German Elections & Politics

[palandio]

[...]
And according to Beatrix von Storch, Merkel is gonna be forced to flee into Chilean exile!

That's an allusion to Margot Honecker, former Eastern German Minister of Education and wife of Erich Honecker, the last leader of socialist Eastern Germany. https://en.wikipedia.org/wiki/Margot_Honecker
Von Storch is comparing Merkel, who has a DDR past, to Margot Honecker.

[palandio]

Exactly. My thought was that most international readers might have wondered "Why Chile?".
(Of course Paul Schäfer was also in a certain sense a German exilee in Chile, but I don't think von Storch wanted to compare Merkel to a pedophile, fascist cult and concentration camp founder.)

[palandio]

[...] Plus a very large percentage of the SPD vote is now coming from immigrants, [...]

I don't agree. At most 10% of people that are allowed to vote in German elections have a "migrational background" which includes having one foreign-born parent and the like. This also includes over 5 million ethnic German Spätaussiedler from the ex-USSR, Poland and Romania (ok, subtract the children), who certainly are not more pro-refugee than non-immigrants.
I would estimate that non-Spätaussiedler voters with "migrational background" make up about 4% of all German voters. About half of them might vote SPD, maybe less. So at most 10% (probably less) of current SPD voters are immigrants and their descendants. If that is a high percentage for you, then so be it.

[palandio]

[...] Plus a very large percentage of the SPD vote is now coming from immigrants, [...]

I don't agree. At most 10% of people that are allowed to vote in German elections have a "migrational background" which includes having one foreign-born parent and the like. This also includes over 5 million ethnic German Spätaussiedler from the ex-USSR, Poland and Romania (ok, subtract the children), who certainly are not more pro-refugee than non-immigrants.
I would estimate that non-Spätaussiedler voters with "migrational background" make up about 4% of all German voters. About half of them might vote SPD, maybe less. So at most 10% (probably less) of current SPD voters are immigrants and their descendants. If that is a high percentage for you, then so be it.

The estimates I've seen are that 60-70% of German Muslims vote SPD, with the rest mostly going for the left as well. But 10% of a party's support is quite a lot; the Democrats or Republicans losing 10% of their support is the difference between a nailbiter like 2000 and a landslide like 1980 or 1996.

So now it's only Muslims we're talking about, not immigrants in general? Muslim German citizens make up at most 3% of German citizens (depends on who you count as a Muslim). Keeping in mind the reform of citizenship laws under Red-Green, a very large chunk of them is probably under 18, while the older generations (35+) have to a large degree not requested German citizenship.

But 10% of a party's support is quite a lot; the Democrats or Republicans losing 10% of their support is the difference between a nailbiter like 2000 and a landslide like 1980 or 1996.

This is obviously different in a multi-party system. While I agree that losing 10% of a party's support matters (but 10% is a rather high estimation, given the fact that turnout among Muslims is likely lower than among non-Muslims), this is a bad example.

Exactly, even among those German Muslims that are allowed to vote, turnout is clearly below average. Altogether German Muslims are not (yet) a relevant voting bloc like e.g. Maghrebis in France or Pakistanis/Bengalis in the UK. They are also quite diverse. The relatively largest group probably has a Sunni Turkish "guest worker" rural Anatolian background. But what about the Alevis, the Bosniaks, the Kurds, the Lebanese, the ex-stateless Palestinians and the nominally Muslim hard-left Turkish and Iranian 80s political emigrants? Do they all vote for the SPD, the Greens and the Left for the same reasons?

[palandio]

Yes, it is likely and yes, of course taking power in BW is much better for Merkel than staying in opposition in BW. But compared to repeating the 2011 desaster everything looks good. Not taking power in BW is unimmaginable for the CDU. Losing five percentage points compared to the desaster of 2011, but somehow taking power in a coalition with the SPD or the Greens, would be better, but still very bad.

[palandio]

Saxony-Anhalt has quite a big potential for a right-wing protest vote. In the 1998 regional elections the far-right DVU https://en.wikipedia.org/wiki/German_People's_Union reached as much as 12.9%. This was quite a shock at that time. They had electoral spots like this:
https://www.youtube.com/watch?v=gG8bZ_gWBbQ

[palandio]

I'm scared that the main political divide in Germany tends to not be anymore between two or more respectable positions, but increasingly between an alternative-less mainstream consensus and the hateful anti-system fringe.

YouGov came out with a new online poll for the Baden-Württemberg and the Rheinland-Pfalz elections.

The new average of the last five polls (published between March 3 and March 10) for Baden-Württemberg is:

CDU 28.7% (-10.3% from 2011)
Greens 32.3% (+8.1%)
SPD 13.3% (-9.8%)
FDP 7.2% (+1.9%)
Left 3.6% (+0.8%)
AfD 11.7% (new)
Others 3.2% (-2.3%)

The new average of the last five polls for Rheinland-Pfalz is:

CDU 35.4% (+0.2%)
SPD 34.6% (-1.1%)
Greens 6.4% (-9.0%)
FDP 5.6% (+1.4%)
Left 3.8% (+0.8%)
AfD 9.6% (new)
Others 4.6% (-1.8%)

[palandio]

@DL: Yes, basically. The only other thing that could happen is that FDP or Greens fall below 5% because of the head-to-head race between SPD and CDU.

The current 5-poll average for Sachsen-Anhalt (Infratest dimap, FGW, INSA, Forsa, the fifth being UniQma instead of YouGov):

CDU 30.4% (-2.1%)
Left 20.0% (-3.7%)
SPD 16.1% (-5.4%)
Greens 5.3% (-1.8%)
FDP 4.3% (+0.5%)
AfD 18.0% (new)
Others 5.9% (-5.4%) [UniQma: Free Voters at 4%]

Personal predictions tomorrow or on Saturday.

[palandio]

Baden-Württemberg:

32.0% Greens
27.5% CDU
13.5% AfD
13.0% SPD
  7.5% FDP
  3.0% Left
  3.5% Others

26 constituencies won by CDU: Neckar-Odenwald, Biberach, Ehingen, Sigmaringen, Balingen, Main-Tauber, Freudenstadt, Tuttlingen-Donaueschingen, Rottweil, Aalen, Bruchsal, Wangen, Calw, Hechingen-Münsingen, Pforzheim, Schwäbisch Gmünd, Kehl, Neckarsulm, Eppingen, Enz, Hohenlohe, Backnang, Villingen-Schwenningen, Geislingen, Böblingen, Bretten
44 constituencies won by the Greens: All others (34 gains from CDU, 1 gain from SPD)

Rheinland-Pfalz:

35.0% SPD
34.0% CDU
11.0% AfD
  7.0% FDP
  5.5% Greens
  3.5% Left
  4.0% Others

24 constituencies won by SPD: Like last time plus Pirmasens (gain from CDU)
27 constituencies won by CDU

Sachsen-Anhalt:

30.0% CDU
20.0% Left
19.5% AfD
15.5% SPD
  4.5% FDP
  4.5% Greens
  3.5% FW
  2.5% Others

1 constituency won by The Left: Halle I
1 constituency won by AfD: Wolfen (gain from CDU)
43 constituencies won by CDU: All others (2 gains from Left, 1 gain from SPD)

[palandio]

Just voted for Baden-Württemberg at ca. 11 am local time. Cloudy, cold wind, but no rain or snow, should be good conditions for turnout. I saw quite a few people on the streets, clearly going to or returning from the polling place. Still turnout until now hasn't been very high, which is not too surprising given the nature of my ward, and even the (not so many) church-goers probably are voting only now.

Exit-polls at 6 pm, projections based on real counts at ca. 6:30 pm. Partial counts are not as usual as they are in other countries.

[palandio]

LOL. Take a look at Pforzheim...

http://www.welt.de/politik/deutschland/article153152717/Ergebnisse-aller-Wahlkreise-in-Baden-Wuerttemberg.html

Does not Pforzheim have an unusual number of unemployed and refugees? Seems like a good combination for the AfD vote.

Pforzheim has a high percentage of ethnic Germans from Eastern Europe, too. [Edit: Typo]

[palandio]

The Greens actually lost 0.1% in their absolute stronghold of Stuttgart I. Green gains and CDU losses are bigger in the districts that are close to the average, resulting in close Green wins in many districts.

[palandio]

[...]
   Concerning what I term ethnic de-Germanization, I'm referring to the situation where in many German cities, the % of schoolchildren whose parents or grandparents are from what is termed a "migration background" is quite high and growing higher.  

That's right, but these "migration background" numbers are somehow inflated. Does it count as ethnic de-Germanization when children have an Austrian parent or a half-Italian parent or a Transsylvanian Saxon parent? It's only natural that the % of children from a "migration background" is growing higher with his definition. (And yes, I do know that Muslims in Germany on average have higher birthrates, but this is only part of the picture.)

[palandio]

Most important net voter migrations in Baden-Württemberg (preliminary estimates according to ARD infratest dimap exit polls, categories being CDU, Greens, SPD, FDP, Left, AfD, others, none):

207k from none to AfD
188k from CDU to AfD
160k from SPD to Greens
151k from others to AfD
133k from none to Greens
109k from CDU to Greens
88k from SPD to AfD
86k from CDU to FDP
68k from Greens to AfD
61k from SPD to none.

[palandio]

Rheinland-Pfalz:

93k from Greens to SPD
77k from none to AfD
66k from none to SPD
65k from none to CDU
46k from CDU to AfD
43k from others to AfD
34k from SPD to AfD
22k from none to FDP
21k from Greens to CDU.

[palandio]

Sachsen-Anhalt:

104k from none to AfD
52k from others to AfD
42k from none to CDU
38k from CDU to AfD
33k from none to others
29k from Left to AfD
22k from SPD to CDU
21k from SPD to AfD.

[palandio]

   Do we have a breakdown of support by social classes in these elections?  Wonder if the AFD support is coming from working class groups, especially in SA?

According to ARD infratest dimap exit-polls the AfD in Sachsen-Anhalt got
36% among the unemployed
35% among blue-collar workers
22% among the self-employed and entrepreneurs
21% among white-collar workers
18% among the retired

http://wahl.tagesschau.de/wahlen/2016-03-13-LT-DE-ST/umfrage-job.shtml
http://wahl.tagesschau.de/wahlen/2016-03-13-LT-DE-BW/umfrage-job.shtml
http://wahl.tagesschau.de/wahlen/2016-03-13-LT-DE-RP/umfrage-job.shtml

[palandio]

Nice spaghetti diagrams on voter migrations:

http://www.zeit.de/politik/deutschland/2016-03/waehlerwanderung-landtagswahlen-parteien-cdu-afd-nichtwaehler

"Andere Parteien" = "Other parties"
"Erstwähler" = "First-time voters"
"Verstorbene" = "Deceased"
"Zugezogene" = "Moved in"
"Fortgezogene" = "Moved out"
"Nichtwähler" = "Non-voters".

Regarding Pforzheim: After WW2 Pforzheim saw a rapid expansion with a thriving export industry, specialized on jewelry, resulting also in a high influx of "guest workers". 46.6% of the population of Pforzheim have a "migrational background", which is the second-highest number among German cities. Since the 80s the jewelry industry has almost completely been resettled to Asia, resulting in the number of jobs declining from ca. 65.000 to ca. 50.000. Pforzheim today has an unemployment that is not high by national standards, but high by Baden-Württemberg standards. Also the economic decline has lead to a high share of affordable housing, exactly at the time when millions of ethnic Germans from Eastern Europe moved to Germany. Estimates on their percentage are mostly between 15% and 20%. (They also are counted among the 46.6% having "migrational background".) In recent years Pforzheim (as well as parts of Mannheim, interestingly) have also seen immigration of ethnic minorities from Bulgaria and Romania.

[palandio]

RP is by far the Greens' structurally weakest land in Western Germany (with the exception of Saarland), as you can see from many federal elections. Their core vote there is ca. 5%, the rest comes and goes (yesterday: goes). BW is on the other side of the spectrum (with the exception of the city-states Berlin, Hamburg and Bremen).

[palandio]

The Greens are the most pro-refugee party, hence SPD voters (and to a minor degree CDU voters) that are against the EU-Turkey deal from a pro-refugee standpoint, would go to the Greens.

Altogether I think that the vote transfers are a little bit more complicated than just CDU -> AfD.
Besides CDU voters the AfD is attracting non-voters and voters of splinter parties. In the East also The Left voters and in the West some SPD voters.

When it comes to AfD gains materializing on the federal level, keep in mind that on the one hand voters are often using state elections for a protest vote, but on the other hand the AfD's main selling points are federal issues and federal elections tend to have high turnout which would mean even more politically disaffected people voting.

[palandio]

Wonder how CDU\CSU are polling after this last two weeks especially in Bayern

Federal polls:
Forsa 07/18 - 07/22: CDU/CSU 35% (same as the week before), AfD 9% (same).
INSA 07/22 -07/25: CDU/CSU 31.5% (-0.5 compared to the week before), AfD 12% (+1.0).

[palandio]

Keep in mind that the published numbers usually aren't raw numbers and that pollsters are continuedly adjusting their secret formulas to account for all kinds of over- and underpolling.

[palandio]

Why would so many Greens prefer Merkel over Gabriel? The perception of Merkel being more of a "left-winger" on migrant issues than Gabriel? Is that also the reason why relatively many SPD voters would prefer Merkel over Gabriel?

Gabriel has been showing his talent of annoying everybody for quite some time. You often hear some inconsiderate remarks from him on all kinds of issues like TTIP, immigration, punishing people by withdrawing their drivers licenses, etc. and within a few days you hear him saying the opposite. You don't know what he stands for, and unlike Merkel who in the past often suffered from the same problem, he is very vocal about it. So Gabriel is not a strong candidate.

On the other hand Merkel has the incumbent bonus and the CDU and the Green electorate have been converging for years. Green voters today are much less anti-establishment, economically leftist and pacifist than 20 years ago. From the other side the CDU has modernized its stances on social issues. Plus the refugee issue has been an icebreaker, of course.

Regarding the positions of SPD voters: Despite everything that has occurred during the last 15 years I would think that many SPD voters today vote that way because they perceive the SPD as the party of the "little man", at least compared to the CDU. It's not that they are all super-progressives on all issues. Secular, tolerant and pragmatist mostly yes. But that's where the political spectrum becomes far more complicated than one or two dimensions. The recent right-wing surge does not only come from conservative CDU hardliners, but from many parts of the political spectrum, and hence it is an electoral problem not only for the mainstream center-right.

Time to start the prediction game for Mecklenburg-Vorpommern:
SPD 26.5%
CDU 19.0%
Left 13.4%
Greens 5.1%
NPD 3.0%
FDP 3.0%
AfD 24.0%
Others 6.0%

Turnout 61.0%

[palandio]

Berlin votes on sunday. The incumbent government is a SPD-CDU grand coalition under mayor Michael Müller (SPD). My prediction:

SPD 24.0% (-4.3)
CDU 17.0% (-6.4)
Greens 15.5% (-2.1)
Left 14.2% (+2.5)
Pirates 2.0% (-6.9)
FDP 6.0% (+4.2)
AfD 14.0% (new)
Others 7.3% (-1.0)

Turnout: 66.6% (+6.4).

Most likely coalition: R₂G.

[palandio]

Which estimates are you referring to? Infratest dimap (ARD) or Forschungsgruppe Wahlen (ZDF)?
Here's a graphic based on data from Infratest dimap: http://www.zeit.de/politik/deutschland/2016-09/waehlerwanderung-berlin-wahl-afd-cdu-abgeordnetenhaus

According to this ca. 46k AfD votes came from Others. Many of these probably came from the NPD which went down from 31k to 9k. But it seems that the AfD was also able to drain voters from all over the "other" spectrum.

The ARD data on former non-voters that came out to vote are different from the data you are referring to. According to ARD 69k out of 251k voted AfD. The reason for this is that probably the ZDF is calculating with net flows and that the ZDF doesn't differentiate between non-voters, natural demographic changes and migrational changes. In my opinion the ARD approach makes more sense.

Regarding the voter flow from Linke and Pirates towards AfD: Keep in mind that the Linke in 2011 came out from an unpopular government and was electorally reduced to an absolute physiological minimum. There just weren't that many generic "protest voters" left that possibly could go to the AfD. The typical Pirates voter on the other hand was more of an inner-city alternative type, although the Pirates also got above-average results in some of the Eastern housing-estates.

As far as I can judge most vote flow estimates today come from a combination of exit polls ("Which party did you vote for last time?") and district-based (multinomial logit) regression analysis. They are resonably accurate, but I would estimate their MoE to lie in the region of ca. 10k when speaking about the more interesting numbers. A comparison of ZDF numbers and net flows calculated from ARD numbers might give a hint.