2013 Elections in Germany

[minionofmidas]

The fake-distribute-among-the-states-first* extra step was what the CDU thought up to arguably get around the (first) SC verdict. It didn't work. As anyone with good verdict-reading skills could have told them in advance. It's funny that it made it into the final compromise.
The final law means that the parties' national seat tallies will from now on be proportional, the seat distribution between states will not become proportional, the intra-state distribution between parties will not become proportional (but closer to it) in states where there is overhang, and will cease to be proportional in states where there isn't - the party that has overhang elsewhere now being underrepresented in these states - and the intra-party distribution between states will remain proportional for parties that have no overhang, and will remain unproportional for parties that have them. However, and crucially for CDU support, it will not become worse - unlike the quick fix the Greens' had suggested at the time of the first verdict, which otherwise shares all these characteristics without blowing up parliament like a balloon (on a 2009 rerun or worse. Which it doesn't actually look like happening; the CDU looks like increasing its share of the vote without increasing its share of the direct seats.) Indeed, the NRW or Lower Saxony CDU will continue to have a seat number proportional to their vote and the nominal (rather than actual) size of parliament, it just happens that these two will diverge further than in the past.


*note that only the federal sum from this distribution is guaranteed, not the per-state. On 2009 numbers, CDU Saxony-Anhalt gains an extra seat compared to the the real result on the first step, then loses it to some western state in step 5. Producing an extra CDU seat in step 3 and four extra equalization mandates, for SPD, FDP, Left and CSU, in step 4.

[peterould]

So basically the final result will be a proportional carve-out of all parties that get 5% or more, but the final number of MPs might be over 598 if some parties get an overhang.

[minionofmidas]

So basically the final result will be a proportional carve-out of all parties that get 5% or more, but the final number of MPs might be over 598 if some parties get an overhang.

Yes. (5% and/or three direct seats.)

[Franknburger]

So basically the final result will be a proportional carve-out of all parties that get 5% or more, but the final number of MPs might be over 598 if some parties get an overhang.

Basically yes, except for
1. The proportional allocation being done state-by-state, which, in the smaller states, tends to work against smaller parties. OTOH, these smaller states are either quite urban (Saar, Bremen, Hamburg), or in the East (MV, Thüringen), meaning that Grüne and/or Linke are not really small there. As such, the new system is especially disfavouring one party, namely the FDP.
2. A change in the allocation method from d'Hondt (favouring larger parties) towards Ste. League (favouring smaller parties), which was introduced to compensate for (1) above.

I will have to set up a spreadsheet over the next days to figure out how the system will work in detail.  However, my impression is it will work in favour of especially Grüne and Linke, and to some extent (provided they stay as "rather small" as they are now in several states) also SPD. This could make the election night even more interesting ....

A detailed description of the system (including tables of FPTP and "regular" mandates per state) can be found here in German:
http://www.wahlrecht.de/bundestag/

[Yeahsayyeah]

You must be logged in to read this quote.

This first distribution among the states is just a kind of fake because in the last step every party
is awarded a number of seats proportional to its vote share on the federal level. Seats will be added until every party matches at least the number of seats of the first step distribution by state, but each party's number of seat will still be proportional to its vote share. (This is actually done in the old method of distribution by party first and then by state). There is no real need for the first step as seats could be easily awarded until overhang is fully compensated on the state and federal level, which makes it so funny.

[freefair]

I think it was a stupid move on the part of the CD/SU not to allow the FDP to win 3 or 4 direct district seats and allow it to enter the Bundestag at any vote percentage. Could it have been so difficult or damaging to find 3 retiring constituency MPs in safe seats, ask the local branch to stand aside, and tell CDU backers to "first vote" the liberals? That's what I'd have done.

[Yeahsayyeah]

Such deals are not very common in Germany, at least they have not happened since 1957, and could likely backfire for perceived as an attempt to rig the election. It is common sense that every party has to stand for its own. It is one thing that some voters decide to vote for another party for tactical reason. Trying to force one's voters to vote for this other party would not make them very happy, particularly if the other party is the very unpopular part of the administration.

The other thing is, that the FDP is not that beloved by the CDU/CSU as well an there are even elements in the party that think a grand coalition would work much better.

[Middle-aged Europe]

I think it was a stupid move on the part of the CD/SU not to allow the FDP to win 3 or 4 direct district seats and allow it to enter the Bundestag at any vote percentage. Could it have been so difficult or damaging to find 3 retiring constituency MPs in safe seats, ask the local branch to stand aside, and tell CDU backers to "first vote" the liberals? That's what I'd have done.

Question is how you're gonna quell the ensuing rebellion within the CDU over this outragous procedure...

[Filuwaúrdjan]

Germany isn't France, basically. And even when that kind of dodgy nonsense is tried in France (the PS-Greens deal, say), there is often local resistance.

[minionofmidas]

Some local CDU grandee would stand. The vote split might kick the seat to the SPD. Worse, in a little known quirk of German election law, should the independent win, his voters' list votes would be disregarded.
And, of course, the FDP would get about one percent of the list vote in those three seats - 1% of the country after all, possibly votes it can not afford to lose.

But yeah, the real reason is that back in the 50s when such deals were occasionally tried, they were viewed (and, in the 60s - by now they're long forgotten - remembered) as attempts to rig the election. And of course there's the CDU and SPD (not so much CSU, interestingly) custom, which I'm not sure how old it is but is only now slowly - and I stress slowly - falling into abeisance - of reserving the first n seats on the list to the constituency direct candidates, n being the number of direct seats available. In the states as well as federally. Because otherwise that constituency party has been wronged and ignored, you see. (Really the effect, the intended as well as the resulting effect, is to ensure that there are, for instance, CDU MdBs from the Ruhr. And that its constituency parties there do not fall to the same joke level as Labour's in the Home Counties.)

[Franknburger]

Germany isn't France, basically. And even when that kind of dodgy nonsense is tried in France (the PS-Greens deal, say), there is often local resistance.

To give a specific example: The obvious choice for such a deal would be southern Würtemberg, where the FDP reached above 20% in the 2009 election. Rottweil-Tuttlingen, the FDP's best constituency in 2009 (21.9%), is out, as it is represented by the CDU parliamentary whip, Volker Kauder.

But neighbouring Schwarzwald-Baar (20.8% FDP), represented by younger brother Siegfried Kauder, could be a good choice. Except that in the local primaries, Siegfried Kauder was ousted by the party's rising star, 39-year old mayor of the city of Donaueschingen Thorsten Frei. Imagine the Federal CDU had tried a "safe FDP seat" deal there. It is first of all very questionable whether the local base had accepted it at all. Secondly, it might have lead to Mr. Frei running as independent, or even switching parties (Greens?), which could have been quite devastating to the local party.

Let's look a bit east, to Zollernalb-Sigmaringen (21.5% FDP in 2009), which has been represented since 2005 by Thomas Bareiß. Bareiß has been leading the CDU youth organisation in the state, and is a prominent exponent of the Merkel-critical right wing of the party. Outspoken against gay marriage, calling for "initiatives to control poverty immigration from Romaina and Bulgaria", etc. In other words- exactly the kind of person the AfD would love to have as their regional leader..

There are more drawbacks to the approach: Firstly, in CDU/CSU dominated states, you need to win a FPTP seat to become a CDU representative, there is hardly a chance for entering the Bundestag via the PV list.
Secondly, a good part of the CDU-FDP split is running along socio-demographic and especially confessional  lines. The CDU is traditionally the Catholic party, while the FDP represents the protestant/secular middle class. This is why south-western Germany has been and still is the FDP's  heartland. This is more than just a confessional divide, it is also a divide between protestant (Calvinist) small towns and the Catholic countryside. Now, in south-western Germany countryside does not mean agricultural - the terrain is quite mountainous, and already in the 19th century farm plots had become too small to secure family income.  Here, countryside means commuting some 30-50 km to work in manufacturing at Mercedes, Porsche, BASF or some medium-scale machine-builder, and spend the evening home in the village on the own vineyard. In other words - people that would be prime SPD electorate, if they weren't Catholic. If you provide these people only a choice between a (protestant, white collar) FDP candidate and a (equally protestant, or possibly Muslim, but blue-collar) SPD candidate, you can't be sure how that turns out.

[freefair]

Well, one can dream. I think that looking at the polls, the CDU, CSU, FDP are going to need to fight for every doggone single seat, tooth and nail, they can obtain especially if AfD (who are, lets face it, Germany's answer to Conservative Populism) or the Pirates surprisingly make it past the threshold. Given the difficulty of a Green or Grand coalition, it seems suicidal to be obeying this partisan code of honour right now, as giving them 4 seats (one for safety!) would effectively wipe out the threshold for just the FDP and give Black-Yellow a real chance of getting 299, or however many seats they need to govern stably.
The FDP are pretty much the liberal wing of the Merkel Party right now- If they fail to gain representation, the political consequences for Germany would from their point of view would be dire- when was the last time a workable traffic light coalition could have been formed?

[minionofmidas]

There would be an even easier solution to that: End the charade of having the FDP around. (You'd probably not be losing a statistically significant number of votes that way. Some economically conservative gays, and, well, that's it. There are other people who genuinely prefer the FDP, of course, but their second choice should be not in doubt.)

[freefair]

There would be an even easier solution to that: End the charade of having the FDP around. (You'd probably not be losing a statistically significant number of votes that way. Some economically conservative gays, and, well, that's it. There are other people who genuinely prefer the FDP, of course, but their second choice should be not in doubt.)

Raising the questio why have "any" seperate market liberal/libertarian and christian democrat/conservative parties anywhere?Why do they exist? Why bother with the Left or the Greens? Why not go the whole way to it's logical conclusion and have a 2 party system everywhere??

[Filuwaúrdjan]

Ah, but the FDP is a sham (or at least 50% of one), while those other parties aren't.

[minionofmidas]

More to the point, would Left voters have voted SPD in 2009 if the Left had not been around?
Some of them. More would have scattered all over the "others" or just stayed home.

Of course, not all the FDP extra voters from when parts of the country thought it hip (god, was that just four years ago?) would have voted CDU otherwise either. Especially in the East (in the West, it's pretty clear that almost all of them would have. Though they were anything but a representative sample of CDU supporters).

[Filuwaúrdjan]

(god, was that just four years ago?)

I know, right? It's almost as much fun as it is to recall that 'Cleggmania' was but three years ago...

[Franknburger]

This seems to be a good moment to inform the non-German observers about a fundamental feature of the German party system, namely the religious split.

First, take a look at the historical confessional split in Germany, as created and cemented by the 1648 Peace of Westphalia that ended the Thirty Years War. Below is the confessional map according to the 1890 census. Catholics are blue, protestants are brownish-orange. Note that this is 1890 borders and especially includes Alsace-Lorraine to the South-West. The Eastern border now runs a bit east of the page split along the Oder and Neisse rivers.

 

In general, the confessional divide runs from north-west to south-east, starting a bit south of the mouth of the Ems, and finishing a bit south of the western tip of Bohemia. In addition, you have:
1. A protestant "island" in the eastern lower Rhineland (Kleve) and along the Ruhr towards Eastern Westphalia.
2. A protestant belt running into south-western direction from Northern Hessen via the Rhein-Main region (Frankfurt) towards Western Palatinate (Kaiserslautern) and the Saar.
3. A second, parallel protestant belt that runs through Franconia (Nuremberg) and Würtemberg (Stuttgart) towards Bale/Switzerland.
4. The catholic exclave of the Eichsfeld east of Gottingen.

Below, a confessional map according to 2011 census results (note that the colours are reversed, catholic is orange, and protestant is violet. Blue indicates "no confession").



Aside from secularisation, especially in former East Germany and in/around major metro areas, the spatial patterns have remained stable.

This is the 2002 federal election. The left-hand side map is FPTP, the right-hand side is PV:


Alright, 2002, with a protestant Low Saxon (Schroder) running against a Catholic Bavarian (Stoiber) enhanced the confessional divide in the north, while diluting it in Bavaria. Anyway, the pattern is pretty obvious, and showing up in other years as well.

Finally, the FDP PV in 2009:



Protestant, small-town middle class in catholic rural areas that are, accordingly, dominated by the CDU.

[minionofmidas]

That's actually random slices of Baden-Württemberg, Catholic and Protestant alike, and the poshest parts of suburban West Germany.



Indeed... apart from the suburbs (but they are a relatively Catholic part of Hesse, as a matter of fact. And not only because of the incomers' majority, either, but ancestrally as well) and Waldeck (now THAT's a bizarre resurgence of traditional FDP-as-Protestant-Conservatives votes, though not one due to neighborhood of Catholics) the right half of this image is, basically, a map of Catholicism in Hesse.

[palandio]

Next sunday we will vote in Bavaria. Poll of polls (including FGW 09/06, Infr. 09/05, GMS 09/03, Emnid 08/28, I'm still hoping for a Forsa poll; when there where no number for Linke and Pirates I assumed 2%):
CSU 47.5% (+4.1% from last time)
SPD 19.3% (+0.7% from last time)
Free Voters 7.8% (-2.4% from last time)
Greens 11.8% (+2.4% from last time)
FDP 4.0% (-4.0% from last time)
Left 2.8% (-1.5% from last time)
Pirates 2.3% (+2.3% from last time)

Seat projection (180 seats):
CSU 100; SPD 40; Greens 24; Free Voters 16

My evaluation:
- The CSU has good chances to regain an absolute majority. There remains the risk for them that they underperform pre-election polls like they did in 2008 and like CDU/CSU has done in every federal election since at least 2002.
- Deluding expectations for the SPD, though the federal trend from Greens to SPD and tactical voting by Left Party sympathizers could provide them with (very modest) gains over their disastrous 2008 result.
- The FW outperformed their polls in 2008 by 3%, so maybe they are underestimated a bit this time as well.
- If the federal trend has a noticeable effect on the Bavarian state election, the Greens might stagnate. On the other hand a relatively good result in Bavaria (ca. 12%) might also help the federal party to reverse the trend.
- The FDP is campaigning hard on the message that a coalition government is much better than an absolute CSU majority. The Bavarian FDP lacks a traditional base, though, so the 5% hurdle might be to high for them.
- The Left is clearly under 5% in all polls. In recent state election this has translated in even worse election results. Their goal will be to limit the damage, because a very bad result in Bavaria could diminish their chances in Hesse.
- The Pirates will probably get their typical core vote result similarly to Lower Saxony where they got 2.1%.

What about the constituencies?
* München-Milbertshofen was the only constituency won by the SPD in 2008 by their front-runner Franz Maget. This time he will be replaced by Ruth Waldmann. Waldmann remains a slight favorite. Some areas in Southern and Western Neuhausen have been added to the constituency. From the East to the West these areas include inter-war social housing, co-operative housing (particularly railroaders), partially gentrified inner-city quarter, gentrified inner-city quarter, post-war social housing. These areas have a strong SPD potential (at least in the mean).
* München-Altstadt-Schwabing (named Schwabing in 2008) was the closest constituency in 2008. The SPD even got more "second votes" than the CSU, but Ludwig Spaenle (CSU) got slightly more vote than Isabell Zacharias (SPD). Margarethe Bause (Greens) placed third, but obtained a very strong result. All three compete for the constituency this time, too. The shape of the district has been changed and this might not favor the SPD. The constituency has lost Southern and Western Neuhausen (see above). It has gained Altstadt-Lehel. The Altstadt (old city) normally shows strong FDP results and relatively strong CSU results. The SPD is particularly weak. The Lehel, a very expensive inner-city residential quarter shows slightly weaker CSU results. The second area gained by the constituency is Isarvorstadt, which consists of a variety of quarters like Gärtnerplatzviertel, Glockenbachviertel (Munich's "gay quarter") and Schlachthofviertel. These are all typical inner-city quarters which have for a long time had a not so good reputation. Since the late 80s there has been a continuing gentrification process particularly in the Gärtnerplatz- and Glockenbachviertel. These quarters show the strongest Greens results in Munich, very low CSU results and dwindling SPD results (in some places still slightly above average). Vote splitting will hurt Greens and SPD.
* Other Munich constituencies where the SPD has slight chances are München-Giesing, München-Moosach, München-Bogenhausen, München-Hadern and München-Moosach. Note that by some kind of gerrymandering these are all very heterogenous samplings of Munich quarters and because of this they show very similar results. München-Giesing for example includes not only Giesing, a quarter with a "working class" reputation, but also Harlaching and Solln, which are mostly upper class residential quarters.
* In Franconia there might be a clear swing away from the CSU, particularly in the Nuremberg metro. The reason is that in 2008 the CSU's incumbent governor was Günther Beckstein from Nuremberg. Constituencies like Nürnberg-Nord, Nürnberg-West and Fürth might show much closer results this time, though they are likely still out of reach for the SPD
* In Freising the match will be between CSU and Greens. The CSU is still favored, but the Greens have obtained a very strong result last time. In the south of the constituency there is the Munich Airport which the CSU wants to extend. Protest against the extension is huge. Additionally the city/town of Freising is home to some agricultural/forest science faculties of the Technical University of Munich and to a Hochschule (college) specialized in agriculture/forestry/gardening/foor processing etc. Freising has been one an ecologist "stronghold" as early as 1978 when the ideologically rather interesting AUD obtained their strongest result there.

P.S.: I like the confessional map, Franknburger. I think that in 2009 the FDP spatial distribution does not really a strong correlation (because the suburban/wealthy factor is stronger), though in individual voting this correlation still exists. And it shows up in Norther/Central Baden-Württemberg.

[minionofmidas]

And it shows up in Norther/Central Baden-Württemberg.

Odenwald-Tauber being, of course, not only more Catholic but also much more rural (out of easy commute reach, unlike anywhere south. The road network playing a role in keeping the western half of the constituency remote - no motorways here) and, of course, very largely in Baden. Though no Baden vs Württemberg boundary shows up in the south.

[palandio]

Hopefully we will see an election in which the FDP is reduced to its core vote which will allow us to spot its core strongholds. I think that in 2009 the CDU loan/protest vote has caused a strong dilution in Southern Baden.
Pre-WWI Baden and Württemberg had a very clear divide between a catholic Center Party on the one side and mainly protestant liberal parties on the other side (the SPD was strong in cities with a large working class). Later (in 1952/53) the first governor of united Baden-Württemberg was liberal Reinhold Maier. Since then the liberals have faded, reaching only 5.3% in the 2011 state election. And so has most of its support in areas where it once had a majority. But not quite.
The following election atlas allows us to look at the FDP's strongholds in the 1996, 2001, 2006 and 2011 state elections more in detail:
http://vis.uell.net/bawue/11/atlas.html
To me most of the FDP stronghold look like traditional protestant DVP/DDP stronghold, with the exception of Donaueschingen and Bodensee.

[Franknburger]

The following election atlas allows us to look at the FDP's strongholds in the 1996, 2001, 2006 and 2011 state elections more in detail:
http://vis.uell.net/bawue/11/atlas.html
To me most of the FDP stronghold look like traditional protestant DVP/DDP stronghold, with the exception of Donaueschingen and Bodensee.

Wow - that's pretty interesting stuff.  Take a look at the top 5 FDP constituencies in 2006 (state election) aside from Stuttgart (which is specific due to the Stuttgart 21 debate), and their swing in the 2011 state election:

Freudenstadt (19.8% FDP in 2006)
FDP         -12.2
Grüne     + 9.4
CDU        + 1.2
SPD        + 2.0

Tuttlingen - Donaueschingen (16.4 %)
FDP         -  9.4
Grüne     + 8.2
CDU        + 0.3
SPD        + 0.6

Schwäbisch-Hall (14.4 %)
FDP         -  6.0
Grüne     +12.1
CDU        -  2,9
SPD        -   2.5

Waiblingen (14.1 %)
FDP         -  6.1
Grüne     +14.9
CDU        -  5,4
SPD        -   3.0

Rottweil (14.0 %)
FDP         -  8.2
Grüne     +10.0
CDU        -  2.9
SPD         +-  0

Now the same procedure for Rheinland-Pfalz 2006 vs. 2011:

Bernkastel-Kues (11.2 %)
FDP         -  5.5
Grüne     + 8.5
CDU        + 4.4
SPD         - 9.9

Betzdorf - Kirchen/Sieg (11.0 %)
FDP         -  6.1
Grüne     +10.3
CDU        + 2.1
SPD         -  6.9

Mainz (10.1 %)
FDP         -  4.5
Grüne     +15.8
CDU        + 2.4
SPD         -13.2

Bad Dürkheim (9.8 %)
FDP         -  4.8
Grüne     + 9.9
CDU        + 3.9
SPD         - 9.4

Wittlch (9.7 %)
FDP         -  4.9
Grüne     +  9,5
CDU        +  4.7
SPD         - 11.6

Now Lower Saxony 2008-2013. The pattern becomes clearer when you shift half of the FDP percentage to the CDU - 2013 was probably the most extreme case of FDP loan votes in German history.

Holzminden (14.2 %)
FDP         -  0.2
Grüne     +  7,5
Pirates    + 1.6
CDU        -   6.6
SPD         + 2.6
Linke       -  3.6

Diepholz (13.7%) (predominantly Catholic)
FDP         + 0.8
Grüne     +  5.9
Pirates    + 2.2
CDU        -   6.2
SPD         + 1.8
Linke       -  3.7

Hannover-Döhren (12.8%) - yeah, that is where the expensive housing is (Kirchrode, Bult, Südstadt), though it also includes a lot of blue-collar areas
FDP         + 1.3
Grüne     +  5.4
Pirates    + 2.1
CDU        -   8.6
SPD         + 4.6
Linke       -  3.9

Osnabrück (12.3%)
FDP         + 0.1
Grüne     +  8.5
Pirates    + 1.9
CDU        -   7.4
SPD         + 0.7
Linke       -  3.8

Oldenburg-Land (11.3%) (substantial Catholic share)
FDP         + 2.8
Grüne     +  7.7
Pirates    + 2.0
CDU        -   6.6
SPD         -  0.6
Linke       -  4.0

Any more questions on why the Greens are decreasing after they have emulated the SPD programme-wise?

B.t.w - Schleswig-Holstein 2009 vs. 2012 displays a similar pattern, except that here FDP losses in their traditional strongholds are mirrored by strong Pirate results (oh, those good old Dithmarschen and Frisian traditions...).

[ERvND]

The poll of polls numbers for Bavaria seem about right. My assessment of the parties:

CSU: Is virtually guaranteed to regain its absolute majority. Mind you, they fell only two seats short of an absolute majority when they were at 43%, so 45% should already be enough. I predict ca. 48%, though.

SPD: Ca. 19%. This will be perceived as a huge embarrassment to Christian Ude, while in reality, it will be a success. Without him, the state SPD would have fallen below 15% (as some early polls already suggested). This will also be the last time the SPD offered a credible candidate for Prime Minister. Next time, they'll nominate some unknown party apparatchik (again), resulting in 10-15%. The SPD's days as a major party are clearly over.

Greens: Ca. 11%. The federal trend may stall their progress, but they will clearly regain their position as third-largest party.

Free Voters: Ca. 7%. While they over-performed the polls last time, I don't see something like this happening today. Their main argument has always been "We are exactly like the CSU, without actually being the CSU." With the people back in love with the CSU, this attitude won't be convincing this time.

FDP: Ca. 3%. Bavaria is one of the very few states where the FDP's natural base is below 5%. Moreover, there are no reasons this time for tactical voters to switch their votes to the FDP. Their main argument, "We are the CSU's main opponent, that's why you should vote for us so that we can govern together with the CSU.", is utterly moronic and won't work. Ironically, the FDP's failure in Bavaria will help them federally, leading lots of loan votes in their direction.

Others: Ca. 12%. The "others" are traditionally strong in Bavaria, comprising not only the Left and the Pirates, but also "gems" like the ÖDP (a conservative environmentalist party) or the "Bayernpartei" (a Bavarian secessionist party), and, of course, Nazis, whereof the "Republikaner" are still shockingly strong in some parts of the country. All of those votes are wasted, and mainly serve the CSU's absolute majority.

[minionofmidas]

Catholic lakeside Baden is traditional DDP / DVP (the 1868-1910 DVP that is) stronghold country. Of course these Democrats were always Catholics (though they may well have been the smalltownfolk as opposed to the yeomanry. I don't have any numbers on that.)