Growth of World Cities (1940-2010)

[Yuri]

Back to the topic, this chart by UN:


Estimated urban growth per hour through a combination of natural internal growth and migration in selected world cities. UN World Urbanisation Prospects 2014

As expected, full of obscure and clearly inaccurate data.

[New Brisavoine]

Originally Posted by yuriandrade

Firstly, I do think we would see some colour changes all over the board.

The colors would change only very little, and would quickly stop changing after a few iterations. I know because I actually looked at the figures from the UK 2011 census (you haven't).

Originally Posted by yuriandrade

Also, the blue areas would bring London to 2,000km2 only. They could go much further, up to 5,000km2 for instance, including, therefore, some centres of emplyment. Then, we would see significant changes on the map.

Lol. They could enlarge the territory of the Greater London Authority to cover the entire island of Great Britain, that wouldn't make London a metro area of 60 million people.

What you seem unable to understand is that metro areas are not based on the borders of the central municipality. Metro areas are based on the central urban area. In most cases, the central urban area is the central municipality plus some suburbs beyond, but sometimes when the central municipality has been enlarged so much that it covers more than the central urban area, then the urban area used to calculate metro areas figures (commute figures) is smaller than the central municipality. That's for example the case with Chongqing in China: its urban area and metro area are much smaller than the administrative city of Chongqing. It's not because the Chinese have artificially enlarged the administrative borders of the city that we now consider that Chongqing has 30 million inhabitants.

So the exact location of where they place the administrative border of Greater London is totally irrelevant to this discussion.

[Yuri]

Originally Posted by New Brisavoine

Lol. They could enlarge the territory of the Greater London Authority to cover the entire island of Great Britain, that wouldn't make London a metro area of 60 million people.

What you seem unable to understand is that metro areas are not based on the borders of the central municipality. Metro areas are based on the central urban area. In most cases, the central urban area is the central municipality plus some suburbs beyond, but sometimes when the central municipality has been enlarged so much that it covers more than the central urban area, then the urban area used to calculate metro areas figures (commute figures) is smaller than the central municipality. That's for example the case with Chongqing in China: its urban area and metro area are much smaller than the administrative city of Chongqing. It's not because the Chinese have artificially enlarged the administrative borders of the city that we now consider that Chongqing has 30 million inhabitants.

I had the exact same discussion with Minato as he insisted to use Chongqing as an example. Obviously, that's an extreme case. However, what about Beijing or Shanghai or Tianjin municipalities? Everybody, regardless their municipality size, equals municipality to metro area in those cases. The same would be the case with a 5,000km2 or a 10,000km2 Greater London. No need to use Chongqing's 80,000km2 as an example.

Originally Posted by New Brisavoine

So the exact location of where they place the administrative border of Greater London is totally irrelevant to this discussion.

Not at all. They would take this 5,000km2 or 10,000km2 Greater London as reference and this would automatically increase the commute rates between this new bigger London and the regions outside it.

[Yuri]

Originally Posted by yuriandrade

Back to the topic, this chart by UN:


Estimated urban growth per hour through a combination of natural internal growth and migration in selected world cities. UN World Urbanisation Prospects 2014

As expected, full of obscure and clearly inaccurate data.

Some stuff I posted on SSC about it:

Between 2001-2011, PWV (Pretoria-Witwatersrand-Vaal Triangle) added more people (3 million) than São Paulo official metro area (2 million between 2000-2010). And even excluding Pretoria, taking the Witwatersrand (Johannesburg plus East/West Rand) only, it's probably adding more people than São Paulo metro area.

And then we have, London, which added 1 million people (2001-2011) in the city proper only, while the entire New York CSA added 700,000 people (2000-2010).
__________________

[New Brisavoine]

Originally Posted by yuriandrade

I had the exact same discussion with Minato as he insisted to use Chongqing as an example. Obviously, that's an extreme case. However, what about Beijing or Shanghai or Tianjin municipalities? Everybody, regardless their municipality size, equals municipality to metro area in those cases. The same would be the case with a 5,000km2 or a 10,000km2 Greater London. No need to use Chongqing's 80,000km2 as an example.

Everybody, no. In the case of Beijing, the metro area is definitely smaller than the size of the municipality of Beijing. It's only because you ignore the geography of Beijing that you equate them. For example the city of Pinggu is fully separate from Beijing and not morphologically or functionally linked to it, despite lying inside the administrative territory of the municipality of Beijing.

For Tianjin and Shanghai, some parts of their administrative territories may or may not form part of their metro areas, depending on the definition used for the metro area. One would have to look at figures from the Chinese census. For example Jinshan, which lies inside the municipality of Shanghai, is clearly distinct from Shanghai, and I'm not sure it sends that many commuters to Shanghai. Wikipedia even says that Jinshan people speak a dialect quite distinct from the Shanghai dialect and unintelligible by people from Shanghai.

Originally Posted by yuriandrade

Not at all. They would take this 5,000km2 or 10,000km2 Greater London as reference and this would automatically increase the commute rates between this new bigger London and the regions outside it.

No, "they" wouldn't. Trust me.

[Yuri]

Brisavoine, most of the lists on metro areas describe the three Chinese municipalities as being also their metro areas, and not even Beijing, the bigger one, is sliced. And Shanghai, many definitions actually include Suzhou.

Anyway, let's go back to the thread? Growth of metro areas. Do you have figures, in fixed borders, for cities all over the world?

[New Brisavoine]

Originally Posted by yuriandrade

Do you have figures, in fixed borders, for cities all over the world?

Yes I do, but it's mostly French and continental European cities, and we both know all you care about is London, São Paulo, London, São Paulo, and London, so I won't bother posting my figures here.

[daniel]

I tried to do this with Israel. By far the biggest metro would be Tel Aviv, which isn't particularly large on a global scale, it is bigger than some of the ones listed here already.
notes:
1) I call it the Tel Aviv metro to keep in with the thread norm, but Israelis don't usually think of it in this way, Tel Aviv just means the city itself. The area around it is called the centre or Dan Bloc.
2) The area included is Tel Aviv district and centre district. These isn't the exact current definition of the metro but it's close and wouldn't be a terrible definition for it, it is in keeping with the spirit of the thread, and I wouldn't be able to get historic population numbers for other definitions.
3) Exact numbers for 1950 are impossible, so I went with 1948 (The year Israel was founded) and from 1960 continued every 10 years.

Year	population      change
1948	427,700         -
1960	1,090,300	+54.92%
1970	1,420,100	+30.25%
1980	1,795,100	+26.41%
1990	2,126,500	+18.46%
2000	2,611,500	+22.81%
2010	3,139,900	+20.23%
2015	3,440,300	+9.57%

[muppet]

Just bear in mind guys, the threshold for calculating a CSA in the US relies on much lower levels than the equivalent in UK or Europe. For example the commuting thing relies on as low as 10% for the entire county to be included, farms and all (plus the other 90% of population that don't commute).

The other glaring point in the smallprint: one only needs to commute into the next county/ town along to be considered - not necessarily into the main city. For example less than 5% of those from NYC's exurbs actually reach NYC proper, despite taking up more than half the statistical area. The vast majority of commuters there only commute into the next town.

[iheartthed]

Originally Posted by muppet

Just bear in mind guys, the threshold for calculating a CSA in the US relies on much lower levels than the equivalent in UK or Europe. For example the commuting thing relies on as low as 10% for the entire county to be included, farms and all (plus the other 90% of population that don't commute).

The other glaring point in the smallprint: one only needs to commute into the next county/ town along to be considered - not necessarily into the main city. For example less than 5% of those from NYC's exurbs actually reach NYC proper, despite taking up more than half the statistical area. The vast majority of commuters there only commute into the next town.

I think this would only be an issue if this were measuring land area instead of population. If the methodology causes a bunch of farmland counties to be added then it really wouldn't affect the overall population numbers of large metro areas.

[SHiRO]

Originally Posted by muppet

Just bear in mind guys, the threshold for calculating a CSA in the US relies on much lower levels than the equivalent in UK or Europe. For example the commuting thing relies on as low as 10% for the entire county to be included, farms and all (plus the other 90% of population that don't commute).

The other glaring point in the smallprint: one only needs to commute into the next county/ town along to be considered - not necessarily into the main city. For example less than 5% of those from NYC's exurbs actually reach NYC proper, despite taking up more than half the statistical area. The vast majority of commuters there only commute into the next town.

Muppet! Good to see you back! Please post more

Is the blog in your sig yours?

[muppet]

Yep

[muppet]

Originally Posted by iheartthed

I think this would only be an issue if this were measuring land area instead of population. If the methodology causes a bunch of farmland counties to be added then it really wouldn't affect the overall population numbers of large metro areas.

It does when it inflates a population count of a 'city' by 5 million. So yes, I hear ya, it includes the farmland - but also the towns and villages in the farmland (that's not part of the city, nor even the commuters to the city).

[Docere]

How about Washington DC? It seems to have grown at a similar rate to Toronto since the 1940s.

[iheartthed]

Originally Posted by muppet

It does when it inflates a population count of a 'city' by 5 million. So yes, I hear ya, it includes the farmland - but also the towns and villages in the farmland (that's not part of the city, nor even the commuters to the city).

Maybe I need an example to actually follow, but if there are that many people being picked up as part of a statistical area then they probably are part of the system of the central city. Five million people don't just live in the middle of nowhere. Half of the population in the United States lives in less than 5% of the counties in the country, which are largely clustered around the 30 largest metropolitan areas.

[JManc]

Originally Posted by muppet

Just bear in mind guys, the threshold for calculating a CSA in the US relies on much lower levels than the equivalent in UK or Europe. For example the commuting thing relies on as low as 10% for the entire county to be included, farms and all (plus the other 90% of population that don't commute).

The other glaring point in the smallprint: one only needs to commute into the next county/ town along to be considered - not necessarily into the main city. For example less than 5% of those from NYC's exurbs actually reach NYC proper, despite taking up more than half the statistical area. The vast majority of commuters there only commute into the next town.

They should go by the density of Giants/ Yankees/ Rangers/ Knicks fans (New York) or Chelsea/ Arsenal/ Tottenham/ West Ham fans (London) as a gauge to what areas count as "New York" or "London".

That's just my 2 cents...er...2 pence...

[Shawn]

Originally Posted by JManc

They should go by the density of Giants/ Yankees/ Rangers/ Knicks fans (New York) or Chelsea/ Arsenal/ Tottenham/ West Ham fans (London) as a gauge to what areas count as "New York" or "London".

That's just my 2 cents...er...2 pence...

It's so true: New York ends and Boston begins at the Connecticut River, and not because of the physical barrier . . . this is the dividing line between Yankees and Sox hats, Giants and Pats hats, and Knicks and C's hats. Because of the Whalers history, the Rangers / B's line is a bit murkier.

(No one in CT sullies themselves with Jets gear)

[MolsonExport]

Originally Posted by muppet

Yep

I really enjoyed reading your blog.

[SHiRO]

Originally Posted by muppet

Yep

Nice! Really well written and for a change some accurate information about these subjects on the internet!

[McBane]

Originally Posted by daniel

I tried to do this with Israel. By far the biggest metro would be Tel Aviv, which isn't particularly large on a global scale, it is bigger than some of the ones listed here already.
notes:
1) I call it the Tel Aviv metro to keep in with the thread norm, but Israelis don't usually think of it in this way, Tel Aviv just means the city itself. The area around it is called the centre or Dan Bloc.
2) The area included is Tel Aviv district and centre district. These isn't the exact current definition of the metro but it's close and wouldn't be a terrible definition for it, it is in keeping with the spirit of the thread, and I wouldn't be able to get historic population numbers for other definitions.
3) Exact numbers for 1950 are impossible, so I went with 1948 (The year Israel was founded) and from 1960 continued every 10 years.

Year	population      change
1948	427,700         -
1960	1,090,300	+54.92%
1970	1,420,100	+30.25%
1980	1,795,100	+26.41%
1990	2,126,500	+18.46%
2000	2,611,500	+22.81%
2010	3,139,900	+20.23%
2015	3,440,300	+9.57%

This would be more impactful if it showed populations starting in 1900. Tel Aviv was nothing but sand dunes prior to its founding in 1910. Prior to then, what is now the Tel Aviv metro area was sparsely populated and mostly agricultural except for the ancient port city of Jaffa (which was still only in the 50-60k range) and few small towns here and there. It would be interesting to see the population growth from 1900 on.