The changed German MMP

With Germany’s federal election coming up within a week, Manuel has a very interesting post at Electoral Panorama about how the recently enacted changes to the electoral system work, and a simulation of their effect if they had been in place for the 2009 election.

I will not try to summarize it, but rather urge my readers to go over there, and then come back and discuss.

Thanks, Manuel, for the post.

32 thoughts on “The changed German MMP

  1. I suppose with any MMP system, the question boils down in essence to:

    If there are 100 seats (50 district, 50 list), and a party with 20% of the votes wins 25 district seats [*], should it end up having…

    A. 25 seats out of 100

    B. 25 seats out of 105

    C. 25 seats out of 125

    D. 20 seats out of 100 (possibly accomplished by some draconian method such as defeating the five district winners with the smallest magnitudes)

    [*] A bit unrealistic but it illustrates the math(s) more clearly. It is of course realistic that in a particular Land a party might win overhangs.

    • * smallest magnitudes…

      Replace with “smallest margins”?

      In any case, deleting candidates who obtained district pluralities would move the system outside the MMP category, and make it more akin to the PR with nominating districts on the order of a (now replaced?) system in Romania.

  2. Yom Kippur is barely behind us, and already I have committed one of the worst sins of blogging: leaving out a link. And been called on it within a couple of hours. Tough crowd!

    Now fixed. Thanks.

  3. Yes, smallest margins, sorry.

    Goof day for all of us.

    What about some compromise such as: they keep the district seats if they won 50% of the votes, but get tipped out if they had mere pluralities? I’m not advocating, just exploring the limits of the MMP concept.

  4. Thanks, Vasi. I note that the Plant Commission’s 1998 proposal “AV-plus” went for Option A (within each region), and I think Malta (if I understand correctly) has gone for Option C. Germany traditionally went with B.

  5. AV-plus was proposed by the Jenkins Commission in 1998. The Plant Commission proposed supplementary vote in 1993, but did not include a proportional aspect at all.

    I don’t think option A is MMP as it is not truly proportional–see the SNP getting a solid majority of seats on less than 45% of the vote in the 2011 Scottish Parliament elections. I think the disproportionality there was also caused to using smallish regions with d’Hondt, though, and not just letting parties keep district seats above their share of the vote.

    Malta’s system is roughly option C in a two-party parliament. You divide the disadvantaged party’s votes by the advantaged party’s votes per seat, round so the total number of seats is odd, and add those on as best loser seats.

    It appears that the new German system is therefore broadly similar to the Maltese system.

  6. I am not good at math. Please forgive me if I am wrong.

    Am I right in understanding that in the new system, every party receives overhang seats if one party earns an excess of direct mandates to ensure proportionality?

    Or, to use the much simpler description above, does winning 25 direct mandates out of 50 on 20% of the vote lead that party to earning 25 seats in a house of 125?

  7. Yes, it’s option C. However, it’s inaccurate to say every party receives “overhang seats” as there is no longer an overhang in parliament. Instead, the other parties receive “balance seats” or “leveling seats” in order to restore proportionality.

    The result in 2009 would have been a much bigger Bundestag and a much smaller majority for the CDU-CSU-FDP coalition.

    At the next election, Merkel may have a bare majority if the FDP crosses the line. In that case, it’s unclear whether she’d try to continue with a small majority or whether she’d enter a “grand” coalition. If the FDP fails to gain 5%, it’s possible the CDU-CSU alone may win a majority of seats, but it’d be a exceptionally small one, and she still might prefer coalition with the SPD.

  8. Interestingly, one of the scenarios described by Chris would have come to pass had the new electoral system been in place back in 1994: the ruling CDU/CSU-FDP coalition’s Bundestag majority would have been reduced from a narrow but manageable 341-331 to a precarious 347-345, and it’s not clear what path then-Chancellor Kohl would have chosen to pursue under such circumstances.

  9. There seems to be something of a consensus among those in the now that a narrow majority government isn’t something that would work in Germany and that even a grand coalition would be better. Why?

    Is party discipline poor enough in Germany that only a large majority can b workable?

    Is there a great deal more instability between parties in coalitions that means chancellors want security in not having the tail wag the dog?

    Does Germany just want something closer to a Swiss style consensus government than an Anglophone winner take all system?

    I feel that I know more than the average person at large about German elections work, but it seems I don’t know that much about how German politics work.

  10. Mark, I would think the concern is that a narrow Bundestag majority could be undone by a handful of defections, as was the case during Willy Brandt’s first SPD-FDP coalition government of 1969-72 – which in turn led to a failed vote of no-confidence, and an early Bundestag poll in 1972.

    That said, the SPD-Alliance ’90/The Greens coalition government that took office in 1998 continued in power with a nine-seat majority after the 2002 Bundestag election, and as noted on an earlier comment, CDU and FDP renewed their coalition agreement in 1994, even though their joint majority was down to just ten seats.

  11. Am I right in understanding that apportionment between the states might become even worse now, because overhang seats from one state are reduced from a party’s list seats in another?

  12. JD, that is correct. By the way, the mechanism for allocating seats among the state lists of a qualifying party is essentially the same as that used in Norway for the nationwide distribution of Storting seats among competing parties.

  13. Is Germany know becoming like Austria in the sense that coalitions are between the two largest parties? Is this a good thing? Both the CDU/CSU and the SDP are losing votes toward the smaller parties.

    The next election in 2017 will probably lead to the CDU/CSU and the SDP if they are in a coalition together will lose seats. If this becomes the norm, then the two largest parties will soon not be able to form coalitions with each other as they won’t be able to have a majority together. This is perhaps what is going to happen in Austria. Voters might as well not even vote if they can’t even change the government, since all parties are alike, have been in bed with one another, and are going to be in bed with one another.

    The question few seem to ask is why can’t a government be formed with CDU/CSU and the Greens. That would be an interesting experiment. Another possibility is why can’t Germany have minority governments. Is there something in the way the Chancellor is selected that precludes the possibly of minority governments? Is there no negative parliamentarianism in Germany, but somehow the Constructive Vote of No Confidence?

  14. The CDU-CSU and Greens have almost no ideological similarities. The Gruene are not like the Canadian Greens, who are really a fiscally centrist or center-right party on everything but the environment, but instead are generally to the left of the SPD. The Greens and CDU formed coalition together in Hamburg from 2008-2011 but the Greens collapsed it a year before elections were needed and the parties went on to lose to an absolute majority SPD.

  15. JD is correct that this will mean the division of seats between states is even less proportional within a party’s caucus than it has been in the past. However, it’s my understanding that officially, MPs represent the whole country and not one constituency or Land, and therefore there is no constitutional issue with this. That doesn’t change the fact that some regional interests are better represented than others.

    Also worth noting is that the Eurosceptic Alternative for Germany is polling surprisingly close to 5%. It’s possible that either, both, or neither of them and the FDP could cross the threshold. While AfD is broadly a center-right party, I can’t see them being too happy to form a government with Merkel nor she with them if they cross the line. Regardless of whether the FDP get in, AfD getting in will greatly increase the likelihood of a CDU-CSU-SPD coalition as it reduces the number of seats every other party holds proportionately and makes it less likely that either CDU-CSU or CDU-CSU-FDP have a majority.

  16. One opinion poll released today has AfD at five percent; Spiegel Online International has more information here.

    Incidentally, while the linked article notes that “fully six parties may receive enough votes on Sunday to send deputies to Berlin,” in reality that number would be seven, as CDU and CSU are separate parties.

  17. The constitutional court ruling was that more than 15 overhang seats for a party would be unacceptably disproportional.

    The new electoral system does not cap the number of “overhang” seats a party may in. Instead, it grants “leveling” seats to the other parties so that the total share of seats is proportional to the number of votes won nationally. Therefore, if there are any states where a party wins more district seats than its proportional share of the total seats, the Bundestag will be enlarged not only by the overhang, but also by additional seats awarded to the opposition to compensate for this. If this system had been in place at the last election, the Bundestag would have over 670 members.

    The link at Electoral Panorama explains how the process works pretty clearly.

  18. Article 38 of Germany’s Basic Law, on Elections to the Bundestag states that “Members of the German Bundestag shall be elected in general, direct, free, equal and secret elections”; and that “Details shall be regulated by a federal law.”

    Thus, it’s clear that PR for Bundestag elections is by no means a constitutional requirement; nevertheless, the Federal Constitutional Court ruled that the 2008 and 2011 electoral systems violated constitutional guarantees of citizens’ rights to take part in direct, free and equal elections, on account of the negative vote weight effect introduced by overhang mandates.

    In fact, Germany could have, say dropped party lists altogether and switched to a FPTP voting system (an option contemplated back in the late 1960s); that would have eliminated as well the problems associated with overhang seats…albeit at the cost of introducing all of the well-known issues associated with plurality voting on single-member districts.

  19. Manuel, what you say does not add up with Chris’s comments… He claims that the constitutional court commented on the system’s proportionality as well, saying what amount of overhang constitutes “too much disproportionality”, implying that there IS a constitutional provision guaranteeing proportionality… However, I can’t find one. Is this a case of judicial activism?

  20. JD, as I understand it PR has become the accepted the generally accepted system for Bundestag elections, even though it’s not constitutionally mandated – just as the appointment of U.S. presidential electors by statewide popular voting has become the generally accepted system in American presidential elections, even though it’s not constitutionally mandated either.

    From that perspective, the Constitutional Court essentially told the politicians that if they were to hold Bundestag elections by PR, they had to come up with a mechanism that treated all parties and voters in a fair and equal manner.

    As for the Court commenting on what constitutes or not too much dis-proportionality, that may be a reflection of the fact that while overhang seats have been a recurring feature of Bundestag elections, they were few and frequent in number (and therefore a non-issue) until fairly recent times.

  21. Pingback: Tweaks to MMP in Germany? | Fruits and Votes

  22. My understanding is that the Federal Constitutional Court has told the Bundestag several times over the years “You don’t have to adopt a proportional system, but if and once you do, it has to be adequately proportional”. So, for example, an exclusion threshold of up to 5% is permissible, but not anything over (as in Turkey, for example). And the system has to ensure that overhang seats don’t distort the result. However, Germany could adopt single-seat plurality (or single-seat runoff, the system used before 1918) and the CC would allow that, provided the districts had approximately equal population/ voters.
    Seems rather odd logic to me, but hey, Karlsruhe has successfully staved off the resurgence of Nazism for 75 years, so I defer to their judgment.

    • I thought that the Federal Constitutional Court’s concern about overhang seats was less about ‘proportionality’ and more about the concept of ‘negative vote weight’: their concern that a party might, due to some property of the way list seats are distributed between Lander and how overhang seats fit into this process (I would encourage you to google this yourself coz i would undoubtedly get it wrong if I tried to explain it), actually lose seats by gaining votes.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.