[bzbarsky]

Comparing the "urban area" numbers is a bit misleading because the structure of cities is so different in different places (even in the US, where for example Chicago officially annexed various suburbs over time while Boston did not).

If you look at the metro area numbers, Shijiazhuang is 26th in China according to your first link, with about 4M in its metro area (assuming "Built-Up area" is the metro area, of course; it's hard to tell from this Wikipedia page). Tehre are 12 more cities with 3M+ in the built-up area.

For the US there seem to be 10 urban areas that are at 4M or more (see <https://en.wikipedia.org/wiki/List_of_United_States_urban_ar...). Obviously fewer than 26, but not an order of magnitude less. There are 14 urban areas with 3M or more people.

More interesting would be to just flat-out compare list positions. #26 on the list of US cities by actual city population is Baltimore. On the urban area list it's San Antonio. Neither one is what one would consider small (nor large, of course).

[yourapostasy]

Good points, thanks for the feedback. Apart from raw population number comparisons here, I personally find the "urban" metric more interesting because that represents the population groupings already in densely-configured spaces. With density comes greater efficiencies across a broad spectrum of measures (energy, face-to-face interactions, waste management, etc.) and those efficiencies, prudently managed, hold out the future promise of unlocking compounding economic advantages. There are of course downsides to dense populations (pollution management, real estate speculation, etc.), as well.

[bzbarsky]

I assume you're talking about the "urban area" column at https://en.wikipedia.org/wiki/List_of_cities_in_China_by_pop... ? It's not clear to me how that's defined, nor what the US equivalent would be....

As for density of US cities, Cambridge is administratively a suburb of Boston, but has a population density of 6400/km^2. At https://en.wikipedia.org/wiki/List_of_United_States_cities_b... it's the 5th densest city on the list. Boston itself has a population density of 5100/km^2, which is less dense.

OK, so maybe Cambridge is weird. But Somerville is right next to Cambridge and has a density of 7400/km^2. Oh, and for economic interaction purposes, Cambridge and Somerville are a lot closer to downtown Boston than many parts of Boston proper. Similar for Chelsea (6100/km^2). Everett (4700/km^2) and Malden (4500/km^2) are up there too.

Other Boston suburbs are a bit less dense: Brookline is at 3300/km^2, Watertown at 3000/km^2, Arlington at 3200/km^2. The interesting thing is to compare those to places like Chicago (4500/km^2) or Los Angeles (3100/km^2). Those are cities that annexed their suburbs, unlike Boston, so if you just compare the "city" population you're comparing apples to oranges. Comparing the population of the "densely zoned/settled area" would be useful, but no one seems to publish that. I assume it would be possible to get that information from census data, though...

Of course as long as we're talking about cities that annexed various surrounding stuff, Houston is at 1300/km^2 and so is Dallas. ;) Which just goes to show that comparing "city" stuff is hard, because cities are just so different from each other.