Here's my prediction about the next German government.

The coalition will include the SPD and the FDP, but not the CDU. SPD and FDP will form a coalition with either the Greens or the CSU, the Bavarian wing of the CDU-CSU.

This prediction is based on the methods of cooperative game theory, with, of course, a few assumptions. One key assumption is that the Free Democrats would rather join the red-green coalition than be in opposition in a grand coalition government with CDU and SPD. If I am wrong on that, everything tilts toward the grand coalition. Another key assumption is that CSU and FDP would oppose any coalition in which the Unthinkable happens -- that the Left is in the government -- even if it brings them into the government too. I predict that the Left will NOT be in the government, but If the entry of the left really is excluded on principle as unthinkable, then a right wing government becomes possible.

Did anybody understand that? There will be a quiz -- but I must say Germany is giving me a helluva lot of fun!

I'll try to get more details up on my web site.

... a tolerated minority government.

SPD + Greens, tolerated by the Linke.

CSU leaving the CDU to form a coaliton with the SPD? Never.
FDP already ruled out a coalition with Red/Green or even with the SPD alone.

In theory not all your thoughts are without any logic. They are just not realistic.

FDP would hate the just previous suggestion -- an informal but de facto SDP-Green-Left "minority government" enough to overcome their rejection of the SDP-Green-FDP coalition. Now, I could be wrong about that, but that's not a matter of logic -- rather of opinion, on which logical people may differ.

Don't get me wrong. Your assumptions and reasoning really amused me and I don't mean it in a belittling way. I really enjoyed these assertions from the "outside" as they showed what could be possible although many in this country wouldn't even waste a second thought about some of these coalitions (SPD/CSU/PDS for example ;)).

I am also unsure whether I fully understood the purpose of this thread or the game theory in specific (I get the general idea) and at the moment I am not able to spend sufficient time for your article.

Nethertheless, I wanted to add why I mentioned the FDP's reluctance to form a coalition with the SPD but I don't know whether that effects your calculations.
The result for the FDP is also part of the fact that many who would have voted for the CDU/CSU, voted for the FDP. Obviously they did so to avoid a grand coalition. Almost 45% (or something like that) of FDP voters said that they rather see the CDU as their preferred party.
So after campaigning hard for a change, they just can't form a coalition with the SPD. They would lose badly in the next election. And yes, they would rather be in the opposition again. The FDP also doesn't have to fear a participation of the PDS as this is pretty much impossible. No party wants such a thing at the moment.

Of course, there are only my opinions and in my view regarding line 3, I wouldn't have chosen a 3 for the FDP or a 1 for the Greens (way lower ranks should be used).
I assume that several people see line 7 as more probable than line 3 but I don't see it happening either. I even regard line 1 to be more probable than line 3.
Concerning the Left (PDS), they always said that their aim was to make a CDU/CSU and FDP coalition impossible. They rather preferred the grand coalition so I think the 9 for them in line 1 is way too high because the grand coalition was what they wanted to achieve. They didn't want to have any participation in a government.

I know that you also included coalitions that are not possible but some questions for your game. What would you say if you included the following points in your model:
- CSU will never leave the CDU and I wouldn't seperate these two. Regard them as one party. That eliminates line 5 and line 6.
- CDU/CSU most certainly will never ever form a coalition with the PDS. But you already used lower ranks for these kind of combinations and as they are all dominated by other options, I assume that this doesn't play an important role anymore.
Considering the current political climate:
- PDS won't form a coalition with any party.
- FDP won't form a coalition with the SPD.
- There is a chance that the FDP would agree with a coalition with the CDU/CSU and the Greens. I still see it as highly improbable. The Greens would really surprise me if they did so. They just don't fit together as the FDP seems to be their nemesis.
- When I look at the behaviour of the parties at the moment, one should consider some other options as well: no stable coalition or new elections.

The game theory IS often sensitive to what one assumes about the preference rankings, and I could only guess at that, based on the principles I mentioned. Your points seem very cogent on that. I included possibilities of separating CDU and CSU mainly because I was interested in what might be behind Schroeder's attempts to do that. It appears, though, that CSU is just irrelevant to the present situation. They are what some approaches call a "dummy player" -- or worse.

On the other hand, FDP seems to be in a very key strategic situation, and their preferences are quite important. If I can find time -- and I am really stealing time for this -- I'll try putting your suggested rankings into the model. I suspect it will be consistent with your reasoning.

If SPD/Greens were to form a minority government relying on Left votes, then from the game-theoretic point of view, that is still a coalition -- even though it would not be formally seen as one. The informal nature of such (game theory) coalition probably would shield SPD and Greens from criticism of "doing the unthinkable," making that outcome more likely rather than less.

In any case, the essence of this situation is that some parties may find themselves choosing between alternatives both of which they hate a lot, and I think game theory is a good tool to capture that.

The logic is there. Judge it for yourself.

If you don't think cooperative game analysis is the right tool, fine. Subjectively, I'm not sure I buy the prediction myself. But this is what the numbers say (so far as I can work it out in a big hurry.)

Actually, as I was writing it all out, I corrected what seemed an error, namely that SDP-FDP-CSU could be stable. (CSU just doesn't matter very much with these numbers). I missed something on that, and may still be missing something. It would be more fun if critics would check it out, use the method, and see what I have missed rather than dismissing it on the basis of what somebody has said or ruled out. After all, FDP "ruling out" a red-green-liberal coalition is just a move in the game!

http://william-king.www.drexel.edu/top/eco/coalitions.html

You'd rather seperate Bush from Concertina Rice.

Still, it is shrewd of Shroeder to toy with the idea. I wanted to include even the "unthinkable" coalitions for completeness and to see what difference they make. As it turns out -- after I corrected my earlier error -- CSU just doesn't matter very much. To anybody. See the website for the corrected analysis, and much more detail -- and, hey, there could be another error using the method in there -- as opposed to differences of opinion -- and I would love to have it pointed out, if so. Lots of bragging rights in that. After all, I do have a book on this topic in print.

Write a few word description to go with each party acronym as I'm having a hard time keeping straight which parties are which.

SPD: center left social democrats

CDU / CSU: right wing conservatives (CSU more right wing than CDU)

FDP: right wing center

Greens: left wing ecological party.

Linke: far left socialists.

Hope that helps.

Forget about any governing coalition involving the Linke. All the other assertions by the various parties are more fungible than their rejection of Linke as a governing partner.

The natural alliances are SPD/Green and CDU/FDP. CDU and SPD are beyond actually cooperating with each other.

As much as they purport to hate the idea, any governing coalition will have to be three parties. Either Greens will split from SPD or FDP will split from CDU.

If I had to bet money I'd put it on the Greens leaving the SPD. Buh bye Schroeder.

My German contacts today said they think there will be another election in fairly short order.

The greens have ruled out forming any government with any "neo-cons" and with only the FTP as a partner Merkel doesn't have the votes for a majority so it looks like they will have a new election next Jan. with Schroeder as PM calling for them.

Although I believe the Green Party will form no government with the CDU, I believe that they would compromise with another party of roughly even strength.

... as of today, I don't see an oncoming election.

There still are lots of possibilities and once the parties have abandoned their current cocky behavior, I predict there will be another Schroeder chancellory.

With the CDU loosing big as they have lost their momentum.

Instead of the two small left parties definitely getting something, they each risk not getting anything if Schroeder only needs one of them after a second election.

Maybe Schroeder is waiting a few days to guage the mood. Maybe he wants a second election so that he can make the coaliting with only the greens. Or maybe the CDU will splinter with some moderates creating a center party. That seems very unlikely on short notice, but that party has got to be wondering WTF after having their worst election results in 46 years.

the CDU will do worse, since they're no perceived as losers.

So the small parties have to decide what the risk is if they don't for a government with Schroerder now, their votes might nor be needed after the next election. It could be that Schroeder and the Greens would get enough votes together to form a government.

(by the way I voted SPD)
After all, FDP "ruling out" a red-green-liberal coalition is just a move in the game! --rogerashton
Several of the parties made election promises not to form a coalition with certain other ones.
* FDP said it wouldn't join Red-Green
* SPD said it wouldn't deal with the left.
* CDU/CSU said they wouldn't join a grand coalition with Schröder still there.

The parties in Germany have long term credibility. The FDP got a reputation decades ago as a slut. She'd jump into which ever coalition she could. They have a voter mandate not to go back on their promises, but of coarse it's up to their constituents to decide if they'd forgive them or not. My personal belief is that a party could stay credible if the people who made these promises would step down.

For example: boot Merkel, they're looking for a way to get rid of her anyway. Let Schröder retire as chancler as part of a coalition deal with the CDU. Thus both parties gave up a little face. Might work with the FDP too.

This is something like what I hoped for anyway: Some right party's ideas are let in to continue to push reforms. The trash talkers are shut up or booted out. They shut up and fix the economy, with less gridlock from the Landestag.

... it already has this "slutty" reputation. It cost them more than a decade to get rid of the memories of their previous defection (from Schmidt's SPD to Kohl's CDU), it is higly unlikely that they will want to go through that experience again.

However, I agree that if someone is to defect, the liberal FDP is the most likely candidate, albeit without the current party leader Westerwelle.

Oh, no I wasn't saying that the FDP could afford to break their promise because they already have a that bad reputation. My point was that such a bad reputation sticks around for decades. The US parties have less of this. When a politician breaks some campaign promises, it doesn't follow his entire party around for decades. It follows the man.

It somehow doesn't seem fair to the FDP and CDU voters if the FDP were going to join the government. About half of the FDP seats are from "borrowed" CDU voters who were against a grand coalition. You could argue those who did vote CDU would be the ones to accept such a grand coalition.

There's also the practical issue that the CDU's reforms aren't far off of what the SPD would want to do. The FDP's suggestions are about as out there as the Left Party's.

It's all so damn interesting!

game theory analysis didn't take into account fully is the longer-term effect on alliances and voter loyalty. And those things are important. I tried to capture them, to some extent, in the preference rankings.

The FDP sounds a little like the Liberal Party of New York. (New York is one of the few American states where coalitions are possible. Indded there probably are no others.) Except the NY Liberals started out as FDR type liberals.

BUT -- from the game theory point of view -- if the grand coalition forms, FDP is irrelevant. Even if they are formally in the coalition, they bring nothing to the table that cannot be gotten without them, so they can't expect to get much out of it. By contrast, they can bring a huge amount to an SPD/Green coalition, and therefore can demand a lot in return. Surely these considerations weigh in for something -- even in Germany?

Their core ideology is less state and more personal freedoms. Their current incarnation though isn't really that serious about it though. They just want less taxes and less state, they still say the subscribe to the social market economy as do all the others.