German Elections & Politics

[Polkergeist]

I have a question about overhang seats.

I understand that if a party wins more first vote constituency seats than their second vote entitlement of seats in any state/lander, then there will be extra compensatory seats for all other parties awarded to their lander list.

If that is so, doesn't that mean if a lander has an accidental overhang then that lander will get more seats then they are entitled to.

Or do I have this wrong?

[Anzeigenhauptmeister]

I have a question about overhang seats.

I understand that if a party wins more first vote constituency seats than their second vote entitlement of seats in any state/lander, then there will be extra compensatory seats for all other parties awarded to their lander list.

If that is so, doesn't that mean if a lander has an accidental overhang then that lander will get more seats then they are entitled to.

Or do I have this wrong?

No. It's quite complicated.
You probably assume that a Land has its own determined number of seats. But in reality, the votes are allocated differently.

First of all, you have to calculate how many seats - according to their second vote figure - a party gets for the whole federal territory. Then you have to take a look at where that party got their votes from:
100 seats are entitled to Party X in total; 60% of the party's votes come from Land A, 30% from Land B, 10% from Land C. Hence, Party X sends 60 party members from Land A to the Bundestag, 30 party members from Land B, and 10 party members from Land C.
If Party X has already won 25 direct seats in Land B, the remaining seats are filled up by 5 members of Land B's party list.
If Party X has already won 12 seat in Land C, they retain the two additional seats - the Überhangmandate ("overhang seats") -, and the number of the Bundestag members will increase by 2.

That method has been ruled unconstitutional by the Bundesverfassungsgericht ("Supreme Court") in 2008 due to the "negative weighting of votes"; that means if you didn't have voted for a certain party, they could have gotten even more seats. Plus, like in 2009 it is possible that a coalition gets more seats in the parliament than they got votes in the election.

Since the last election, the number of the total seats of the Bundestag has to be raised in such a way that Party X theoretically receives 12 seats from Land C via second vote (which are then, of course, set off against the direct seats). The additional additional seats are called Ausgleichsmandate ("compensation mandates").

[Polkergeist]

No. It's quite complicated....

Thank you for the explanation. It is much appreciated.

[President Johnson]

No. It's quite complicated....

Thank you for the explanation. It is much appreciated.

At this website, you can fill in percentage points for each party. It calculates the number of seats that result would translate into. And how many seats per state for each party.

[Anzeigenhauptmeister]

Wow! Awesome website! 👍🏻

It also shows which candidates would receive a Bundestag seat, distinguishing between direct mandates and list mandates. Strangely, it doesn't show which AfD politicians would be elected to parliament...

[Anzeigenhauptmeister]

No. It's quite complicated....

Thank you for the explanation. It is much appreciated.

There are two more things to be added, in case you haven't known them yet:
You can vote for different parties for Erststimme and Zweitstimme. That's what the FDP tried to convince their potential supporters of: Erststimme CDU/CSU, Zweitstimme FDP; for a long time quite successful, the last time they failed miserably with that strategy.
Furthermore, independent candidates can only run for a direct mandate or be a member of a party list, but they cannot run for second votes.

[Middle-aged Europe]

Far-right German candidate promises to get rid of Arabic numerals

LOL. Sounds like an election slogan from DIE PARTEI.

Well, a member of DIE PARTEI asked him the question hahaha

https://www.youtube.com/watch?v=FBiLSqz-DDg

That NPD candidate obviously didn't know what Arabic numerals are or what they look like, since he promised that he would replace them with (quote) "normal numerals" without going into detail which numerals he considers "normal".

Instead, he was solely triggered by the word "Arabic" and worked on the assumption that he has to oppose everything "Arabic". The fact that the question was asked by someone who had clearly identified himself as a member of DIE PARTEI and the fact that this question had caused some laughter from the audience, should have given him with a hint though.

[Polkergeist]

Thanks when I think of MMP, I always think of how New Zealand uses it given I am more familiar with their system. NZ is a unitary state and they just let overhangs.... just hang there.

On your other point, does the FDP not run for constituency seats at all?

No. It's quite complicated....

Thank you for the explanation. It is much appreciated.

There are two more things to be added, in case you haven't known them yet:
You can vote for different parties for Erststimme and Zweitstimme. That's what the FDP tried to convince their potential supporters of: Erststimme CDU/CSU, Zweitstimme FDP; for a long time quite successful, the last time they failed miserably with that strategy.
Furthermore, independent candidates can only run for a direct mandate or be a member of a party list, but they cannot run for second votes.

[Anzeigenhauptmeister]

Here is a map about the polls of the direct mandates on the basis of different pollsters.

https://wahl.tagesspiegel.de/2017/karten/direktmandate

Here's the latest Emnid poll:

[Anzeigenhauptmeister]

On your other point, does the FDP not run for constituency seats at all?

They do. But they have absolutely no chance of winning a direct mandate at all. The last Free Democrat to win a constituency was Uwe Lühr in Saxony-Anhalt in 1990.
Right-wing extremist parties like the NPD or the dissolved DVU often only ran for second votes, mainly because they feared that their votes may be too dumb to know of the difference between first and second vote.
The NPD also tried to run a Zweitstimmenkampagne ("second vote campaign") in the last Landtag election like the FDP used to do - just with the difference that they told their supporters to give them their second vote and their first vote to the AfD, which was enraged by this campaign, firstly because of the importance of the second vote and secondly because of their feared image damage.

[Tender Branson]

A new federal election poll out of NRW (the biggest state) for the ARD has horrible news for the SPD and also some bad news for the CDU, while the FDP is heading for a really nice win:



Changes compared with the 2013 result in NRW:

CDU: -4%
SPD: -6%
FDP: +7%
AfD: +4%
Greens: -1%
Left: +2%
Others: -2%

[parochial boy]

A new federal election poll out of NRW (the biggest state) for the ARD has horrible news for the SPD and also some bad news for the CDU, while the FDP is heading for a really nice win:



Changes compared with the 2013 result in NRW:

CDU: -4%
SPD: -6%
FDP: +7%
AfD: +4%
Greens: -1%
Left: +2%
Others: -2%

That seems like a surprisingly good score for Die Linke in a West German state - I guess it figures, if they are losing out heavily to AfD in their East German heartlands, and holding up better in the West, their might be less of a divide between East and West than there usually is.

[Tender Branson]

That seems like a surprisingly good score for Die Linke in a West German state - I guess it figures, if they are losing out heavily to AfD in their East German heartlands, and holding up better in the West, their might be less of a divide between East and West than there usually is.

The Linke has overpolled heavily this year though, especially in the NRW state election in May.

You probably need to shave off some 2% from their current polling. 5.5 to 6% seems accurate for NRW.

[palandio]

That seems like a surprisingly good score for Die Linke in a West German state - I guess it figures, if they are losing out heavily to AfD in their East German heartlands, and holding up better in the West, their might be less of a divide between East and West than there usually is.

The Linke has overpolled heavily this year though, especially in the NRW state election in May.

You probably need to shave off some 2% from their current polling. 5.5 to 6% seems accurate for NRW.

Before the NRW state election Infratest dimap overpolled the Linke by 0.1%. Before the Schleswig-Holstein state election Infratest dimap overpolled the Linke by 0.7%. Before the Saarland state election Infratest dimap overpolled the Linke by 0.1%. On average that's 0.3% to shave off.

8% for the Linke in NRW seems a bit high to me as well and might indeed be a combination of a slight systematic bias and a polling outlier, but your statement seems a bit exaggerated to me.

[Grand Wizard Lizard of the Klan]

Silesian accent in German elections. Candidate in district number 249 Main-Spessart / Miltenberg.

[Beezer]

Merkel choking?



http://www.wahlrecht.de/umfragen/yougov.htm

[Beezer]

Well, the demise of Schulz may actually have hurt the CDU/CSU in the sense that some of their dissatisfied voters may now once again feel they can afford to send a message to Merkel by voting AfD in light of the fact that Schulz is nowhere close to winning the chancellery.

[_]

Well, the demise of Schulz may actually have hurt the CDU/CSU in the sense that some of their dissatisfied voters may now once again feel they can afford to send a message to Merkel by voting AfD in light of the fact that Schulz is nowhere close to winning the chancellery.

We'll see, but AFD seems to be trending up in the latest polls, so you just might be right.

[Bumaye]

Merkel choking?



http://www.wahlrecht.de/umfragen/yougov.htm

 
 
Today 38,5% (Allensbach) and 38% (Forsa) for the CDU in other polls. YouGov is the only one conducted post-debate though. Merkel was solid but Lindner was way better the day after so that maybe brought two percent from black to yellow.

[Dr Oz Lost Party!]

Merkel choking?



http://www.wahlrecht.de/umfragen/yougov.htm

Literally only one poll and it's YouGov so take it with a fistful of salt.

[mileslunn]

Unless the polls change it looks like it will probably be either a Grand Coalition or Jamaica Coalition, maybe a Black-Yellow if polls are underestimating one or both party.  My understanding is the SPD doesn't want to continue the Grand Coalition while in the past the Greens have generally been unwilling at least at the national level to go into coalition with the CDU/CSU so do you think there will be any difficulty forming a government and if formed what is the most likely one?  Obviously if the Black-Yellow gets a majority than that will most likely occur, but assuming they don't, would it be a Grand coalition again or Jamaica Coalition.

[parochial boy]

How accurate have AfD numbers usually been in state elections?

I am wondering if, as they are a relatively "new" party with a somewhat less settled voter base (ie most of their voters this time round won't have been 2013 AfD voters) they might be somewhat harder to poll - which means their numbers could be off in one direction or the other.

Looking at the recent polling, there does seem to be more volatility in the the AfD numbers than any other party (except perhaps CDU/CSU)

[Beezer]

Yikes. Press F to pay respects for the SPD

[palandio]

How accurate have AfD numbers usually been in state elections?

I am wondering if, as they are a relatively "new" party with a somewhat less settled voter base (ie most of their voters this time round won't have been 2013 AfD voters) they might be somewhat harder to poll - which means their numbers could be off in one direction or the other.

Looking at the recent polling, there does seem to be more volatility in the the AfD numbers than any other party (except perhaps CDU/CSU)

Last poll done before the election by the respective pollster. Negative numbers mean underpolling, positive numbers mean overpolling.

Baden-Württemberg (March 13 2016)
FGW -4.1
YouGov -4.1
Forsa -4.1
INSA -2.6
Infratest -2.1

Rheinland-Pfalz (March 13 2016)
FGW -3.6
YouGov -1.6
Forsa -3.6
INSA -3.6
Infratest -3.6

Sachsen-Anhalt (March 13 2016)
FGW -6.2
Forsa -6.2
INSA -5.2
UniQma -7.2
Infratest -5.2

Mecklenburg-Vorpommern (September 04 2016)
FGW +1.2
INSA +2.2
Infratest +0.2

Berlin (September 18 2016)
FGW -0.2
Forsa -1.2
INSA -0.2
Infratest +0.8

Saarland (March 26 2017)
FGW -0.2
INSA -0.2
Infratest +0.3
Forsa -0.2

Schleswig-Holstein (May 07 2017)

FGW +0.1
INSA -0.9
Infratest +0.1

Nordrhein-Westfalen (May 14 2017)
FGW -0.9
YouGov +1.6
INSA -0.4
Infratest +0.6
Forsa -0.4

Conclusion:
After massively underestimating the AfD in March 2016, the pollsters were quite quick in adjusting their models, leading to relativly accurate (last) polls. What doesn't shine up when looking at the last poll from each pollster is that some pollsters go with the herd only in the last moment, Forsa's Berlin polls being a prime example.

[parochial boy]

Thanks Palandio. I guess that would seem to indicate that the AfD numbers are fairly accurate. Although, having said that pollsters have found a model that works for state elections, that still *may* not work for the Federal elections.

If (I assume) state elections tend to have a higher protest vote, or lower turnout, than the Federal election, it could still be the case that weighting the AfD as if it is a state election might lead to pollsters overrating the AfD score.