Foreign born populations: how many nationalites are represented?

[Docere]

I hope this doesn't violate the no "city vs. city" policy.

Using 5,000 born in a country as a cutoff - how many countries are represented in the foreign born populations of various cities.

Here's US data (2014 ACS, 5-year estimates)

New York MSA 83 (Europe: 21, Africa: 8, Asia: 25, Oceania: 1, Americas: 28)
Los Angeles MSA 57 (Europe: 11, Africa: 4, Asia: 25, Oceania: 1, Americas: 16)
Washington MSA 48 (Europe: 5, Africa: 8, Asia: 15, Americas: 20)
Chicago MSA 42 (Europe: 15, Africa: 1, Asia: 15, Americas: 11)
Bay Area* 38 (Europe: 8, Africa: 1, Asia: 18, Oceania: 2, Americas: 9)
Miami MSA 37 (Europe: 7, Asia: 6, Americas: 24)
Boston MSA 34 (Europe: 10, Africa: 2, Asia: 10, Americas: 12)

* San Francisco-Oakland and San Jose MSAs.

Canadian cities (2011 NHS):

Toronto CMA 70 (Europe: 21, Africa: 9, Asia: 23, Americas: 17)
Montreal CMA 41 (Europe: 12, Africa: 6, Asia: 14, Americas: 9)
Vancouver CMA 24 (Europe: 8, Africa: 1, Asia: 11, Oceania: 2, Americas: 2)

Australian cities:

Greater Sydney 53 (Europe: 17, Africa: 4, Asia: 22, Oceania: 4, Americas: 6)
Greater Melbourne 46 (Europe: 16, Africa: 4, Asia: 21, Oceania: 2, Americas: 3)

And finally Greater London 80 (Europe: 26, Africa: 19, Asia: 22, Oceania: 2, Americas: 11)

Feel free to post more detailed data for these cities (or double check my data), or post for other cities as well.

[Bikemike]

This is an excellent and well thought out methodology. I would love to see a gradation of cut-offs starting with 5,000 and moving on up (10,000, 20,000, 50,000) to see how many nationalities each city sustain.

5,000 is a good starting point, but owing to the huge size of these metros, may be too small to amount to a viable community. Even a population of 10,000 can easily be swallowed up and diluted by a large metro.

[NorthernDancer]

Not surprised at all that Toronto is the most African metro in North America. I suspect others will be surprised though.

[Crawford]

Originally Posted by NorthernDancer

Not surprised at all that Toronto is the most African metro in North America. I suspect others will be surprised though.

It may be the most African metro, but the above analysis doesn't show it. It's just showing # of countries represented at a random cutoff.

Is there data on % African origin by metro?

[Docere]

Pretty sure either New York or Washington have the largest African-born populations in North America.

[Urbanguy]

The one unfortunate thing about ACS data is that it's not very extensive -- the countries of origin shown are limited.

[NorthernDancer]

Originally Posted by Docere

Pretty sure either New York or Washington have the largest African-born populations in North America.

Using the 2011 NHS and the 2014 1-year ACS, I get this:

Metro area (MSA or CMA): pct. of population born in Africa (population born in Africa)

D.C.: 3.24% (195,756)
NYC: 1.26% (253,348)
Toronto: 2.58% (142,180)

[NorthernDancer]

Originally Posted by Urbanguy

The one unfortunate thing about ACS data is that it's not very extensive -- the countries of origin shown are limited.

Yeah, I don't know how OP got the number of African countries. I can't find it. Only a breakdown by region.

[dimondpark]

Originally Posted by Docere

Bay Area* 38 (Europe: 8, Africa: 1, Asia: 18, Oceania: 2, Americas: 9)

I dont think it's just 2 from Oceania--Tonga & Samoa probably are 5,000+ as well, too bad the Factfinder stat doesnt list all nations.

[Bikemike]

Originally Posted by dimondpark

I dont think it's just 2 from Oceania--Tonga & Samoa probably are 5,000+ as well, too bad the Factfinder stat doesnt list all nations.

Yeah comparing metros is only valid within a common census criteria (within a single country). Different countries classify national origins differently. And for many nationalities, there are large sub-political groupings that are distinct ethnicities unto themselves but not counted (eg. Tibetan, Cantonese, Fukinese, Tamils, Gujaratis, Kurds, Uyghurs, etc). Most census-derived categorizations utterly fail to capture the obvious and equally important contribution to diversity that such sub-political cultures make.

For instance, lumping all immigrants from China into a monolithic "Chinese" category potentially dismisses the multitude of ethnicities within China, each of which may be as unique and distinct as, say, "French" or "German". A community of 5,000 Muslim Uyghur immigrants having little in common with the local Cantonese or Fukenese community would therefore be discounted from a metro's diversity etc. whereas the more mutually indistinguishable immigrants from the various British Commonwealths of New Zealand, Canada, and Australia would be given far more relative distinction/weight despite offering much less substance to real-world diversity.

My point is, take these numbers with a coarse grain of salt. But good thread topic all the same. One of these days if I'm bored I'll do metros starting at higher numbers.

[Bikemike]

What would be a more interesting measure of diversity (though measuring diversity can never be a complete science, like chasing rainbows and unicorns or comparing metro sizes) would be a tally of primary and seconary languages spoken at home using the similar cut-offs to the OP's.

[New Brisavoine]

Originally Posted by Bikemike

This is an excellent and well thought out methodology.

Except that it cannot be applied to all cities. For example in France, census figures are only published for the category "immigrant", which excludes the French nationals born abroad, so it cannot be compared with the foreign-born figures in the UK or US censuses, which include nationals born abroad.

Also, INSEE publishes the immigrant figures only for a small selection of origin countries, so it would be impossible to make a tally of countries for Paris.

Anyway, what we know from the census is that in the Paris Region (which equates quite well with the Paris metro area) at the Jan. 2012 census there were 2,782,834 people who were born outside of Metropolitan France (the European part of France), making up 23.4% of the total population.

The figure "people born outside of Metropolitan France" is the one that should be used when making comparisons with UK and US cities, because "foreign-born" in the UK includes people born in the British overseas territories, and in the US it includes people born in Puerto Rico and other US overseas territories (like US Samoa, etc).

In the City of Paris proper, 26.2% of the total population was born outside of Metropolitan France (Jan. 2012 census). In the suburbs and exurbs of the Paris Region, 22.7% of the total population was born outside of Metropolitan France.

[Crawford]

Originally Posted by New Brisavoine

The figure "people born outside of Metropolitan France" is the one that should be used when making comparisons with UK and US cities, because "foreign-born" in the UK includes people born in the British overseas territories, and in the US it includes people born in Puerto Rico and other US overseas territories (like US Samoa, etc).

Are you sure? I'm almost certain the U.S. Census doesn't count people born in Puerto Rico as "foreign born". They're American citizens, born in the U.S. You can see it in the Census tract level data, where, for example, a heavily Puerto Rican neighborhood will generally have a very low % foreign born.

Or did something change recently, or does the Census use "foreign born" for Puerto Ricans in some datasets but not others?

[New Brisavoine]

^^It was always like that. Check the definitions carefully.

[fflint]

Originally Posted by New Brisavoine

^^It was always like that. Check the definitions carefully.

No, you are wrong.

The Census Bureau website clearly states "Persons born abroad of American parents or born in Puerto Rico or other U.S. Island Areas are not considered foreign born."

[Docere]

Yes, there are limits to the methodology and it's far from the "be all and end all" of measuring diversity. However country of birth data allows for more "apples" comparison than an ancestry or origin question, which is measured very differently (and often not at all) in different countries. Language data that's consistent is also harder to come by.

Obviously different cities fare better in some areas than others. London does better than Chicago on country of birth data in terms of diversity, while Chicago would do better with an ancestry question, as London has a lot of "old stock" British/Irish while Chicago's would have more third and fourth generation Polish Americans etc. But I think few would dispute that London is the far international and cosmopolitan city.

[Docere]

Originally Posted by dimondpark

I dont think it's just 2 from Oceania--Tonga & Samoa probably are 5,000+ as well, too bad the Factfinder stat doesnt list all nations.

6,475 in Oceania, nec. Would either nationality hit 5000?

If the "other" country is obvious, I would include it. For instance Somalia, remarkably, still isn't classified as a country of birth but it shows up under Other Eastern Africa, should be counted for Minneapolis-St. Paul and Seattle.

[dimondpark]

Originally Posted by Docere

6,475 in Oceania, nec. Would either nationality hit 5000?

If the "other" country is obvious, I would include it. For instance Somalia, remarkably, still isn't classified as a country of birth but it shows up under Other Eastern Africa, should be counted for Minneapolis-St. Paul and Seattle.

It's not just 'Oceania' tho.

There are 50,000 Native Hawaiians/ Pacific Islanders in the Bay Area CSA(not even including Australians/ New Zealanders), 10,000 Samoans, 10,000 Fijians, 9,000 Tongans, etc and Im fairly certain that the majority of all 3 groups are born in their country, but then we dont know because only Fijians are singled out by the Oceania foreign born stat.

But I digress, I could be wrong too, but there are lots of Polynesians in the Bay Area.

[Docere]

OK, so far it looks like Paris and the Bay Area may be "shafted" by this criteria.

[New Brisavoine]

Originally Posted by fflint

No, you are wrong.

The Census Bureau website clearly states "Persons born abroad of American parents or born in Puerto Rico or other U.S. Island Areas are not considered foreign born."

I've reviewed this and I guess the confusion comes from the fact that I've always looked at 'place of birth' tables, which distinguish people born in the United States from people born outside of the United States (including Puerto Rico and US overseas territories).

You should be aware that what's called "foreign-born" in the UK census refers to all people born outside of the UK. So it's purely defined by place of birth and not by citizenship. See here:

A foreign-born population is defined by birthplace and not nationality or ethnicity. Birthplace, nationality and ethnicity are related, but to varying degrees. The UK’s foreign-born population will include people who have been British citizens since birth and others who have become British citizens since their arrival in the UK.

http://www.ons.gov.uk/ons/rel/fertil...chapters-8.pdf

So the foreign-born population in the UK includes the children of British people born abroad. For example the vast majority of "German-born" living in the UK are in fact children of British military personnel who served in Germany and had their children born there.

Also, "the UK" refers only and strictly to England, Wales, Scotland, and Northern Ireland. It does not include the Channel Islands and the Isle of Man, which are crown dependencies, and it does not include the British overseas territories either. You can see it in this table: http://www.ons.gov.uk/ons/rel/census...le-qs203ew.xls

So comparing what the American Community Survey calls "foreign-born" with what's called "foreign-born" in the UK census is not like-for-like. That's why I always do comparisons based on place of birth only (and not citizenship), and excluding the overseas territories. These are like-for-like comparisons.

In the US that would be people born outside of the 50 states and DC, whatever the citizenship of their parents.
In the UK that would be people born outside of the UK (England, Wales, Scotland, and Northern Ireland), whatever the citizenship of their parents.
In France that would be people born outside of Metropolitan France, whatever the citizenship of their parents.

Making a comparison based on the US definition of "foreign-born", which involves both place of birth and citizenship, would be impossible because as far as I know the UK does not collect information on the nationality of people at birth, so it is impossible to know who, among those born abroad, are children of British citizens, and who are children of foreigners. The children of British citizens born abroad should be excluded to have a like-for-like comparison with the US "foreign-born" population, and these people are very numerous (for example most of the "German-born", as I said).

Now based on these like-for-like figures, here is what we have.

The Jan. 2012 French census should be compared with 2014 ACS 5-year estimates. ACS 5-year is more precise than ACS 3-year or 1-year. The French census is also a 5-year estimate now, but more large-scale and precise than ACS. So the Jan. 2012 French census is in fact the result of aggregation of partial censuses which took place in Jan. 2010, 2011, 2012, 2013, and 2014. This matches with the dates of the 2014 ACS 5-year.

For the UK we only have the April 2011 census. Estimates for years after the census are only very rough, and not very detailed at a geographical level (no further than English regions I believe).

For Paris and London, we have the INSEE definition of metropolitan areas (based on 40% commuters) that I was able to apply to London thanks to the wealth of data from the 2011 UK census that was published. So I can give absolutely like-for-like figures based on two territories defined with exactly the same methodology.

For the US we have MSAs, which are based on 25% of commuters. Their extent is not exactly the same as the metro area defined by INSEE, but that's the only definition of US metro areas we have. I don't think metro areas based on 40% of commuters would have numbers very different from those based on 25% of commuters anyway (that's especially true for relative figures, percentages of the total population who are born abroad; absolute numbers may be more affected by the larger extend of the US MSAs).

So, with all this in mind, here are the figures. For Paris we only have figures for the Paris Region, but like I said it equates quite well with the Paris metro area based on 40% of commuters, so perfectly comparable with the London metro area listed below (see here and here for a detailed explanation of how I determined the extent of this London metro area based on INSEE's definition of metro areas). I have also determined the extent of the Manchester and Birmingham metro areas based on INSEE's definition. For France unfortunately we only have 'place of birth' figures for the départements, so I'm listing the départements which correspond more or less to the metro areas of Lyon and Marseille (Bouches-du-Rhône is slightly larger than the Marseille metro area, while Rhône is smaller than the Lyon metro area). For the UK we have 'place of birth' figures down to the smallest census tract, so I could reconstruct the numbers for the metro areas of London, Manchester, and Birmingham based on INSEE's definition of metro areas.

Absolute number of people born abroad (= outside of the 50 states and DC, or outside of the UK, or outside of Metropolitan France):
- New York MSA: 6,204,770
- Los Angeles MSA + Riverside-San Bernardino MSA: 5,546,755
- London metro area: 3,287,258
- Paris Region: 2,782,834
- Miami MSA: 2,396,773
- SF-Oakland MSA + San Jose MSA: 2,134,984
- Chicago MSA: 1,797,553
- Houston MSA: 1,467,402
- Washington MSA: 1,401,083
- Dallas MSA: 1,249,924
- Boston MSA: 886,988
- Atlanta MSA: 803,580
- San Diego MSA: 799,796
- Philadelphia MSA: 717,865
- Seattle MSA: 680,200
- Phoenix MSA: 677,136
- Orlando MSA: 542,919
- Las Vegas MSA: 474,300
- Tampa-St. Petersburg MSA: 458,095
- Sacramento MSA: 418,862
- Detroit MSA: 416,945
- Birmingham metro area: 390,198
- Denver MSA: 357,200
- Bouches-du-Rhône (15% larger than the Marseille metro area): 357,080
- Minneapolis-St. Paul MSA: 356,902
- Manchester metro area: 323,425
- Portland MSA: 313,295
- San Antonio MSA: 313,269
- Austin MSA: 299,106
- Baltimore MSA: 295,969
- Rhône (hosts 80% of the Lyon metro area's population): 275,946
- Charlotte MSA: 234,045
- Columbus MSA: 150,872
- Kansas City MSA: 148,296
- Cleveland MSA: 147,008
- St. Louis MSA: 143,901
- Indianapolis MSA: 133,456
- Cincinnati MSA: 101,979
- Pittsburgh MSA: 94,681

Percentage of people born abroad in the total population:
- Miami MSA: 41.5%
- SF-Oakland MSA + San Jose MSA: 33.5%
- London metro area: 32.1%
- Los Angeles MSA + Riverside-San Bernardino MSA: 31.9%
- New York MSA: 31.2%
- San Diego MSA: 25.1%
- Orlando MSA: 24.4%
- Washington MSA: 23.9%
- Las Vegas MSA: 23.7%
- Houston MSA: 23.7%
- Paris Region: 23.4%
- Seattle MSA: 19.1%
- Boston MSA: 19.1%
- Sacramento MSA: 19.1%
- Chicago MSA: 18.9%
- Dallas MSA: 18.6%
- Bouches-du-Rhône (15% larger than the Marseille metro area): 18.0%
- Austin MSA: 16.3%
- Tampa-St. Petersburg MSA: 16.1%
- Rhône (hosts 80% of the Lyon metro area's population): 15.6%
- Phoenix MSA: 15.6%
- Birmingham metro area: 15.5%
- Atlanta MSA: 14.7%
- San Antonio MSA: 14.0%
- Portland MSA: 13.7%
- Denver MSA: 13.5%
- Manchester metro area: 12.3%
- Philadelphia MSA: 11.9%
- Baltimore MSA: 10.7%
- Minneapolis-St. Paul MSA: 10.4%
- Charlotte MSA: 10.2%
- Detroit MSA: 9.7%
- Columbus MSA: 7.7%
- Kansas City MSA: 7.3%
- Cleveland MSA: 7.1%
- Indianapolis MSA: 6.9%
- St. Louis MSA: 5.1%
- Cincinnati MSA: 4.8%
- Pittsburgh MSA: 4.0%