2013 Elections in Germany

[palandio]

A closer look on the results:
In Franconia, Augsburg etc. the SPD gains have not been at the expense of the Greens, or at least the Greens have somehow managed to compensate losses to the SPD.
In rural Lower Bavaria on the other hand the SPD has lost votes compared to 2008.
In Munich the SPD has won München-Schwabing from the CSU and may even win a third constituency.

Deputies affected by the recent Spezlwirtschaft scandal have lost votes (this is true for all parties) and are doing worse than their parties.

The secessionist Bavaria Party's manifesto campaign has been very succesful. On the other hand the REPs and NPD were almost invisible during the campaign and are losing heavily.

[palandio]

Other observations: Turnout has increased clearly. CSU and partially SPD have benefitted from this. The Greens in most places (except Daxenberger land) kept their vote stable in absolute terms, but lost percentage points because of rising turnout.

It's interesting that at the moment with 49 out of 90 constituencies completely counted almost none of them is in Munich or Nuremberg and Upper Bavaria and Middle Franconia are generally a bit behind.

[palandio]

The main strategic problem for the FDP is that most voters are merely interested in Angela Merkel to remain Chancellor. And the FDP is not required for Merkel to remain Chancellor.

That's why the FDP needs the immaginary red-red-green bogeyman...

[palandio]

Nice maps!

Looking on the map of the Greens' gains/losses you can clearly detect the border of Upper Bavaria. The second vote in Bavaria is personalized and limited to each of the seven Bezirke. That is, you have the possibility to give your second vote to Christian Ude if and only if you live in Upper Bavaria. Because of the electoral system these type of effects does not fade out slowly but it halts at province borders. I think that this Ude effect accounts for many of the regional differences.

Heavy Green losses in South-Eastern Bavaria mainly come from some reverse Daxenberger effect, while relative strength in the same regions can partially be explained by some kind of residual Daxenberger effect.
The Greens' losses are mostly due to rising turnout. They have actually gained 20.000 votes compared to 2008. In Upper Bavaria they lost 1000 first votes, but 64.000 second votes (mainly Ude effect and partially reverse Daxenberger effect).
But I agree that some regional effects are difficult to explain. For example in the places where not everybody can afford to live. In some places like the Tegernsee valley (district of Miesbach, south of Munich) the Greens are not so succesful, FDP and CSU on the other hand yes. In the Ammersee region both FDP and Greens are very succesful. And in places like Breitbrunn (north-western shore of the Chiemsee, district of Rosenheim, almost Traunstein) both FDP and Greens are strong as well. Maybe this trend in South-Eastern Upper Bavaria is more recent than in the Oberland (Tegernsee etc.)

The Freie Wähler surge in Lower Bavaria? Maybe some combination of effects:
- The FW have become more concentrated on chairman Hubert Aiwanger who is from Lower Bavaria. As with Ude in Upper Bavaria, it was possible to vote Aiwanger with your second vote if and only if you vote in Lower Bavaria.
- The Ude effect in Upper Bavaria may have drained the anti-CSU vote there and sharpened the Upper/Lower Bavarian border.
- Ude was good in relating to people in the SPD "strongholds" (if you can say that in Bavaria), maybe not so good in the Eastern Bavarian countryside.

[palandio]

The interesting phenomenon is that even within the same metropolitan area (Munich, Frankfurt, Hamburg, Stuttgart) some posh suburbs are rather green and others not.

After the first proposal was declared unconstitutional, the federal electoral law has been changed in a way that all overhangs are compensated by additional seats, not only to preserve proportionality between the parties within one Land, but also proportionality between the Länder. (This is what I understood...) This may inflate the size of our parliament heavily, particularly if CDU/CSU get a relatively low proportional result due to loan votes for the FDP. I think that this has been discussed already some days/weeks before.

[palandio]

Thank you for the maps and the description of the situation in Hamburg.

Your explanation for the Upper Bavarian phenomenon is really good. This fits almost exactly the image of the places that I have. Regarding the FDP voters in Green exurban strongholds this is maybe similar to gentrification processes in the inner cities: Munich green strongholds like the Gärtnerplatzviertel have seen a rise in the FDP vote recently. Not only social psychologists like old houses with a lake nearby, but also attorneys, dentists etc. (often after they retire, as you pointed out).

[palandio]

My prediction:

36.9% [+3.1] CDU/CSU
26.7% [+3.7] SPD
  9.2%  [-1.5] Greens
  9.0%  [-2.9] Left
  7.3%  [-7.3] FDP
  4.7% [+4.7] AfD
  2.2% [+0.2] Pirates
  1.1% [+1.1] FW
  1.0%  [-0.5] NPD
  1.9% Others

Turnout: 75.0% [+4.2]

[palandio]

This election is not a good one for party dissidents:

Siegfried Kauder (ex-CDU) got 3.0% in his consituency Schwarzwald-Baar (Baden-Württemberg).
Wolfgang Neskovic (ex-Linke) got 8.1% in his constituency Cottbus-Spree-Neiße (Brandenburg).

Comparison: Martin Hohmann (ex-CDU) got 21.5% in Fulda (Hesse) in 2005.

[palandio]

Who is the winner of our inofficial prediction contest? (I hope that I didn't forget anyone and that I did the calculations correctly.)

According to the sum rule it's
1. Tender Branson 5.6 (best)
2. Vosem 5.8
3. Yeahsayyeah, Vasall des Midas 8.0
5. jaichind 8.2
6. palandio 9.4
7. ERvND 10.8
8. Franknburger 14.4

If we use the pythagorean distance we get
1. Vosem 2.69
2. Tender Branson 3.14
3. Vasall des Midas 3.57
4. Yeahsayyeah 4.15
5. jaichind 4.58
6. ERvND 4.64
7. palandio 5.41
8. Franknburger 7.32

All of us had the FDP wrongly above 5%; Tender Branson, jaichind and Franknburger were also wrong about the AfD.
The result that was responsible for most of the penalties was the CDU/CSU result. ERvND predicted the CDU/CSU surge best, but got too much penalty for the rest of his predictions.

[palandio]

You could call the Munich districts gerrymandered (on the federal and the state level):

On the federal level there are four districts that are all very heterogenous, but show very similar total results:
In München-Nord inner-city upscale Schwabing and Maxvorstadt are combined with the (relatively) poor northern perifery (e.g. Hasenbergl).
In München-Ost you have inner-city green strongholds like Au-Haidhausen, upscale suburban-type quarters in Bogenhausen and Waldtrudering, social housing in parts of Berg am Laim and Ramersdorf and West Germany's largest housing estate Neuperlach.
In München-Süd left-leaning inner-city perifery Giesing and Sendling are combined with upscale suburban Solln and Harlaching.
München-West-Mitte reaches from inner-city green strongholds like Isarvorstadt and Schwanthalerhöhe to suburban-conservative Untermenzing and Aubing.

You could easily carve out some left-leaning consitutuency in the southern inner-city.

On the state level it's the same, though the CSU has moved from the cracking tactic to the packing tactic in München-Milbertshofen when it shifted red Southern Neuhausen into this SPD-held constituency.

P.S.: @Vasall des Midas: You cited the result of some prediction contest and exchanged Left and Greens. If you want me to exclude you from the prediction contest, I will.

[palandio]

If you want me to exclude you from the prediction contest, I will.

God, no. I was just wondering. Why should I object to taking third place in a competition I didn't enter?

Though I also posted another set of figures later - something rumored to be a leaked exit poll, which in hindsight was probably true.

Union 41 #SPD 26 #FDP 4,8 #AFD 5 #Grüne 10 #Linke 9

Just out of interest - it certainly wasn't a prediction. Though I did say the votes to get AfD and FDP both over 5 might just not be there - how does that score?

There is a little problem with the leaked rumors because they don't include a Pirate result; overall they are slightly ahead of Tender Branson according to the some rule and very, very slightly behind Vosem according to the pythagorean rule (mainly because the leaked rumor underestimated the "others" share heavily). Though the prediction that the FDP would remain out would have won you an extra point. Sadly this was a bit too late to be a prediction that could be included without becoming unfair.

@Franknburger: Maybe the NPD result in Duisburg is due to local issues (i.e. conflicts about Roma immigration).

[palandio]

The main AfD platform in the election was against the Euro and I think many not-so-ideological 2009 Left protest voters voted AfD for that reason.

But the AfD has also an ideology other than that. That reaches from the CDU business wing to paleo-libertarians, from New Right ideologues to disappointed liberals and mainstream conservatives. It is really a strange fit for left-wing voters, but I think many 2009 Left voters did not identify as left-wing (I think there were even surveys about that).

[palandio]

@Sibboleth: Thank you for the map, I like it. Can you also make a map that shows relative changes (i.e. from 15% to 6% means -60%, from 10% to 3% means -70%)? That would be great!

Results from a place where I sometimes like to go by bike (it's a bit hilly):

Voting age population: 332
Voters: 190 (There are also postal voters, so turnout is higher; postal ballots are not counted in the village polling station, but down in town)
Valid PR votes: 190
CSU 153 (80.53%)
SPD 8 (4.21%)
REP 6 (3.16%)
Bavaria Party 6 (3.16%)
Free Voters 6 (3.16%)
FDP 5 (2.63%)
AfD 4 (2.11%)
Animal Welfare 1 (0.53%)
The Left 1 (0.53%)

No, I didn't forget the Greens, Pirates etc., they're all at 0.00%.

@ n/e rep butafly: The FDP has definitely a good chance for a comeback in my eyes.

[palandio]

I found an even higher CSU PR vote share, and it's even a whole Gemeinde (very small though).
Oberneukirchen (constituency of Altötting):
Voting age population 633
Voters 475
CSU 391 (82.7%)
SPD 18 (3.8%)
FDP 16 (3.4%)
Grüne 12 (2.5%)
NPD 9 (1.9%)
ÖDP 5 (1.1%)
AfD 5 (1.1%)
Linke 5 (1.1%)
Piraten 4 (0.8%)
Bavaria Party 3 (0.6%)
REP 1 (0.2%)
The Violets (Spiritual politics) 1 (0.2%)
MLPD 1 (0.2%) [That's actually above average ]
The Women 1 (0.2%)
Free Voters 1 (0.2%)

I think it's wrong to confound libertarianism with classical liberalism, ordoliberalism etc. There are no relevant libertarian parties/movements in the American sense in Europe.

The map in the Berliner Morgenpost is very nice. Munich has an election atlas here:
http://www.mstatistik-muenchen.de/wahlatlas_btw2013/html5/atlas.html?indicator=i4&date=Anteile&indicator2=i13&date2=gesamt

[palandio]

On the refugee/expellee issue: In Bavaria there are some post-WWII new towns that were founded by refugees and at least initially had a very high share of refugees. These patterns are now blurred a bit by Spätaussiedler and other people that have found affordable housing there. Nevertheless it may be worth to take a look at them (at first I always state the result for the Kreis level, then that of the town)

Stadt Kaufbeuren (voting age pop. 31405; voters 19628; turnout 62.50%; vote by mail 6617)
CSU 51.14%; SPD 17.48%; Grüne 8.13%; FDP 4.92%; AfD 4.81%; Linke 4.37%
Neugablonz (part of St. Kaufbeuren; voting age pop. 9376; voters 3657 [sadly not including vote by mail])
CSU 53.51%; SPD 17.37%; Grüne 4.31%; FDP 3.79%; AfD 5.56%; Linke 5.89%

Kreis Bad Tölz-Wolfratshausen (voting age pop. 90819; voters 65754; turnout 72.40%)
CSU 53.41%; SPD 14.38%; Grüne 8.26%; FDP 5.52%; AfD 5.56%; Linke 2.84%
Geretsried (part of Kr. Bad Tölz-W.; voting age pop.17006; voters 11002; turnout 64.49%)
CSU 50.83%; SPD 18.17%; Grüne 7.01%; FDP 4.89%; AfD 5.29%; Linke 3.84%

Kreis Traunstein (voting age pop. 131873; voters 91516; turnout 69.40%)
CSU 56.76%; SPD 15.01%; Grüne 9.08%; FDP 4.39%; AfD 3.06%; Linke 3.04%
Traunreut (part of Kr. Traunstein; voting age pop. 15004; voters 8766; turnout 58.42%)
CSU 56.27%; SPD 17.22%; Grüne 7.25%; FDP 4.12%; AfD 2.61%; Linke 4.33%

Kreis Mühldorf (voting age pop. 84331; voters 56608; turnout 67.1%)
CSU 59.4%; SPD 13.9%; Grüne 5.8%; FDP 4.0%; AfD 4.1%; Linke 2.8%
Waldkraiburg (part of Kr. Mühldorf; voting age pop. 16.935; voters 9669; turnout 57.09%)
CSU 57.5%; SPD 18.0%; Grüne 3.6%; FDP 3.9%; AfD 4.9%; Linke 3.4%

Kreis Regensburg (voting age pop. 145691; voters 103510; turnout 71.05%)
CSU 52.27%; SPD 18,20%; Grüne 6.58%; FDP 4.13%; AfD 4.02%; Linke 3.10%
Neutraubling (part. of Kr. Regensburg; voting age pop. 9260; voters 5582; turnout 60.28%)
CSU 50.77%; SPD 21.11%; Grüne 5.49%; FDP 4.91%; AfD 4.40%; Linke 3.86%


Observations:
- CSU share often slightly lower than in their rural surrounding, but higher than in other towns of similar size
- SPD share higher than in their rural surrounding, but similar to other town of similar size
- Grüne share very low
- FDP varies, maybe a bit low for towns of that size
- Linke stronger than in their rural surrounding, but normal for towns of that size
- Finally the AfD share. I would really have expected that some pattern would show up here, but this is rather disappointing... Maybe we could save the argument if we differentiate a bit. For example in the town of Bruckmühl (Kreis Rosenheim) the AfD share (6.39%) is the highest of the Kreis. Bruckmühl has many residential areas that were built a bit later than the typical post-WWII new towns. These residential areas also had a high share of refugees, but were less urban isles. Maybe it's these refugees and the milieu they established that is now a bit AfD leaning?

[palandio]

[palandio]

@ERvND: I think that there might still be distiguishable features in the voting pattern of expellees and there descendants, though more complex than "expellees = AfD", "expellees = CDU", "expellees = SPD" etc.

The Sudeten Germans have been a good example for that in the past. On the one hand their is a clear right-wing tradition (Henlein's party, Sudetendeutsche Landsmannschaft [particularly Witikobund]). Additionally in the late 1960s NPD surge some NPD strongholds were refugee towns. Then they are mainly Catholics which would imply a CDU/CSU tendency. But there among Sudeten Germans there is also a stronger Social Democratic tradition than among other refugees (Seliger, Jaksch etc.) and in rural Bavaria they were probably more SPD leaning than the rest of the voters.

You are right that the statistics from the refugee towns don't show any AfD tendency. When driving further conclusions they should be interpreted carefully. The growth of these towns has continued for several decades. Additionally they have a high Spätaussiedler percentage that probably causes low turnout and a relatively high CSU share.

[palandio]

[palandio]

Our Bundeswahlleiter has similar tables on his website:
http://www.bundeswahlleiter.de/de/bundestagswahlen/BTW_BUND_13/veroeffentlichungen/ergebnisse/
Sadly I found these tables only after I made the maps. His numbers are slightly different mainly because he calculates
[2013 absolute number of LINKE pr votes]/[2009 absolute number of LINKE pr votes]
while I have calculated
[2013 relative share of LINKE pr votes]/[2009 relative share of LINKE pr votes]
The basic pattern remains the same.

I made these quotient maps (for both FDP and LINKE) because I thought that a difference map would not be able to highlight all aspects I wanted to show.

Observations:

- Normally when a party loses votes it loses more in absolute numbers in its strongholds, but relatively more in its weak areas. For example the LINKE has gone down 5.8 points in the East, which is a fifth of their 2009 vote share, and it has gone down 2.7 points in the West, which is almost a third of their 2009 vote share. Similarly for the FDP. The interesting cases are when constituencies deviate from this basic rule (*).

- As Franknburger has pointed out earlier the FDP has hold up relatively good in its urban/metropolitan stronghold, but lost heavily in more rural/small town country, particularly in Catholic areas (e.g. Oberschwaben).
- In the East (except Potsdam), particularly in Saxony, the FDP has lost much more than in their Western weak areas (e.g. Ruhr, Braunschweig)
- The Linke has held up better in Saxony than in the rest of the East despite the basic rule*.
- In Saarland their is some kind of reverse Lafontaine effect.
- For some reason the Linke has remained relatively stable in NRW, in many rural areas (Paderborn?!) better than for example in the Northern Ruhr area, violating the basic rule*.
- In (Eastern) Bavaria and (Northwestern) Lower Saxony the Linke has lost heavily, sometimes (Aurich-Emden, some villages in the Bavarian forest) violating the Basic rule.
- The most striking pattern for the Linke is that they have held up much better in university cities (Münster, Freiburg, Aachen, Tübingen, Heidelberg and so on) in particular and Green strongholds in general. To a great extent this is probably not Green voters switching to the Linke, but sign of an academically educated "alternative" Linke core vote sharing the same milieu with the Greens.

[palandio]

Why there is no equilibrium between the CDU/CSU and the SPD? From 1949 to 2013, the SPD had  the chancellery for only 20 years. There were 18 federal elections and the SPD had more votes than the CDU/CSU only in 1972, 1998 and 2005. The margin in these three elections were small, while the CDU/CSU had huge victories until 1965 and after 1983.

There is much more equilibrium between the Labour and the Tories, the PSOE and the PP, the Democratics and the Republicans.

You should distinguish between the historic perspective and the present.
At the moment the parties left of center have a strucural potential of roughly half of the vote. So basically the SPD should poll about even with the CDU like in 2002 and 2005. But then
1. the SPD has at least two rival parties in its half (Linke, Greens, to some extent even Pirates), while the CDU/CSU at first had only the FDP, now FDP and AfD, but they are (at the moment) weaker than Linke and Greens.
2. the federal SPD brand is still a bit damaged (not as damaged as in 2005, though), while Angela Merkel is very popular. If you look on the state level Lower Saxony and Hesse both are slightly left-of-average, but had reasonably popular CDU/FDP governments. Both lost. The CDU was still stronger than the SPD because of point 1.
3. turnout peaked in 1998 and dropped sharply in 2009. The SPD needs to win back at least some of these voters.
4. AfD seems to have catched some Linke and SPD voters.

[palandio]

Do all municipalities in Hesse have CDU or SPD in second place?
In Bavaria there is quite a number of municipalities where the SPD is in third (or fourth) place. On the other hand there are not many municipalities that were won by the SPD. One example:

Stadlern, Kreis Schwandorf
voting age pop. 445
voters 250 (56.18%)
valid PR votes 247
SPD 96, 38.87%
CSU 94, 38.06%
LINKE 16, 6.48%
NPD 9, 3.64%
FW 8, 3.24%
GRÜNE 7, 2.83%
AfD 6, 2.43%
BP 4, 1.62%
REP 3, 1.21%
ÖDP 2, 0.81%
PIRATEN 1, 0.40%
FDP 1, 0.40%

The city of Teublitz (Kreis Schwandorf) on the other hand which was one of two (!) municipalities in Bavaria that went for the SPD in 2009 was won by the CSU 39.79% vs. 35.65%.

Maybe I will find another municipality not won by the CSU.

[palandio]

I still don't understand the phenomenon of 2009 one-time Left voters in places like Cornberg, Stadlern, St. Oswald-Riedlhütte etc.

Where did they come from? Where did they go?
SPD? CDU/CSU? Minor parties (FW, REP, etc.)? Abstention?

Why did they vote Left in 2009?
Economic crisis and resulting lay-offs/short-hours in some industries?

What explains these massive swings?

[palandio]

The fact that these places have a large SPD and Left potential doesn't surprise me so much. They all have a certain mining and/or industrial tradition.
(Interesting facts about Cornberg though.)

[palandio]

The northernmost Catholic enclave is a little Stadt that until 1802 was part of the Duchy of Westphalia and hence Catholic like the rest of this state (cuius regio, eius religio). In the Napoleonic Era the town changed its ruler several times (not voluntarily, of course). In 1817 Prussia, which had got the ex-Duchy of Westphalia in 1814, ceded the little town to the Electorate of Hesse, until in 1866 Prussia annexed the whole Electorate of Hesse.
Some neighboring villages belonged to Waldeck, which was (by its rulers) Protestant, others belonged to the Electorate of Hesse. But from the Napoleonic era on rulers did not force their subjects to change religion anymore, so the town remained Catholic.