Where are the world's largest cities?

An astronaut's eye view of urbanisation. Image: Getty.

Every year or so, Demographia, the St. Louis-based consultancy, publishes the World Urban Areas Report – a sort of bumper book of city population stats.

The report is a veritable treasure trove of demographic stats for city nerds. If you want to know whether, say, Edinburgh is one of the 1,000 biggest cities in the world, then Demographia will tell you. (It's not, it's just outside at joint 1017th). And if you want to know whether Paris covers a bigger land area than London, it'll tell you that, too. (It does: 2,845km2, as opposed to 1,738 km2.)

It even includes estimates for the size and density of hundreds of towns so tiny that you're slightly surprised to find that anyone on the other side of the Atlantic has even noticed they exist. (Favourite example: Kidderminster*.)

But the most fun is to be found in the first half of the report. Table 1 includes the 1,022 built-up areas in the world that house 500,000 people or more, ranked in order of population; Table 2 ranks the same cities in order of land area.

I don’t know about you guys, but we are stoked.

via GIPHY

(*55,000 people in 16 km2, giving a population density of 3,400 per km2, since you ask.)

Understanding the numbers

Explanations first. The report defines urban areas as

...a continuously built up land mass of urban development that is within a labour market (metropolitan area or metropolitan region)... [and] contains no rural land.

We should probably also note the reports caveat, about the limits of its own methodology:

Revisions are made as more accurate satellite photographs and population estimates become available. As a result, Demographia World Urban Areas is not intended for trend analysis.

Year-to-year changes indicated in population and land area may merely reflect better data that was not used before and may not, therefore indicate a trend.

In other words, if this year's report has a larger estimate for a city's population than last year's report, that might mean the city has grown. Then again, it might just mean that new, more accurate data has appeared.

Moreover, nearly all of the data is estimated. Appropriate caution is therefore advised.

Spoilsports.

The biggest cities, by counting people

Anyway, let’s get to the fun bit. Here are the 10 largest cities in the world by population:

It’s pretty familiar stuff. The top 10 is dominated by Asian cities, with a strong showing from New York City. Tokyo is far in the lead, with Jakarta and Delhi coming up behind. For anyone who pays any attention to this stuff, there aren't at first glance any massive surprises.

We're not supposed to compare with last year's figures for all the reasons laid out above, but what the hell, here's the 2015 top 10:

There are two big differences worth noting. One is that only one Chinese city is now in the top 10, down from three last year. (Beijing and Guangzhou are now ranked 11th and 13th.) All three have seen their populations revised downwards: this seems to the result of previous over-estimates, rather than mass evacuations.

The other noteworthy trend is the sudden appearance in the top 10 of Mumbai, with nearly 23m people. Last year, it ranked 13th with 17.7m. In a section headed "Revised data: highlights", the report notes:

The Mumbai built-up urban area has been expanded to incorporate the Bhiwandi, Kalyan and Vasai-Virar urban areas.

In other words, it isn't that a population the size of Madrid has moved to Mumbai over the last year. Rather, it's become clearer that the megacity has expanded to swallow surrounding areas.

You can see why ranking cities is a complicated business.

The biggest cities, by measuring land

As noted, most of the biggest cities in the world by population are in Asia. Most of the biggest cities in the world by physical size are, well, somewhere else.

Suddenly, with the single exception of Tokyo, Asia doesn’t even feature. Eight out of 10 are in the US. Atlanta has an estimated 5.1m residents, so measured by population it’s the 79th largest city in the world. Measured by land area, though, it ranks fourth.


I guess this is what happens when you build your cities around the car.

Incidentally, the report estimates Atlanta's population density to be 700 people per km2. The comparable figure for Dhaka in Bangladesh – 16.2m people, the 16th largest city in the world - is 44,100.

Dhaka is 63 times more crowded than Atlanta.

All the megacities

One more chart: this is the top 36 by population.

The reason we've stopped at that arbitrary point is not to get London into the rankings (well, not only for that reason). It's because the common definition of megacity is that with a population over 10m.

This is all of them. We've colour coded them by continent:

That's three each in North America, South America, Europe and Africa.

But the remaining 24 – two thirds of all the world's megacities – are in Asia. Five of them are in China alone.

Here, because we love you, is the data used in that graph. Enjoy.



Rank City Population
1 Tokyo-Yokohama 37,750,000
2 Jakarta 31,320,000
3 Delhi 25,735,000
4 Seoul-Incheon 23,575,000
5 Manila 22,930,000
6 Mumbai 22,885,000
7 Karachi 22,825,000
8 Shanghai 22,685,000
9 New York City 20,685,000
10 Sao Paulo 20,605,000
11 Beijing 20,390,000
12 MexicoCity 20,230,000
13 Guangzhou-Foshan 18,760,000
14 Osaka-Kobe-Kyoto 16,985,000
15 Moscow 16,570,000
16 Dhaka 16,235,000
17 Cairo 15,910,000
18 Bangkok 15,315,000
19 Los Angeles 15,135,000
20 Kolkata 14,810,000
21 Buenos Aires 14,280,000
22 Tehran 13,670,000
23 Istanbul 13,520,000
24 Lagos 12,830,000
25 Shenzhen 12,240,000
26 Rio de Janeiro 11,815,000
27 Kinshasa 11,380,000
28 Tianjin 11,260,000
29 Lima 10,950,000
30 Paris 10,870,000
31 Chengdu 10,680,000
32 Lahore 10,355,000
33 London 10,350,000
34 Bangalore 10,165,000
35 Ho Chi Minh City 10,075,000
36 Nagoya 10,035,000

 

And if you want more of this stuff, here’s our report on the 2015 edition of the Demographia World Urban Areas Atlas.

Jonn Elledge is the editor of CityMetric. He is on Twitter, far too much, as @jonnelledge.

Want more of this stuff? Follow CityMetric on Twitter or Facebook.

 
 
 
 

What is to be done? Some modest suggestions on solving the NIMBY problem

Lovely, lovely houses. Image: Getty.

The thing about NIMBYism, right, is that there’s no downside to it. If you already own a decent size house, then the fact a city isn’t building enough homes to go round is probably no skin off your nose. Quite the opposite, in fact: you’ll actively benefit from higher house prices.

So it’s little wonder that campaigning against property development is a popular leisure activity among those looking forward to a long retirement (don’t Google it, it’ll only depress you). It’s sociable, it’s profitable, it only takes a few hours a week, and, best of all, it makes you feel righteous, like you’re doing something good. In those circumstances, who wouldn’t be a NIMBY?

To fight the scourge of NIMBYism, then, what we need to do is to rebalance the risks and rewards that its participants face. By increasing the costs of opposing new housebuilding, we can make sure that people only do it when said development is genuinely a horror worth fighting – rather than, say, something less than perfect that pops up a Tuesday afternoon when they don’t have much else on.

Here are some reasonable and sensible ideas for policies to make that happen.

A NIMBY licence, priced at, say, £150 a month. Anyone found practicing NIMBYism without a licence faces a fine of £5,000. Excellent revenue raiser for the Treasury.

Prison sentences for NIMBYs. Not all of them, obviously – we’re not barbarians – but if the planning process concludes that a development will be good for the community, then those who tried to prevent it should be seen as anti-social elements and treated accordingly.

A NIMBY lottery. All homeowners wishing to oppose a new development must enter their details into an official government lottery scheme. If their number comes up, then their house gets CPOed and redeveloped as flats. Turns NIMBYism into a form of Russian roulette, but with compulsory purchase orders instead of bullets.

This one is actually a huge range of different policies depending on what you make the odds. At one end of the scale, losing your house is pretty unlikely: you’d think twice, but you’re probably fine. At the other, basically everyone who opposes a scheme will lose their entire worldly wealth the moment it gets planning approval, so you’d have to be very, very sure it was bad before you even thought about sticking your head above the parapet. So the question is: do you feel lucky?


NIMBY shaming. There are tribal cultures where, when a member does something terrible, they never see them again. Never talk to them, never look at them, never acknowledge them in any way. To the tribe, this person is dead.

I’m just saying, it’s an option.

A NIMBY-specific bedroom tax. Oppose new housing development to your heart’s content, but be prepared to pay for any space you don’t need. I can’t think of any jokes here, now I’ve written it down I think this one’s genuinely quite sensible.

Capital punishment for NIMBYs. This one’s a bit on the extreme side, so to keep things reasonable it would only apply to those NIMBYs who believe in capital punishment for other sorts of crime. Fair’s far.

Pushing snails through their letter boxes. This probably won’t stop them, but it’d make me feel better. The snails, not so much.

Reformed property taxes, which tax increases in house prices, so discourage homeowners from treating them as effectively free money.

Sorry, I’m just being silly now, aren’t I?

Jonn Elledge is the editor of CityMetric. He is on Twitter as @jonnelledge and also has a Facebook page now for some reason.

Want more of this stuff? Follow CityMetric on Twitter or Facebook.