Worldwide Demographics

[New Brisavoine]

The four France(s). Population growth from the 1962 census to the 2023 census in the Paris metro area vs the largest French metro areas outside Paris vs the rest of France vs the overseas regions.

1962 is when the the largest metro areas outside Paris started to grow more than Paris. Since 1962, growth has essentially taken place in those large metro areas outside of Paris (+85.7% for all the green metro areas).

For the entire France, the population growth from 1962 to 2023 has been +44.3%.

In comparison, the population growth in the US, UK, and Germany between 1962 and 2023 has been:
- USA: +81.5%
- France: +44.3%
- UK: +28.1%
- Germany: +12.6%

[Yuri]

Originally Posted by Yuri

France 2025 numbers are out: 66,351,959 inh.

Negative for the 2nd year: 610,065 births and 631,145 deaths, for a negative natural growth of –21,080 (–737 in 2024).

Natural growth in France was at +280,480 as late as 2006 and up to 2014 it was above 200k. In 2010, 802,000 births were registered and the number fell to a full 1/4 within 2025.

610,000 births were the lowest number ever recorded in France for the past 200 years except for the 1915-1919 and 1940-1942 periods. In 1825, for example, 1,026,604 births were recorded in France.

France and Argentina were always appointed as successful cases over holding on to high fertility levels decades after their demographic transition. France fell below 3 children per woman in 1949; Argentina in 1964. France got above 2 as late as 2010; Argentina in 2018. Now France plunged to 1.59; Argentina to 1.33.

And countries that tried to emulate the "French model", like Germany and Japan displayed some success over while making people believe the model worked. Japan rising from 1.26 in 2005 to 1.45 in 2015 only to fall to an all-time low of 1.12 in 2025; Germany 1.24 in 1994, kept on 1.3 up to 2011 and saw a spike surpassing the 1.5 for every year between 2015 and 2021, peaking at 1.59 in 2016. In 2024, back to 1.35.

This new declining trend is clearly universal and apparently unstoppable.

I'm waiting for the final 2025 numbers for the United Kingdom. As they have separated registrations for the four countries, they take longer to update.

We have the number of deaths already (648,894; flat against 2024) but births will decline. I guess it will be around 645k and Britain will join France into the negative natural growth club. Total population of 69,487,000 as 2025, and assuming a net migration of only 220k or so. The lowest level since the early 2010's. In 2023 it 745k, the highest ever recorded. In 2015, before Brexit, it was 300k.

In Europe, we'll only have Sweden, Switzerland, Norway, Ireland, Cyprus, Luxembourg, Malta and Iceland left on positive ground. Put together, they have 30 million people.

Till now, British and the French were the last big ones left, but now they're firmly on negative. Midsized countries like the Netherlands and Belgium plunged into negative few years ago.

But back to Britain, I wouldn't compare them with France and Argentina which were more peculiar cases, but they also enjoyed high fertility rates till recently like 1.96 children/women in 2008, 1.92 (2012) and 1.79 (2016). As the rest of the world, they started falling from there and in 2025 it reached 1.39, lowest level ever recorded. France didn't report the TFR for 2025, but I'll guess it will be around 1.54, also the lowest level ever recorded except for three years during WWI.

[New Brisavoine]

Your figures for France are... wrong.

The population figure of France (without including the overseas collectivities and New Caledonia), as published by INSEE last month, was 69,081,996 as of January 1, 2026, which is 2.7 million more than what you say, because you did not include the overseas departments of France, which are an integral part of France and the EU. It would be like giving the population of the UK without integrating Northern Ireland in it!

That 69,081,996 figure is itself an underestimate. The reason for that is because INSEE, unlike the ONS in the UK, doesn't have a migration survey. The net migration is known only AFTER the yearly census of population is published, and the yearly census is published 3 years after the census takes place, so in this case the last net migration we have is for the year 2022. For years after the census, INSEE simply use the average of the last 3 known years (in this case 2020, 2021, and 2022), which is a stupid method, but that's how it is.

Immigration in France has risen a lot after Covid, like in the rest of Europe, so using the 2020-2022 average leads to an underestimate of the population of France in 2026. Statistics from the ministry of the Interior show that immigration in France has actually continued to rise after 2022, reaching a historical peak in 2025 (unlike the UK where immigration fell back significantly after mid-2024).

So based on these statistics from the ministry of the Interior, the actual population of France on January 1, 2026 could be around 69,975,000 on January 1, 2026, i.e. considerably above the current estimate. The final figure will be published only when the 2026 census is published, i.e. in December 2028. That's when we'll know for sure. But the bottom line is: population of France on January 1, 2026 is closer to 70 million than to 69 million. And including the overseas collectivities and New Caledonia, the population of the French Republic has already passed the 70 million mark.

Originally Posted by Yuri

Till now, British and the French were the last big ones left, but now they're firmly on negative. Midsized countries like the Netherlands and Belgium plunged into negative few years ago.

France will return to positive territory in the coming years when the numerous cohorts born in 2006-2014 start having babies. Currently it's the small cohorts of the 1990s that are of child-bearing age. Also, this will correspond to the smaller cohorts of the 1950s dying (as opposed to the very numerous cohorts of 1946-1950 currently dying). So both movements will lead to the natural change in France returning to positive territory, even if fertility remains low.

Originally Posted by Yuri

France didn't report the TFR for 2025, but I'll guess it will be around 1.54, also the lowest level ever recorded except for three years during WWI.

INSEE estimates the fertility rate at 1.56 in France in 2025, but this is a preliminary estimate. Final figures published last week showed that the natural change in 2025 was not as negative as they had preliminary published in January (preliminary figure was a natural change of -6,000 in 2025, but final figures show a natural change of -3,045). The fertility was not readjusted yet (it will be readjusted only in January next year).

More importantly, contrary to what happens in all other European countries, the issue of births and fertility is a national obsession in France (and it's been like that since the 1890s due to the threat of a rapidly growing and militaristic Germany), so the declining fertility in France is a major topic in the French media at the moment, and I expect candidates in the 2027 presidential election will propose some policies to redress the fertility (such as improving family subsidies, which were cut by Hollande in the 2010s and which are largely blamed for the falling fertility, and also proposing measures to increase the construction of housing, which the prime minister already did last month). The matter is not going to be ignored as is the case in our neighboring countries, but unfortunately as long as Macron was president nothing was done because Macron just doesn't care about French demography (he himself has no children, and he was an advisor to Hollande when the cut in family benefits were implemented in 2013-2015).

[New Brisavoine]

Originally Posted by New Brisavoine

the declining fertility in France is a major topic in the French media at the moment

Ah, an example just from today! Les Echos is the French equivalent of the Financial Times and Wall Street Journal.

Translation: "Decline of births: The French housing market has become the most effective contraceptive in our country."

And the quote above: "France is facing a historic demographic turning point. It can no longer afford to ignore the fact that without a roof over their heads, there are no cradles."

[Yuri]

Originally Posted by New Brisavoine

Your figures for France are... wrong.

I told you it's European France. All long-term comparisons for France use Metropolitan data for compability.

Originally Posted by New Brisavoine

France will return to positive territory in the coming years when the numerous cohorts born in 2006-2014 start having babies. Currently it's the small cohorts of the 1990s that are of child-bearing age. Also, this will correspond to the smaller cohorts of the 1950s dying (as opposed to the very numerous cohorts of 1946-1950 currently dying). So both movements will lead to the natural change in France returning to positive territory, even if fertility remains low.

Not really. France lost lots of people during WWII that would be dying now and they're already missing from the death cohorts. Before that, it was the 1930's, the nadir of France birth rates, when population remained stagnant.

France also enjoyed a massive baby boom starting in 1946 going up qo 1973 (over 800k births every single year), coinciding perfectly with the post-war economic miracle. All those people will start dying off by the end of this decade and this process will end only by 2060. Till there, number of deaths in France will be higher and higher.

[New Brisavoine]

Originally Posted by Yuri

I told you it's European France. All long-term comparisons for France use Metropolitan data for compability.

Your post was not a long-term comparison. There's no point giving figures for Metropolitan France only when talking about France today. Why not remove the Canary Islands from Spanish figures, Hawaii and Alaska from US figures, Northern Ireland from UK figures, etc?

Besides, INSEE now publishes historical series for the entire France since the year 1975. That's more than 50 years. And before 1975 the overseas departments were so little populated as to make no difference in the overall figures of France (birth rates, fertility rates, etc).

Originally Posted by Yuri

Not really. France lost lots of people during WWII that would be dying now and they're already missing from the death cohorts.

France is actually one of the countries that lost the least people in WW2. Barely 550,000 died, as opposed to 6 million for Germany, also 6 million for Poland, 25 million for the Soviet Union. When France lost lots of people was in WW1 (1.4 million French soldiers died), but in WW2, thanks God, France suffered far far less.

WW2 was rather odd for France. France suffered considerably more material damage in WW2 than in WW1 (we were the most destroyed country in Western Europe after Germany), but considerably less deaths than in WW1. WW1 was a butchery for France, but material damage was circumscribed to a few towns and cities.

Then starting in 1946 the French birth rate exploded due to the return of the prisoners of wars which Hitler had kept in German camps for 5 long years in violation of international conventions, but by 1950 the birth rate declined, and was lower in the 1950s than either during 1946-1950 or during the 1960s, when it experienced a revival.

The big generations to die are those born 1946-1950 and those born 1960-1972. Those born in the 1950s are relatively less numerous. And in the 2030s, it is the people born in the 1950s who will be most concerned by death, just when the numerous generations born between 2006-2014 will be having babies, so the natural change of France will of course become positive again.

[Yuri]

Originally Posted by New Brisavoine

Your post was not a long-term comparison. There's no point giving figures for Metropolitan France only when talking about France today. Why not remove the Canary Islands from Spanish figures, Hawaii and Alaska from US figures, Northern Ireland from UK figures, etc?

Besides, INSEE now publishes historical series for the entire France since the year 1975. That's more than 50 years. And before 1975 the overseas departments were so little populated as to make no difference in the overall figures of France (birth rates, fertility rates, etc).

INSEE starts in 1975? Metropolitan France series start at 1820...

Sources I use mostly (Wikipedia, citypopulation.de) bring Metropolitan France numbers and I'm personally way more interested on those data then a weird combination between France, Martinique and Mayotte.

But few free to share other numbers. For the point I was making (birth and death curves) it doesn't matter.

P.S. Canary Islands has been Spanish for almost six centuries and it's mostly ethnic-Spanish. Ireland is linked to England in a way or another for one millenium. Alaska is full of White Americans. Hawaii is like 15% Native. Mayotte has been France since when? 40 years? It's like what? 0.5% European?

Originally Posted by New Brisavoine

Then starting in 1946 the French birth rate exploded due to the return of the prisoners of wars which Hitler had kept in German camps for 5 long years in violation of international conventions, but by 1950 the birth rate declined, and was lower in the 1950s than either during 1946-1950 or during the 1960s, when it experienced a revival.

The big generations to die are those born 1946-1950 and those born 1960-1972. Those born in the 1950s are relatively less numerous. And in the 2030s, it is the people born in the 1950s who will be most concerned by death, just when the numerous generations born between 2006-2014 will be having babies, so the natural change of France will of course become positive again.

Not really. Every year in the 1950's were above 800,000. I don't think those 2006-2014 people having children will be enough to offset around 800,000 deaths that will be expected for the next three decades.

There were only 610,000 births in France last year. It would require a massive increase there.

[New Brisavoine]

Originally Posted by Yuri

and I'm personally way more interested on those data then a weird combination between France, Martinique and Mayotte.

Ok. I'll think about it next time you post data about a weird combination between US, Hawaii and Alaska.

P.S. Canary Islands has been Spanish for almost six centuries and it's mostly ethnic-Spanish. Ireland is linked to England in a way or another for one millenium. Alaska is full of White Americans. Hawaii is like 15% Native. Mayotte has been France since when? 40 years? It's like what? 0.5% European?

Oh dear... Don't go to Ireland and tell them that. We might never see you again.

Martinique, Guadeloupe, Réunion, French Guiana have been French for longer than most part of Brazil have been Brazilian. As for ethnicity, are the Black people in Salvador de Bahia ethnic Brazilians.

If you don't understand how the French nation is defined, don't project your distorted view on reality. Whether you like it or not, a Réunionese is as French as a Provençal, and the population of France that matters for all international purposes is the one of the entire France, not just Metropolitan France (be it for the number of seats we have in the European Parliament, the contribution we must pay to the EU, our percentage share in the IMF, etc).

You are like a guy who would give the population of Atlanta only after subtracting the Black people, because somehow the Black people are not really American, only White people are. How ridiculous is that!

[New Brisavoine]

Originally Posted by Yuri

INSEE starts in 1975?

Actually, 1957. Even earlier than I thought. So you have data for the entire France since 1957 on the INSEE website. Therefore no "oh but we don't have long historical series" excuse. You don't even have data for Northern Ireland before 1957, so please!

It's just a 'race' thing, as you confess yourself. Just because a lot of people in these overseas departments are Black (but far from all of them in fact; in Réunion the Black people are only a rather small minority of the population), you don't consider them French. Let's subtract the 1 million Syrians from the German population then. And the Turks. And the Erythreans, etc. So, Germany with less than 80 million people?

Originally Posted by Yuri

Every year in the 1950's were above 800,000

There were 864,037 births per year between 1946 and 1950, vs 815,496 per year between 1951 and 1960, vs 851,587 per year between 1961 and 1970, so it's as I said, lower number of births in the 1950s compared to the 1946-1950 period and compared to the 1960s.

As for more than 800,000 births needed in the 2030s, you forget the lengthening life expectancy. You also forget that part of the people born in the 1950s have already died. As of January 1, 2026, only 7,653,170 people born between 1951 and 1960 were still alive, i.e. 765,317 on average for each year. And it's not like 765,000 people are going to die every year in the 2030s, because due to the lengthening of life expectancy theirs deaths will be spread over more years than their births.

[streetscaper]

Originally Posted by New Brisavoine

don't project your distorted view on reality.

lol yup, that guy doesn't care about actual data, just feels the need to be right, it's giving 'alternative facts'

[montréaliste]

There’s about 33,000 folks in my little suburb.

[New Brisavoine]

I'm extracting this from the Greenland thread:

Originally Posted by Yuri

There were some undercounting due fakenews (it was Bolsonaro era and they loved to spread all kinds of conspiracy), but not in Amapá, rural areas, smaller cities. Data there was collected quickly. There was indeed a reversal on migratory flux: lots of people from North moved to the three Southern states, something unimaginable till 15 years ago. Southern Brazil usually provided migrants for the farming frontier in Mid-West.

Undercounting was heavily centered on the big metro areas (hence several came negative: Belém, Recife, Salvador, Rio de Janeiro, Porto Alegre and others mostly flat: Fortaleza, Belo Horizonte).

They counted 203,080,756 people in Brazil in 2022. In Brazil, emigrants and immigrants usually cancel each other. Using natural growth as a parameter, some 5 million people is "missing". The most likely number for 2022 it would be 207-208 million.

Brazil's census, like all censuses, undercounts people. This happens more and more in our mobile society, and this was particularly the case for the 2022 Brazil census that took place during Covid.

That's why several countries have now stopped conducting a full (100%) census every 5/7/10 years, and instead conduct a partial (5%/10%/40%) census every year. France led the way in this (due to the undercounts of its last full census in 1999), and now various other Western countries are following French lead (Spain and Italy also now conduct yearly censuses, various smaller European countries, Netherlands and Nordics don't have full censuses anymore, US too has started a yearly census, called ACS, but due to Constitutional requirements they still have to conduct a full census every 10 years).

Maybe eventually Brazil will switch to a yearly census based only on a sample of the population. These yearly censuses actually count the total population better, because the national statistical institute can count and study more in depth the sample of the population selected. In the French case, 40% of the population is sampled over a 5-year period (8% every year), and in small municipalities 100% of the population is sampled every 5 years.

Anyway, due to the undercounts, IBGE (Brazil's federal statistical office) has had to "redress" the figures after each census. The redressing in 2022 was pretty large (more than 3%). A good census is considered to have a margin of error of only 0.5%, so more than 3% is pretty large.

They have published their redressed figures, and this changes quite a bit the vision of Brazil's population growth.

These are the raw (unredressed) census figures:
- 2010 census: 190,755,799
- 2022 census: 203,080,756

2010-2022 growth: +6.5% (i.e. +0.52%)

And these are the redressed figures, i.e. the real population of Brazil at each census:
- 2010 census: 194,749,329
- 2022 census: 210,862,983

2010-2022 growth: +8.3% (i.e. +0.66%)

So the population growth of Brazil between 2010 and 2022 has been higher than initially thought.

For comparison, 2000-2010 growth based on redressed figures: +11.5% (i.e. +1.09% per year)

Regarding Amapa, it was massively undercounted contrary to what Yuri says. The 2022 census counted 733,759 people in Amapa, but the actual figure (redressed by IBGE) was 795,457. So here the margin of error was more than 8%, i.e. considerably more than at the national level.

Redressed figures are unfortunately not published at the municipal level (as far as I know), so it's impossible to give the population of the Brazilian metro areas with the redressed figures, but it's important to keep in mind that all the population figures of the Brazilian metro areas underestimate their real population. For Rio de Janeiro for instance, 16,055,174 people were counted in the State of Rio de Janeiro at the 2022 census, but the actual (redressed) figure of that state was in fact 17,211,760. So that means the actual population of the Rio de Janeiro metro area should be around 1 million more than published by IBGE. So around 12.8 million instead of 11.8 million (raw data).

Originally Posted by Yuri

Regarding births, you're wrong: hospitals report births immediately (SUS) and these data is easily crossed with birth certificates and the Statistical Offices use the combination of those sources and release a year after to garantee all births are there. Births registered on later years represent something like 0.5% those days, and mostly due births on December. We're not in the 1980's anymore.

Check the IBGE website. A lot of births officially registered between 2010 and 2022 are actually births that took place before 2010. This would be particularly the case in a frontier state like Amapa. So it's impossible to know the actual natural change of Amapa, and therefore deduct its net migration.

[New Brisavoine]

Originally Posted by New Brisavoine

the declining fertility in France is a major topic in the French media at the moment

Another example of what I was saying yesterday. This is on the main page of Le Figaro today (one of France's two largest national newspapers):



And there were even more articles about redressing the French birth rate this morning when I checked, but they've removed some from the main page.

And this guy is one of the prominent French journalists on Twitter, with a long career. He's commenting on a news show he watched on the news channel LCI apparently. Translation: "It is obvious that France must have again a true and large pro-family policy." (pro-family policy means policies supporting the birth rate here, such as child benefits, free childcare, housing construction objectives, etc).



And it's like that just about every day here. The nation is obsessed with its (declining) birth rate. You see that in no other European (or Western) country. It will be an important topic during the 2027 presidential election.

[Yuri]

Originally Posted by New Brisavoine

Martinique, Guadeloupe, Réunion, French Guiana have been French for longer than most part of Brazil have been Brazilian. As for ethnicity, are the Black people in Salvador de Bahia ethnic Brazilians.

They were colonies, not France proper. What I meant was there no ethnic French/European in those places unlike say Alaska, Hawaii, Northern Ireland or Canary.

Brazil is a colonial society so there is no Brazilian ethnicity. In Salvador there are Mixed Brazilians, Black Brazilians and White Brazilians.

Originally Posted by New Brisavoine

If you don't understand how the French nation is defined, don't project your distorted view on reality. Whether you like it or not, a Réunionese is as French as a Provençal, and the population of France that matters for all international purposes is the one of the entire France, not just Metropolitan France (be it for the number of seats we have in the European Parliament, the contribution we must pay to the EU, our percentage share in the IMF, etc).

You are like a guy who would give the population of Atlanta only after subtracting the Black people, because somehow the Black people are not really American, only White people are. How ridiculous is that!

I understand but apparently you don't. "France" might mean different things. I'm not discussing about the official French Republic and its administrative divisions. I was talking about a very specific context (where it made sense to have a specific and frozen definition of France) and I made it clear. As Wikipedia does, as citypopulation does, as most who deal with France in this specific context.

If we were discussing saying GDP, I'd certainly bother to subtract those regions that are part of France.

[Yuri]

Originally Posted by streetscaper

lol yup, that guy doesn't care about actual data, just feels the need to be right, it's giving 'alternative facts'

People don't even know what's been discussing but needs to hate. Who are you again? What did I do that offended you. I have no recollection.

Yeah, alternative facts, I'm not alone on "alternative facts":

https://en.wikipedia.org/wiki/Demographics_of_France#Vital_statistics

The vital statistics below refer to Metropolitan France and do not include the Overseas France, territories and New Caledonia. Source: [45][46][47][48][1]

For the purpose of compatibility, all data refers to Metropolitan France.

Or:

https://citypopulation.de/en/france/

National (Metropolitan France)

Overseas Departments: Arrondissements and Communes
For each overseas department, the population of all arrondissements, communes and urban agglomerations.

Overseas Collectivities (collectivités d'outre-mer)

But sure, Wikipedia, City Population and I are all in a big conspiracy to make France to be 3 million smaller. It's our only reason.

[New Brisavoine]

Originally Posted by Yuri

They were colonies, not France proper.

Ireland was also a colony of Britain. So then, let's deduct the Northern Irish population from the figures of the UK? Be my guest!

Originally Posted by Yuri

What I meant was there no ethnic French/European in those places unlike say Alaska, Hawaii, Northern Ireland or Canary.

There exist no "ethnic French", just as there exist no ethnic Brazilians and no ethnic USians. The concept makes no sense whatsoever for a country made up of people from various language and ancestral groups, and largely shaped by immigration since 1850 (with more than 40% of French people having an immigrant background, including me by the way). Heck, even de Gaulle had an immigrant background! (from Ireland and Germany)

Originally Posted by Yuri

as citypopulation does, as most who deal with France in this specific context.

Citypopulation, the website that famously ascribes 14 million people to a so-called "London agglomeration", but only 12 million to Paris?

Citypopulation is good at certain things, like reporting census figures for administrative territories, but — at other things (like inconsistently reporting metro area figures with completely different definitions and comparing apples and oranges).

Originally Posted by Yuri

If we were discussing saying GDP, I'd certainly bother to subtract those regions that are part of France.

No one publishes the GDP of Metropolitan France anymore. IMF, World Bank, Eurostat, OECD, etc: it's the GDP of entire France that is published. Just as no one publishes the GDP of 48 contiguous US states, or GDP of Great Britain (without Northern Ireland).

[Yuri]

Originally Posted by New Brisavoine

Check the IBGE website. A lot of births officially registered between 2010 and 2022 are actually births that took place before 2010. This would be particularly the case in a frontier state like Amapa. So it's impossible to know the actual natural change of Amapa, and therefore deduct its net migration.

As I said, a tiny proportion these days. And as IBGE delays its releases, they manage to fill the gap. That's why there are small differences on the Ministry of Health numbers (DATASUS), Transparency Portal - Civil Registers and IBGE.

So regarding Amapá, we can definitely tell what's their migration patterns (and a small undercounting) with a 10% of certainty. Maybe 67,000 people haven't left Amapá, but it will be something between 40k-60k. Amapá has always attracted migrants (virtually all from the Northeast and Pará state), but on the last decade they've became a sender.

Lots of explanations: 2015-2016 economic crisis, 6 years of right wing governments where public and social investments were cut. Big migrations in Brazil were something from the past, ending in the 1990's but it surged by the end of the 2010's where people from poorer regions resumed their migration to the wealthier, dynamic ones.

Economy is very strong now, social programs are back, but I don't believe on reversal anymore. Migration will probably slowdown but I don't see the North (Amazon) resuming its role as a migrant magnet.

[New Brisavoine]

Originally Posted by Yuri

But sure, Wikipedia, City Population and I are all in a big conspiracy to make France to be 3 million smaller. It's our only reason.

I'll ask some Wikipedians to edit and change it. There's no reason why they shouldn't have the figures for the entire France when INSEE publishes historical series since 1957.

And in any case, in the main France article, it's the figure for the entire France that appears, not the figure for Metropolitan France (again, as explained yesterday, that 69.1 million figure is itself an underestimate, and the final figure will probably be around 69.9 million).

[Yuri]

Originally Posted by New Brisavoine

Ireland was also a colony of Britain. So then, let's deduct the Northern Irish population from the figures of the UK? Be my guest!

The Act of Union happened in 1801. Vital stats series start for Britain in 1838: https://en.wikipedia.org/wiki/Demographics_of_the_United_Kingdom#Vital_statistics

Mayotte has been a French colony and more to the point, part of France since...?

Originally Posted by New Brisavoine

There exist no "ethnic French", just as there exist no ethnic Brazilians and no ethnic USians. The concept makes no sense whatsoever for a country made up of people from various language and ancestral groups, and largely shaped by immigration since 1850 (with more than 40% of French people having an immigrant background, including me by the way). Heck, even de Gaulle had an immigrant background! (from Ireland and Germany)

That's why I used "ethnic French/European". Now you woundn't dare to say France and the French Empire were not racialist societies. That would be beyond absurd. "White" has a social meaning in France as much as in the US or Brazil. Each one with its own particularities, but race is a massive thing everywhere in Americas and Europe.

Originally Posted by New Brisavoine

Citypopulation, the website that famously ascribes 14 million people to a so-called "London agglomeration", but only 12 million to Paris?

I've never looked into their yearly agglomeration list and never referred to it. I don't believe in any global or even regional definition for metropolitan areas.

Aside that, I have pretty much all their data on my Excel tables, ranging from tiny US census tracts to big groupings provinces to match historical regions (e.g. Scania, Gotaland, Svealand in Sweden etc.).

They're amazing. Too bad you don't like it. The world goes way beyond London being bigger than Paris.

Originally Posted by New Brisavoine

Citypopulation is good at certain things, like reporting census figures for administrative territories, but — at other things (like inconsistently reporting metro area figures with completely different definitions and comparing apples and oranges).

They use official definitions in most places they exist: Canada, the US, Mexico, Brazil, France, Australia, etc.

In places that don't (Britain, Germany), they provide their own metro/urban area definition, with maps and they're also great.

Originally Posted by New Brisavoine

No one publishes the GDP of Metropolitan France anymore. IMF, World Bank, Eurostat, OECD, etc: it's the GDP of entire France that is published. Just as no one publishes the GDP of 48 contiguous US states, or GDP of Great Britain (without Northern Ireland).

That's what I said. Demographically, however, most do and that's precisely my point.

[Crawford]

I think the usage of Metropolitan France as apples-apples France is defensible. Usually, overseas territories are not counted in population, GDP, etc., even if the territory residents have citizenship, voting rights, etc.

Counting territories would have been pretty weird during the colonial era. Portugal would have probably been the most populous and poorest country in "Europe" in the 1970's, given its extensive African holdings.

And I don't understand the analogy with 48 states. There are 50 states. Puerto Rico, Guam, etc. are not counted as U.S., even though those are citizens with voting rights.