Census 2011: Germany has 80.219.695 inhabitants

[Tender Branson]

One thing is already clear: Germany will have MUCH fewer people than what is now estimated (more like 80 Mio. and not 81.6 Mio. like it was estimated on Census Day May 9, 2011).

It will also be interesting to see how the population changed in the Eastern German states, where the last (GDR) Census was in 1981.

There could be really big flaws in the people's registers there, because many young people could have left the states for Western Germany without de-registering. Also, many dead people could still be on the registers. Also, if people moved from abroad to Germany and after a few months/years back to where they came from, without de-registering. And so on ...

Another important thing to watch if Hessen has enough inhabitants not to lose one of its 5 seats in the Bundesrat.

[Tender Branson]

Here are the benchmark population estimates for Germany and each state on May 9, 2011 (Census Day):

I calculated them by using the "Statistics Portal" monthly population estimates from 2011 and divided the End-April and End-May figures by 3, and added that to the End-April number.

10.760.000 - Baden-Württemberg
12.548.000 - Bayern
  3.473.000 - Berlin
  2.499.000 - Brandenburg
     660.000 - Bremen
  1.790.000 - Hamburg
  6.071.000 - Hessen
  1.638.000 - Mecklenburg-Vorpommern
  7.915.000 - Niedersachsen
17.836.000 - Nordrhein-Westfalen
  4.000.000 - Rheinland-Pfalz
  1.015.000 - Saarland
  4.140.000 - Sachsen
  2.326.000 - Sachsen-Anhalt
  2.834.000 - Schleswig-Holstein
  2.229.000 - Thüringen

81.734.000 - Germany

...

Here are my projections of what the OFFICIAL Census release will show next Friday:

10.545 - Baden-Württemberg
12.297 - Bayern
  3.369 - Berlin
  2.399 - Brandenburg
     640 - Bremen
  1.736 - Hamburg
  5.950 - Hessen
  1.573 - Mecklenburg-Vorpommern
  7.757 - Niedersachsen
17.479 - Nordrhein-Westfalen
  3.920 - Rheinland-Pfalz
     995 - Saarland
  3.974 - Sachsen
  2.233 - Sachsen-Anhalt
  2.777 - Schleswig-Holstein
  2.140 - Thüringen

79.784 - Germany

...

I assume an average error rate of ca. 2% in West German states, and 4% in East German states and 3% for Berlin compared with the benchmark estimates.

Of course there are always some surprises, but let's see how accurate my predictions are.

Hopefully my overall estimate for Germany is within 500.000 (79.3 million to 80.3 million) and my state estimates are within +/- 1% ...

[Tender Branson]

I'd think the overcount is most likely to be where the transient migrants are - Berlin, Hamburg and the rich Southern states.
There are of course also all sorts of shenanigans with official residency of German young adults (registered where their parents are rather than where they themselves are), but I doubt the Census actually caught that.

No, this Census only counted people where they are registered with their MAIN residence.

In the 1987 Census, people were also counted if they had a secondary residence somewhere.

This is why the current estimated numbers, especially in the big cities, will not match up with the new Census results next week.

The 1987 Census numbers have been fortgeschrieben (projected) to now, using main and secondary residences, therefore inflating the population almost everywhere.

For example, let's take Frankfurt, which according to the projected estimates has close to 700.000 people on Census Day 2011, but actually will have only 650.000 to 680.000 next week, because of the missing secondary residences, for example students - who like you said - are registered at home with their parents at their main residence, but study there. Or migrant workers, who have their main residence in let's say Hungary, but a secondary residence in Frankfurt, or well-off people who own a second or third property in Frankfurt while actually having their main residence in Berlin or so.

[Tender Branson]

Lewis, do you know if Hessen will automatically lose its 5th seat in the Bundesrat if it has less than 6 Mio. people, or does it depend on the overall size of the German population as well ?

I have read somewhere that states with more than 6 Mio. people must have 5 seats, but will this criteria change if Germany also has significantly fewer people ?

[Tender Branson]

BTW: Frankfurt already changed its reporting style to "Hauptwohnsitz" only in their latest release:

http://www.frankfurt.de/sixcms/media.php/678/07_Bev%C3%B6lkerung_4Q.pdf

678.691 people had their main residence in Frankfurt at the end of 2012, while more than 25.000 additional people had their secondary residence there.

...

We also have double counts here:

For example in Saalbach-Hinterglemm, which is a popular ski-area, there are 2.881 main residences and 4.848 secondary residences (mostly foreigners).

[Tender Branson]

Cologne is particularly adamant about putting only the misleading "including second residences" numbers in visible places... they usually have over a million inhabitants in that, as they like to think to have, but have never in fact had in the real numbers.

Just checked that:

Cologne has fewer secondary residences than Frankfurt actually. Only 20.000 vs. 25.000 in Frankfurt. Cologne also reports 2 population figures: They passed the million in 2010 and by Census Day it was ca. 1.010.000 with main residence and ca. 1.030.000 with secondary residences.

The Census will show if it really has more or less than 1 Mio., because of the register validations.

10.000 people is not a safe buffer.

[Tender Branson]

After thinking more about it, I will update my previous estimates because I think the errors will be slightly different (more like 1.75% in Bayern and BW, 2% in other West-German states and 2.25% in East German states, this time incl. Berlin):

10.572.000 - Baden-Württemberg
12.328.000 - Bayern
  3.395.000 - Berlin
  2.443.000 - Brandenburg
     647.000 - Bremen
  1.754.000 - Hamburg
  5.950.000 - Hessen
  1.601.000 - Mecklenburg-Vorpommern
  7.757.000 - Niedersachsen
17.479.000 - Nordrhein-Westfalen
  3.920.000 - Rheinland-Pfalz
     995.000 - Saarland
  4.047.000 - Sachsen
  2.274.000 - Sachsen-Anhalt
  2.777.000 - Schleswig-Holstein
  2.179.000 - Thüringen

80.118.000 - Germany

...

Average error rate for Germany: 1.98%

[Tender Branson]

Germany seriously hasn't done a Census since 1987?!

West Germany, yes.

And that was the costly traditional one (where Census takers went from door to door with their questionaires, like it is still done in the US).

In Eastern Germany, the last Census was taken in 1981 ...

So, the new updated data is needed very badly and the statistics folks need it as well to see how they can make the census process even more efficient for the 2021 Census.

[Tender Branson]

Just 1 day anymore ...

I'm already really excited.

The Census databank is already up (without data of course, which starts tomorrow at 11am):

https://ergebnisse.zensus2011.de/?locale=en

[Tender Branson]

I'm already really excited.

Im Ernst jetzt?

"Seriously ?"

Yes, I'm always very excited when Census-related stuff is released somewhere.

Especially now after 30 years.

[Tender Branson]

change compared to register / compared to Tender's estimate

Schleswig-Holstein -1.2 / +0.8
Hamburg -4.6 / -2.7
Lower Saxony -1.7 / +0.3
Bremen -1.4 / +0.6
NRW -1.7 / +0.3
Hesse -1.6 / +0.4
Rhineland-Pfalz -0.2 / +1.8
Baden-Württemberg -2.5 / -0.8
Bavaria -1.2 / +0.6
Saarland -1.5 / +0.5
Berlin -5.2 / -3.0
Brandenburg -1.7 / +0.5
McPomm -1.7 / +0.6
Saxony -2.0 / +0.2
Saxony-Anhalt -1.7 / +0.6
Thuringia -1.8 / +0.5

Ha, my Germany-estimate was within 100.000 and I got my goal of being withing 1% in my state estimates right in 13 out of 16 states. Only Berlin, Hamburg and RP were beyond my 1% goal.

After thinking further, I should have used a higher error rate for Berlin and Hamburg, considering they are huge cities with many foreigners.

RP was really unexpected.

And the Eastern states did not do as bad with the missing people. Maybe their registers weren't so outdated and flawed after all.

It seems that out of the 1.5 million people that were too much in the previous estimates, only 400K were from Germans and 1.1 Mio. from foreigners.

So, the German population only had an error rate of slightly more than 0.5%, while the foreigner population had an error rate of ca. 15% !

Very likely many of those just went home to their countries without de-registering.

Also, 1.9% of the population is Muslim - but that is probably too low, because the Religion question was voluntary and many people likely opted for "none" or "not stated".

[Tender Branson]

Religion by West/East:

West: 74% Christians, 6% Other Religions, 6% No Religion, 13% Not Stated
East: 27% Christians, 1% Other Religions, 33% No Religion, 39% Not Stated

Germany: 67% Christians, 5% Other Religions, 11% No Religion, 17% Not Stated

Those godless East-Germans ...

[Tender Branson]

They already published the new end-2011 estimate based on the May 2011 Census:

80.327.900

End-2012 is not yet ready, but maybe in the next months. Likely a gain of ca. 200.000 people, because of a immigration saldo of +400.000 and a birth/death deficit of -200.000

[Tender Branson]

Hessen had 5.993.771 people at the end of 2011.

So, what will they do ?

Will they use the 2011 Census numbers or the End-2011 numbers for Hessen (loses 1 seat), or will they re-apportion later, using the End-2012 numbers (keeps all seats, because more than 6 Mio.) ?

[Tender Branson]

Among cities with 100K+ populations, Aachen had the biggest negative error rate:

Aachen .................................... 236.420 258.246 -21.826 -8,5%

Berlin had the worst numerical error:

Berlin........................................ 3.292.365 3.471.756 -179.391 -5,2%

...

The best positive error rate (overcount) was in

Bergisch Gladbach..................... 108.878 105.690 3.188 3,0%

and the best positive numerical error was in

Bielefeld.................................... 326.870 323.146 3.724 1,2%

[Tender Branson]

States with highest share of migration background:

27.5% Hamburg
25.2% Baden-Württemberg
25.1% Bremen

States with lowest share of migration background:

3.3% Thüringen
3.5% Sachsen-Anhalt
3.7% Mecklenburg-Vorpommern

Cities (100.000+) with highest share of migration background:

Offenbach am Main....................... 48,9%
Pforzheim...................................... 46,6%
Heilbronn...................................... 46,1%
Frankfurt am Main.......................... 42,7%
Stuttgart........................................ 38,6%
Ingolstadt...................................... 38,1%
Ludwigshafen am Rhein................. 36,5%
Nürnberg....................................... 36,2%
Augsburg....................................... 36,0%
Mannheim..................................... 35,7%

[Tender Branson]

Citypopulation already has the 2011 Census numbers and End-2011 numbers for each district, using Google Maps:

http://www.citypopulation.de/php/germany-admin.php

[Tender Branson]

Hessen had 5.993.771 people at the end of 2011.

So, what will they do ?

Will they use the 2011 Census numbers or the End-2011 numbers for Hessen (loses 1 seat), or will they re-apportion later, using the End-2012 numbers (keeps all seats, because more than 6 Mio.) ?

The number of votes (they aren't, of course, actually seats) amends monthly. According to the state data as retroactively emended for the Census result, Hesse crossed 6 million inhabitants in June. They keep the vote.

Of course that still means we used a sixth vote we shouldn't have had for 16 years. (Or maybe there were ups and downs and we'd have been over 6 million for part of that time anyways? Like I care.)

States with highest share of migration background:

27.5% Hamburg
25.2% Baden-Württemberg
25.1% Bremen

That surprises you? Ever had a look at Switzerland? (Also, be sure to check the definition of Migrationshintergrund.)

I knew that BW had a high foreigner/migration background share, but I thought it was more in the Bavaria dimension (15-20%).

[Tender Branson]

States in which more German citizens were counted than what was estimated:

Bayern
Bremen
Rheinland-Pfalz
Saarland

https://www.destatis.de/EN/FactsFigures/SocietyState/Population/_Doorpage/Census_CurrentMigration.html?nn=142236

[Tender Branson]

(Also, be sure to check the definition of Migrationshintergrund.)

You must be logged in to read this quote.

[Tender Branson]

"Kazakhstan"

...

Germans in Kazakhstan coming back to Germany in the past 20 years I suppose, then counted as someone with "migrant background" ?

"Aussiedler und Spätaussiedler"

[Tender Branson]

States with the highest share of people older than 65:

24.8% Sachsen
24.3% Sachsen-Anhalt
23.2% Thüringen
22.6% Brandenburg
22.1% Mecklenburg-Vorpommern

States with the lowest share of people older than 65:

19.0% Hamburg
19.3% Berlin
19.4% Baden-Württemberg
19.5% Bayern
19.8% Hessen

States with the highest share of people younger than 18:

17.7% Baden-Württemberg
17.4% Niedersachsen
17.0% Bayern
17.0% NRW
17.0% Schleswig-Holstein

States with the lowest share of people younger than 18:

12.7% Sachsen-Anhalt
13.1% Thüringen
13.4% Sachsen
13.4% Mecklenburg-Vorpommern
13.7% Brandenburg

[Tender Branson]

The newspaper "Die Welt" has an article called "Did 1.5 million people just vanish ?"

http://www.welt.de/wirtschaft/article116703461/Sind-1-5-Millionen-Menschen-einfach-verschwunden.html

With a clickable map showing the error rates.

The article notes that the error is smallest in Rheinland-Pfalz with just 0.2% and they say that RP has a central population register, which the other states do not have.

Germany and the states also have formed a commission about how to improve the population registers and the procedures of registering people to make it more accurate (probably also for the 2021 Census).

[Tender Branson]

It should be noted that there is currently no centralized population register in Germany.

The reluctance to set up a Germany-wide CPR is due to abuses of such an institution by the Nazis to single out Jews and other groups who the Nazis wanted to find.

[Tender Branson]

The new census numbers will also impact the state fiscal transfer payments.

Wealthy German states like Bayern, BW, Hessen, Hamburg etc. are so called net-payers, while other states (mostly the Eastern states) are net-recipients.

The new numbers are brutal for Berlin and not so brutal, but still bad for Hamburg.

Berlin, as a net-recipient, will have to pay back 1 Bio. € for 2012 and 2013 and will receive ca. 500 Mio. € less in the coming years.

Hamburg as a net-payer will have to pay ca. 100 Mio. € more each year.

Rheinland-Pfalz on the other hand, which had the lowest error, will get a better position and will receive 150 Mio. € more each year.