3 or 4 of those are over a million people.

I agree, by the way. I think the definitions are so vague and vary so much between country, you'd have to be very, very careful with any results. I was just pointing out that the term cities in these studies refers to metro areas, more akin to the colloquial use of the word "city" (at least, outside the US), rather than actual municipalities that tend to have pretty arbitrary borders.

The Credit Suisse report seems to use a definition of "city" that consists of "a settlement with over 20,000 people". Its hard to tell because its a long report, packed with filler.

This might be an OK definition for ancient Sumeria, but my mother and my wife would be surprised to find out that they came from cities. They both thought they grew up in small towns (both of which are just over the 20,000 threshold) and moved to cities as young women.

Since the population of the world has increased by three billion people in the last forty years, I would imagine that there are more places with a population of over 20,000 people than ever! I'm not sure how you can distribute seven billion plus people and somehow have most of them live in settlements that don't hit the 20,000 mark.

Other than that, yes I suppose rising GDP and growing city populations are correlated.

Wow VT is down there with WV, MS, SC and KT in GDP/hours worked!

Figure 15 in the report suggests (to me anyway) that the productivity advantage of urbanization are getting *larger* over time. Which is, in its own way, even more interesting that the finding that productivity in more urbanized places is higher than productivity in less urbanized places.

Wow, look how productive DC is! If only every state could follow their economic model of massive rentseeking.


DC: yet another sign that GDP doesn't really measure what we think it does.

There's a lot of tech industry around there too. But yes a great deal of lawyering and lobbying for sure.

All of that is really beside the point. What is the primary industry in DC? Spending other people's money, seized by threat of force.

Don't let Richard Florida see that post.

DC has third highest GDP per hour? Could someone kindly explain to me what it is that's produced in DC?

candy distribution pipelines

Why, rents of course!

For some time I've been suspicious of GDP as we know it as a meaningful measure of economy, and now I know that it's worthless. Progress!

Depositions and Congressional golf outings.

In the second graph each country occurs seven times - once every five years. This skews the conclusion horribly: In many ways, it's coincidence that the last 30 years on this planet have seen massive growth in BOTH urbanization and energy/CO2 efficiency. A far better way would be to examine individual states/districts/prefectures over several different countries. Great discussion going on here, though.

If you're considering carbon emissions, what about carbon sequestration? Agriculture, along with natural forests are the only major carbon sinks we have and they offset a lot of the carbon released in both urban and rural settings. The major carbon sinks are, by nature, rural. I'm not 100% sure this wasn't accounted for, but if it wasn't, the effect on net CO2 emissions/GDP might be interesting. I could also be underestimating the disparity in productivity between the 2 settings.

Comments for this post are closed