China fact of the day

Chinese cities are less densely populated than many people think:

While China is a very populous country, density in Chinese cities is typically much lower than major American metropolitan areas. For example, the first tier cities have population density varying from 1,000 to 2,000 people per square kilometer in 2010, whereas New York City has density of 10,000 people per square kilometer and the top 100 American MSAs all have density above 4,000 people per square kilometer.

That is from Glaeser, Huang, Ma, and Shleifer “A Real Estate Boom With Chinese Characteristics,” an interesting and useful paper.  They argue that current Chinese real estate prices might prove sustainable, but there is a basic dilemma that cannot be avoided:

…that path may create significant social costs. Construction employment would plummet. Millions of Chinese may lose the apparent productivity advantages associated with living in Chinese cities (Chauvin et al. 2016). Local governments would lose the financial autonomy from land sales and taxes that has been their institutional basis. The alternative for the Chinese government is to accommodate high levels of construction and housing supply. As we showed, this will lead to very low or negative expected returns to investment in housing. The welfare of potential new buyers will rise, but current owners will suffer losses.

Recommended reading.


Don't Chinese municipalities tend to across large amounts of surrounding rural areas?

It could be useful to have a common rule of thumb for deciding across countries what is and what isn't in the city part of a massive metropolitan area.

Maybe something to do with what % of workers commute downtown?

Raj Chetty's current work involves divvying up the U.S. into "commuting zones" -- that seems promising.

Yes, political boarders are not useful.

At least in China, outside of the area of main urban development, there are two further levels of government below municipal: town, and then village.

So if you're in Town B in Municipal "jurisdiction" C, you're in a less urban area, and if you're in Village A in Town B in Municipal "jurisdiction" C, you're certainly not in an urban area.

(I think ... but in fact there is much more diversity in jurisdictional and institutional arrangements in China than most know, including most Chinese.)

So if you're map of China includes villages and towns then the statistical corrections might not be that complicated.

When an area of urban development expands into non-urban areas, the former political jurisdictions are rolled into the urban municipal government (I think), so the data should actually be pretty decent without having to resort to any fancy tricks.

Beijing's official land area is about 6,000 square miles which makes it slightly bigger than the state of Connecticut. No way you can compare it to New York City, which is "only" 300 square miles.

Yeah, this statistic seemed fishy to me too. Difficult to ever compare city size by land area. Same in the states.

Exceptions being a handful of megacities and "municipal" "jurisdictions" with little urban development.

Yes. A casual look at the map shows this very easily. The units are not "cities", they're more like French prefectures, but unfortunately there's a history of bad translation adding to the confusion.

But you can't expect modern academics to have this sort of knowledge.

Yeah a lot of Chinese cities include vast areas of surrounding suburbs and even semi-rural areas. Having been to a number, they aren't all that different from anywhere else (depending on what your expectations were I suppose)

Among other things, way more parks, way more space for pedestrians and cyclists.

But the cities really are otherwise less dense, even with all the other stuff mucking up the numbers. For example, apartment buildings are most often 5-10 floors not 20-40 or more, although this is rapidly changing.

In how many American cities is it the norm to have 20-40+ floor apartment buildings? Not very many, I'd wager.

Thanks, the Other Jim. I was wondering why a 46 year old white guy was arrested for those shootings. Now I know. He was a Bacon, Lettuce and 'Marto activist.

White people being arrested due to PC pressure, Obama and media are putting intense pressure on local police to cover up BLM terrorism.

I don't know if China and the US are comparable. population density is registered inhabitants divided by area. Commercial property is full of people but adds zero to density. Same for government buildings, schools, train stations, very crowded places that are not accounted in density because people does not "live" there. If people in China commute more to the city instead of inhabiting it, that may explain the lower density. But...the use or commercial property remains intensive.

"The welfare of potential new buyers will rise, but current owners will suffer losses."

What losses will owners of homes suffer by others gaining homes? It will still provide the same shelter, ie utility?

After all, in a free market, price with converge on cost of construction or manufacturing. At least that's what I learned in the 60s.

It's not a free market.

"[D]ensity in Chinese cities is typically much lower than major American metropolitan areas". Are the authors comparing apples (cities) and oranges (metropolitan areas) or just being sloppy with the terms they use?

Some comparisons I use in class. Chinese cities are far less dense than American cities if comparing legally bounded areas of cities. By law, all Chinese cities must remain about 85% "non-construction" land, meaning farmland or otherwise vacant. A better comparison would be the districts of Chinese cities to American cities, or Chinese cities to American metropolitan areas.

Place Area, km2 Population, x 106 density

Liaoning Province 145,900 42.17
Zhèjiāng Province 101,800 47.2
Illinois 143,959 12.8

Shenyang 12,980 6.7 516/km2
Dàlián 13,237 5.7 430/km2
Hángzhōu 16,847 6.4 380/km2
Shànghăi 7,037 18.7 2650/km2

Cook County 4,235 5.1
City of Chicago 580 3. 5170/km2
Chicago planning area 10,387 8.5 (seven counties)

Thanks. Urbanization is only a recent phenomenon in China, which also could affect density. For example, in 1980 the population of Shenzhen was less than 100,000. Today, the population exceeds 10 million. Do planners promote less density, with such things as the law you mention and better public transit?

Chinese urban planning has been modernist, car- oriented since its rejuvenation after Mao. But is is complicated. Chinese cities sprawl tremendously. There are now lots of newly developed areas in which lack of a car is as much of a detriment to travel as it would be in someplace like Naperville, outside Chicago. This is for many reasons, not the least of which is the manner of control of land by different levels and units of government or state owned businesses. Lots of cities have developed, or are developing, subway systems in the midst of development, which delay need not have happened but the systems are now really necessary to combat the traffic problems. And even more than is the case in the US, Chinese urban planners are often stymied by developers and their political cronies, so what is built can be far different than what was planned. In addition, the coordination between land use planning and transportation planning can be poor. So the planning can now argue for more transit, and better environmental planning, but the realities on the ground are often quite different.

The numbers for China are far off. In a typical town of northern China where resident houses are one story the population can be easily 4000 per suqare kilo meter. Shanghai spans over an area of 7000 SQ KM with actual population more than 30million. You can do the math.

My population numbers are a bit old - dating from about 2007. The point is to compare like quantities. There are no farms in New York or Chicago, but there are farms an hour away, in the suburbs. Every Chinese city of more than a million has substantial farmland. Discussions of urban planning and transportation planning and real estate development need to be cognizant of differences in local definitions as well as law and custom. The high modernism of Chinese planning worked for a while, but now it only isolates entire areas from the rest of the city. Almost back to the village life idea - travel costs make it too difficult to leave, so lives get constructed rather locally again.

You sound like talking about administrative area of cities. It is true most big cities are composed of two parts, suburbon and urban. Farmers live in suburban area with farmlands and people with city HuKou live in urban area. An urban area is similar to New York metropolitan area, crowed with large population, residential and commercial buildings. Recent years, real estate developers and government have built lots of high-rise buildings on the suburban farmland making the cities more crowded.

Interesting. I suppose that having much agricultural activity near every city is good for food security in the case of any widespread war.

The Chinese urban area is still under change. When I was in Shanghai, I was struck by the number of low-rise (~3 story) apartment blocks that look like they were built in the 1960's (but who knows) which were cordoned off with big signs saying something along the lines of "don't go back into condemned building". The old buildings were being replaced with very tall apartment towers.

For example, see:

Chinese cities are by almost any standard considerably more dense than American cities. The population density of the five largest urban areas in China:
Shanghai — 6100 people / km^2
Beijing - 5500
Guangzhou - 6000
Shenzhen - 6900
Tianjin - 5400

And in America...
New York City - 2040
Los Angeles - 2400
Chicago - 1300
Dallas - 1200
San Fran - 2100

That's from here:

The discrepancy is even larger if you look at neighborhoods in the urban core. Manhattan, with a population of about 1.6 million has a density of 29,000 per km^2. Some comparably sized districts in major Chinese cities:

Xicheng (Beijing): population 1.3 million, density of 25,000 per km^2
Huangpu (Shanghai): population 900,000, density of 33,000 per km^2
Hexi (Tianjin): population 900,000, density of 24,000 per km^2
Yuexiu (Guangzhou): population 1.2 million, density of 34,000 per km^2

That's from here:

The urban core of most Chinese cities is built at densities somewhere around the density of Manhattan (often a bit lower). How many American cities can boast that? And then the urban area, including all the suburbs, will generally have an overall density comparable to that of the city of Boston or Chicago. Chinese cities aren't quite as dense as, say, cities in South Asia, but they're clearly built on more of a 2nd world model than a 1st world model.

I don't know if this makes sense, but consider density per population, to compare like and like.

Say, compare a US city of 10 million and a Chinese city of 10 million. With a similar total population, is density higher or lower? This avoids the problem of trying to compare a city of 30 million with a city of 8 million and then trying to draw conclusions where density planning (or absence thereof) is relevant.

See the Max Bloom post above. Chinese cities are composed of districts and counties. Districts are "urban," although they may contain farmland, and counties are "rural," although they may contain substantial cities of their own. Even using my old data from my post above, the City of Hangzhou has an area almost 30 times that of Chicago. Hangzhou is larger in area than all of the Chicago metropolitan area, including anything you would want to define as a suburb of Chicago, using commuting sheds or whatever data you wish. It is larger than what we think of as northeastern Illinois. As of about 2007 or so, the population of Hangzhou was about 6 million; the population of Chicago was about 3 million; the population of northeastern Illinois was about 8 million. The population of the districts of Hangzhou was about 3 million, roughly the size (population) of Chicago. The most dense parts of Chinese districts of major cities are built out at Manhattan densities per unit area. Otherwise, districts - as in Hangzhou, as in Dalian, as in other major cities - can contain miles of suburban looking development, car oriented, that also contain multiple nodes of substantial downtown looking development. For a US comparison, think of Cook County as the districts of what we call "Chicago," and the surrounding counties of "Chicago" including some developed areas of their own, including Naperville, Aurora, Waukegan, Gary, perhaps Milwaukee. That would be a rough idea of what Hangzhou is like.

Is there a metric to imperial conversion area here? Those numbers would sound more plausible if the American ones were expressed in terms of sq miles.

The idea that Chinese cities are less dense than US metros is laughable. As mentioned above, Los Angeles--Long Beach--Anaheim, metro has a density of 2,702 per sq km

Comments for this post are closed