The impact of M=96 and no legal threshold

The decision of the German Constitutional Court to invalidate the legal threshold for election of MEPs has been predictably consequential. Given the single 96-seat district, a very large number of parties has won at least one seat, and some have won with less than 1% of the vote.

There will be thirteen parties (counting the CDU and CSU separately) in the German delegation. Seven of them had less than the former 3% threshold; the biggest of the sub-3% parties had not even quite 1.5%. The German government reports the votes; seats are shown at Wikipedia.

Assuming the Wikipedia list is accurate (and it looks likely to be so), these parties that won representation thanks to the Court ruling are: Free Voters, Pirates, Human Environment Animal Protection, National Democrats (yes, a German neo-Nazi will be in the European Parliament), Family Party, Ecological Democrats, and some outfit called Die PARTEI. The last three of these have vote totals ranging from 0.69% down to 0.63%. The NPD’s vote percentage was 1.03.

Also noteworthy is that the Free Democrats continued their slide, winning only 3.36%. They had just missed the 5% threshold for the federal Bundestag elections last year. The Alternative for Germany (AfD), which also had just missed the threshold for the Bundetag won 7.04%.

29 thoughts on “The impact of M=96 and no legal threshold

  1. Germany’s Federal Returning Officer’s website does have the country’s EP seat distribution as well, in both German and English (in fact, the voting results are available in both languages as well). At any rate, the seat figures published in Wikipedia appear to be accurate.

    As for the consequences of doing away with the threshold, I can’t help but comment that the resulting distribution of seats is truly worthy of the Weimar era. From that perspective, it wouldn’t surprise me at all if Germany adopts British/French-style regional constituencies for the next European election.

    • Yes, Germany should consider adopting regional constituencies for the next European election, but British/French-style Euro constituencies are not the only basis for that consideration. Scotland has, with no threshold, limited undue fragmentation for four elections: 1999, 2003, 2007 and 2011, with 16-MP regions. In actual operation they amount to a 5% regional threshold, but not a national one. In 1999 the Scottish Greens got only 3.59%, but elected one MSP in their best region, a foot in the door which led to a 6.9% vote in 2003, and seven MSPs. These are certainly more representative models than those of Ireland’s EUs and Poland’s EUs. But they are even better than France (average DM of 10) and the UK (average DM of 6), similar to Italy’s Euro DM of 15.6. Given the Germany inclination to as much proportionality as reasonably achievable (is that the AMPARA principle?), they would likely choose no more than six regions (DM 16) if they can make the sixteen Länder fit into six.

      • Except that Italy’s Euro DM is 14.6. (Why won’t this blog let me edit?) However, I still think the Germans will want to do better than Italy.

      • I think this would be satisfactory:
        Baden-Wurttemburg 13
        Bavaria 15
        East Germany 19
        North Rhine-Westphalia 21
        Saarland+Rhineland-Pfalz+Hesse 13
        Lower Saxony+Schleswig-Holstein+Hanseatic cities 15

      • If Germany would switch to districted PR, the scheme suggsted by jdmussel is probably the only reasonable one. If I recalculate the result, these six district would give the following seat shares (Sainte-Lague):

        CDU 30(+1), SPD 27, Grüne 11, Die Linke 8(+1), AfD 7, CSU 7(+2), FDP 4(+1), Freie Wähler 1 and NPD 1 (with Piraten, Tierschutzp., Familie, ÖDP and Die PARTEI losing their only seat)

        The effective tresholds would be: NW 4,49% (last seat to SPD), East 5,28% (NDP), BY 6,12% (FDP), North 6,54% (CDU), BW 7,25% (Die Linke) and HE+RP+SL 7,43% (FDP)… too high for the BVerG?

  2. Die PARTEI is a satirical party, like Sweden’s Donald Duck and Britain’s Monster Raving Loony. Its programme includes rebuilding the Berlin wall and the Iron curtain, war on Liechtenstein and changing the first article of the German constitution (“Human dignity is inviolable”) to exclude certain television CEOs… The party’s leader has already announced he will resign from the European Parliament after a month, to let the next one on the list have a go, and that the latter would do the same so that all the candidates have a go as MEPs,

  3. Things would be slightly less absurd if they had used D’Hondt. It cuts off seats at the NDP and gives the leftovers to the CDU (+2) and SPD (+1).

    • If Germans were any good at math they would recall that the Weimar deflation, not the hyperinflation, bought a certain regime to power and moderate the demands they make of countries in the EU southern tier. They might even recall that it was the bizarre powers of the Weimar presidency which opened the door to that regime and wonder why they concentrated on the electoral system which was essentially irrelevant to the seizure of power.

  4. Am I the only one who agrees with the court and doesn’t have a problem with not having a threshold? If people want outliers, microparties, amnd joke parties to represent them in a body that doesn’t need to be stable, I am all for it.

    • The electoral system has too much impact on the outcome for ‘what people want’ to be a truly meaningful criterion. The fragmentation-inducing effects of having no threshold prevent a united opposition in South Africa, but to say that that is really what anti-ANC voters want would be silly.

    • I agree with you Mark. I don’t see any moral imperative to deny smaller parties a right to a seat in parliament by setting a threshold. Eliminating a threshold helps to reduce the number of wasted votes, and I do believe that increasing the number of effective votes is a legitimate and worthy goal in creating better electoral systems. PR is supposed to hold up a mirror to the electorate and that includes the absurd (die partei) to the revolting (National Democrats neo-nazis) to the legitimate. The only question that I think that relevant is which type of proportional method is the most appropriate in seat allocation calculations. I have no issue with choosing a proportional system that favors larger parties, particularly, D’Hondt or Droop, as opposed to those that favor smaller parties.

      • There was a lengthy discussion on the merits of thresholds last year:

        It’s also worth noting that the high court’s ruling applied only to European elections and not Bundestag elections, the logic being that as the Bundestag was a governing body, the thresholds a valid democratic implement to ensure government stability. It ruled that the European Parliament was not such a body (though I believe the Greeks would strenuously object to that assessment at present), and as such, the threshold served no valid purpose.

        In my view, once you leave direct democracy behind for some form of representative democracy, you are by default introducing disproportionality to the system to some extent. At this point, it is simply a question of degrees. A system which gives a party with 0.66667% of the vote a seat in parliament while denying a seat to a party with 0.66666% (the Dutch system) still has a threshold and it’s hard to see how it is therefore substantially different than a system with a threshold of 3% or 5%. If each legislator received a nun number of votes within the body equal to the number of votes he received, it might be a somewhat different story, but I know of no system which does so.

        Therefore, once you’ve decided to establish a threshold of some sort, it makes sense to me to establish a threshold that promotes good governance rather than merely using 1/m or 0.5/m, where m is the entire assembly and is always some arbitrary number less than the number of qualified electors.

        I think once you get to the Turkish level which has at times effectively disenfranchised 45% of those who’ve cast ballots, it may be crossing the line into undemocratic, but I don’t believe that the Greek elections of May 2012 were undemocratic simply because 19% of the voters chose to cast votes for parties which had little chance of passing the 3% threshold (a fact that I’d assume most of the voters were aware of when they cast their ballots).

      • On the other hand, the last German election featured the FDP unexpectedly getting only 4.8%, and the new AfD unexpectedly getting only 4.7%. Polls has consistently shown the FDP above 5%, and the AfD at 5% or between 4% and 5%. So clearly 9.5% of voters got an undemocratic result, not approaching Turkish levels, but beyond normal European tolerances. Germany has been struggling for decades with the arguable need to lower the threshold to the 4% used by their neighbours Austria, Sweden and Norway (Denmark being even lower at 2%). I think the time has come.

      • Dominic: I don’t see why ‘holding a mirror up to the electorate’ should be such an important end. The end, I think, should be along the lines of good government (open to your interpretation, indeed, but more useful as a goal whatever the interpretation), and the question should be what do we do to achieve that. I personally don’t believe extreme proportionality contributes much to good government – there have been occasions where it has even proven a problem, either in terms of stability or intraparty/regional representation. Indeed, for the European Parliament only the second issue is truly relevant. Therefore, I think that a set of constituencies, combined with an open list system, might serve to better represent Germans in the EP in terms of oversight over and accountability of MEPs.

        Wilfred Day: A 4% threshold can bring the same problem, as shown in Bulgaria’s last election where some 25% got no representation because they voted for sub-threshold parties. I think a better solution, for Germany as well as Bulgaria, would be to make the ballot preferential, so that votes for AfD and FDP would have gone to their second preference. This eliminates complete lack of representation for those voters, as well as the vicious cycle associated with the threshold and strategic voting (people don’t vote for a party because it won’t reach the threshold because people don’t vote for it). In fact, I think having used such a system last year would have brought both parties into parliament as a result of people voting their true preference instead of CDU.

      • JD, I completely agree with your points, especially with regards to permitting multiple preferences with transfers for parties below the threshold. I’d actually argue that the system which comes closest to this, the New South Wales upper house system, is the world’s best electoral system in present use for its combination of governability and fair representation (the quota is 4.55% with the posibility of transfers).

        As far as the German need to reduce the Bundestag threshold, I don’t think that need was at all apparent prior to the previous election. As far as memory serves me, there was only one situation at the federal level which a party was near even 4% and failed to cross the threshold (the PDS/Linke in 1998, I want to say). Perhaps that need is now apparent, as a larger share of voters than normal were “disenfranchised by the threshold” (is there a more precise term for this in comparative psephology?), but as far as governability is concerned, I’m not sure that “reducing the need for a grand coalition” is really a reason to change the electoral process.

  5. Either a threshold or a medium to small multi member district magnitude reduces the number of parties. The Dutch seem to be the only Europeans not to bother with an electoral threshold and Dutch Democracy is one of the world’s best. South Africa has no threshold and it is perhaps a good thing or else the ANC would get a solid seat bonus to a two thirds majority. It is a mystery of what makes a democracy succeed.

      • Are Scandinavian countries democracy better than the Netherlands? Why would you disagree?

      • I disagree for two main reasons:
        1) the lack of electoral efficiency and the resulting lack of accountability and
        2) the lack of local representation and the resulting low quality of representation

        Part 1 is about the interparty dimension, part 2 is about the intraparty aspect. The first part is indeed more a result of PR than of the specific brand of PR used, but my views on PR are not exactly mainstream (among this blog’s commentators). Besides, political fragmentation in the Netherlands is on the increase, so the relevance of the specific brand is increasing.

        As to Scandinavian countries, I certainly think they have a better system on the intraparty dimension, while the interparty dimension is not as problematic due to the formation of pre-election coalitions, something the Netherlands has not had in a long time.

    • The Cypriots also don’t use a threshold, though the m=56 assembly results in an effective threshold of around 1.8%.

      Also, under both the Dutch and Cypriot systems, a party must earn at least one full quota (1/m) to receive a seat, even if it otherwise would win a seat by largest remainder. This results in an effective threshold in the Netherlands of 0.67%.

  6. The man who tried to implement proportional representation in Ontario, John Gerretsen, was a Dutchman who moved to Canada with his parents when he was 12, but was very well aware of the differences between the Dutch and German systems. He always said he supported PR “the German system, not the Dutch.” And he, like most PR advocates for Ontario, wanted a regional model (three German states use regional models). Unusually, for fans of the German model, he wanted regions with DM as small as 11, maybe average 13, mostly because of Ontario geography, but also because he was no fan of Dutch extreme proportionality, and he correctly thought most Ontario voters would want all Members of Provincial Parliament to be accountable to a defined group of voters. Too bad his brainchild, the Ontario Citizens Assembly, didn’t design the model he wanted.

    Does the European Parliament not need MEPs who are accountable, and parties that are accountable? Surely the recent result shows the very opposite.

    In fact, maybe Germany will not only move to a six-region model for their MEPs, but switch to mixed-member proportional? (Likely a 50/50 model.)

    • I was only aware of Bavaria using regional MMP – what are the others?

      I am very sceptical of the possibility of introducing accountability to the European Parliament, and hardly think MMP would make a difference. Constituencies would have to contain more than a million voters each. Besides, German electoral-system designers seem unwilling (or unaware of the possibility) to not compensate for overhangs (as in Britain), and Germany certainly can’t have more than 96 MEPs.

  7. The other two are Baden-Wurttemberg (four regions) and Rheinland-Pflaz (an unusual optional regions model). As for MMP without overhangs, the first Bavarian electoral law actually provided for that: a party winning too many local seats would forfeit as many local deputies as necessary. It never happened, and then they changed that point, so this is a very little-known fact.

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.