Erin's Erfurt III Experience

[Alex Richards]

Got to say 'we have two Baden regions and two Württemburg regions and three of them need to agree to restore the old states' seems a bit off.

 

[Erinthecute]

Got to say 'we have two Baden regions and two Württemburg regions and three of them need to agree to restore the old states' seems a bit off.

The whole thing was an exercise in gerrymandering. Both sides knew that Württemberg would support the merger and Baden probably wouldn't, and were just trying to figure out how to slice up the territory to get their preferred option to win.

 

[Erinthecute]

Continuing the theme of federal reform, Mitteldeutschland is a proposed state combining Saxony, Thuringia, and at least part of Saxony-Anhalt. These specific borders were (roughly) pitched by social scientists Meinhard Miegel and Adrian Otthad in 1990 and 1997, respectively. A big theme in reorganisation proposals is centring states around one or more of the officially recognised metropolitan areas, and Mitteldeutschland conforms to the metropolitan region of the same name. It is based around Leipzig and Halle - two closely-linked cities awkwardly separated by a state border - and takes in surrounding cities such as Chemnitz, Zwickau, Dresden, Dessau-Roßlau, Gera, and Jena. With these borders, it would have a population of 7.4 million and an area of 43,700 square kilometres, similar to Lower Saxony.

The proposal is particularly popular in the Saxony-Anhalt, and it's by no means fringe - in 2013, state premier Reiner Haseloff endorsed the idea, much to the chagrin of the other two states' governments. State finance minister Jens Bullerjahn of the SPD also spoke out in favour. The late Bernhard Rothe, an SPD MdL from Halle, long campaigned for a referendum to form the state. In 2015, he gathered 8,000 signatures for a petition for a referendum, although it was dismissed by the interior ministry and a legal challenge was rejected by the Federal Constitutional Court. Mayors of the various cities in the region have also indicated their support. Much of its mainstream popularity is fueled by the financial difficulties encountered by the small states, Saxony-Anhalt and Thuringia, particularly in the face of the debt brake, which could be solved by unifying into a more economically powerful state. It's also viewed as a good way to reduce bureaucracy and the size of government, which makes it popular among various FDP associations. Predictably, it is a lot less popular in the larger Saxony, where politicians tend to insist the other two states should deal with their own problems.

Politically, Mitteldeutschland overlaps almost perfectly with the AfD's heartland. The party won a plurality in Saxony and Thuringia in the last federal election, and they consistently win the southern districts of Saxony-Anhalt compared to its more moderate north. On the other hand, cities like Leipzig and Jena are bastions of the progressive left, in stark contrast to the ultraconservative countryside. I made this scenario to explore the potential dynamics of a united Mitteldeutschland. Backdating its formation to 1990, it would have been easily dominated by the CDU throughout its history, but face a brutal challenge from the AfD from 2015 onwards. The CDU manage to cling onto first place in 2019 by a very narrow margin against an AfD led by none other than Björn Höcke. Despite serious losses, Thomas de Maiziere secures a full term as Minister-President (IOTL, it was often speculated that he would return to Saxony state politics as successor to Stanislaw Tillich.) He is also forced to expand his coalition to include the Greens, though due to the re-entry of the FDP into the Landtag the government's majority is slimmer than they would like.

The map is almost equal parts black and blue. The CDU take 39 constituencies to the AfD's 29, while The Left secure four and the Greens three. The AfD wins primarily rural areas - with some exceptions such as the cities of Gera (constituency 20), Chemnitz south (64), Zwickau (72), and Plauen (75). The CDU win peripheral heartland as well as suburban areas, and sweep most cities thanks to a divided left and weak AfD. An urban shift enables the left-wing to pick up some seats, including the Greens, who record their best-ever performance. They win the city centres of Leipzig and Dresden (38 and 57), and in an upset, Anja Siegesmund wins her constituency of Jena (18). The Left also eke out pluralities in Erfurt south (13) and Chemnitz central (63), while retaining two seats in the south of Leipzig (39 & 40).

 

[Ares96]

So I assume this would entail the northern half of S-A merging with Brandenburg?

 

[Erinthecute]

So I assume this would entail the northern half of S-A merging with Brandenburg?

Yes. The full proposal is for seven states:

 

[vjw]

Yes. The full proposal is for seven states:

Even after the end of the world, Bavaria remains itself

 

[Erinthecute]

The People's State of Hesse is one of two Hesses in Germany, the other being the former Prussian province of Hesse-Nassau. The Volksstaat Hessen is often referred to as Hesse-Darmstadt to distinguish the two. The state was founded in 1919 and in its early years was ruled by a stable coalition under the moderate SPD leader Carl Ulrich, who became so broadly popular that even the former Grand Duke whom he helped depose congratulated him upon his 75th bithday. The retirement of the "red grand duke" and the difficulties of the 1930s saw a decline in the SPD's fortunes, and the Zentrum defected to a centre-right coalition with the DVP and Bauernbund.

Compared to most other states, the party landscape in Hesse-Darmstadt has changed surprisingly little in the last century. The SPD has retained first place in every election, and has governed more often than not, in a variety of coalitions. The fall of the völkisch right and consolidation of urban conservatives and right-wing liberals around the DVP established them as the major opposition and on-and-off coalition leader. The Centre Party has managed to retain a strong position, especially on the left bank of the Rhine. The Bauernbund's strength has declined due to urbanisation and the rise of the DVP, but they are guaranteed influence as a necessary part of any right-wing government. They still dominate the sparse rural areas of Oberhessen. In Hesse, the crisis in the KPD manifested as a rupture between its wings, and the moderate left-socialist faction won out, continuing today as the Hessian Socialists. Together with the Progressive liberals, they are the most common coalition partners for the SPD.

 

[Erinthecute]

Couple more voting precinct maps. First off is the federal election results in Leipzig. It's important to note that the data only maps in-person votes, which made up 61% of votes cast. The main difference from the true total is that the AfD and Greens are over- and underrepresented by about 3 percentage points, respectively. Based on the distribution of postal votes (mostly in the city centre, less in the periphery), basically imagine less AfD on the outskirts - replaced mostly by the CDU - and more Green in the centre.

With that out of the way, analysis time. This is a particularly interesting city, being something of a leftist bastion surrounded by deeply conservative AfD-voting territory. Accordingly, the SPD came out on top in 2021, winning 20.9% overall (and in both constituencies.) The Greens placed second with 18.5%. The CDU, Left, and AfD all recorded 13-14% each, and the FDP won 10%. The party list vote trends are fairly simple: Greens in the centre, transitioning to SPD as you get further away, and AfD in the outlying areas. However, there are notable pockets of Left voters in Connewitz in the south and Volkmarsdorf east of the city centre. The former is a well-known hub for the alternative scene and squatters and has undergone substantial urban renewal in recent decades. Volkmarsdorf, on the other hand, features a number of GDR-era apartments built right at the tail end of the 80s, which fell into disrepair after reunification. The area experienced population decline and low turnout, as well as a strong support for the NPD, though things have begun to turn around in the last decade. The Left and Greens perform strongly in both areas.

Things change significantly in the constituency vote. Leipzig I, in the north, saw CDU representative Jens Lehmann narrowly beat out the SPD candidate. This can likely be attributed to a combination of incumbency advantage, the normal tendency of FDP voters to support CDU candidates, and tactical voting by progressives. This latter factor, ironically, probably cost the SPD the seat. However, Left candidate Nina Treu also noticeably outperformed her party and won strong pluralities in both Lindenau, west of the city centre, and the areas around Volkmarsdorf to the east; these went Green in the party vote.

This may be explained by the buzz around the situation in Leipzig II. The south of Leipzig is noted as young, cosmopolitan, and progressive - good territory for the modern Left party. In 2017, the perfect storm of a declining CDU and strong Left vote gave Sören Pellmann a narrow and unexpected win over incumbent Thomas Feist. He became the first member of his party to win a constituency in Saxony, and the only directly-elected Left MdB outside Berlin in that election. This interesting tidbit became absolutely crucial in 2021 when the party found itself slipping in the polls and, on election night, teetering on the brink of failing to pass the 5% threshold. They needed three constituencies to ensure they would remain in the Bundestag, and after the shock defeat of Petra Pau in Marzahn-Hellersdorf, everything came down to Pellmann in Leipzig II. With the CDU and AfD both on the back foot, his prospects for re-election looked good, and this was confirmed as his margin grew from 0.7% to 4.4%. Compared to 2017, he saw a small swing against him, but this was compensated by support from left-leaning voters of other parties: both the Green and SPD candidates underperformed compared to their party lists. Pellmann's strongest challenge came from Greens candidate Paula Piechotta, who took 18.4%. Geographically, his support extended around the centre of the city through Plagwitz, Südvorstadt, and Reudnitz-Thonberg, though he was unable to overcome the Greens in most of the centre itself, nor penetrate the outlying areas where the CDU and AfD held on.

The 2017 election in Hamburg. This result was interesting because it showcased a historically weak SPD pushed to second place by the CDU, alongside strong performances from the Greens and SPD. I was not expecting the extent of the Left's success in Mitte and inner-city Altona. At the same time, the SPD put on a respectable performance throughout much of the city suburbs; despite losing the party vote overall, they still won five of the state's six constituency seats. The AfD won only a single plurality, but they won it with over 40% of the vote. You can also see where, south of Bergedorf, the city gives way to more rural areas where the CDU was able to run up comfortable margins.

 

[Thande]

Wonderful work @Erinthecute . Have you considered doing individual party heatmaps as well as leading party - not for everything, I mean, but maybe for a big one like Berlin?

 

[Erinthecute]

Wonderful work @Erinthecute . Have you considered doing individual party heatmaps as well as leading party - not for everything, I mean, but maybe for a big one like Berlin?

I've thought about it, but it's time-consuming enough to make one precinct-level map for a city, let alone ones for each party (or even just for a few).

 

[Erinthecute]

Bit late to posting these, but after Elon's lovely attempt to legitimise Russian annexation of the south of Ukraine by quoting the 2012 election map, I decided to give Ukrainian democracy the respect it deserves by making better maps. I intended to do presidential elections as well but the divisions used are different and I don't have the boundaries, plus the 2010 election already has a pretty good map, so I figure it's not really worth it.

The 2006 and 2007 elections used a simple system of party-list PR with a 3% national threshold. The divisions shown are electoral okrugs, which as far as I can tell are pretty arbitrary divisions and didn't serve any purpose in allocating seats or candidates. The current electoral system was adopted from 2012 onwards: parallel voting, with half the Rada elected via party-list PR with a 5% threshold, and the other elected via single-member constituencies.

To make my point: the Party of Regions was a big-tent pro-Russian federalist party. Despite what ignorant tools like Musk believe, the Party of Regions did not advocate separatism from Ukraine or annexation by Russia; while many of its leading figures did have ties to the Russian government, its key electoral appeal was its advocacy for the Russian language, cultural, diplomatic, and economic ties with Russia, decentralising power to the regions, and support for the welfare state. The annexation of Crimea and the war in the Donbas massively damaged pro-Russian politics: the Party of Regions was the most popular party in the country between 2004 and 2014, but its successor groups have scarcely been able to exceed 10%. The contrast between the 2012 and 2014 maps is clear.

 

[Meppo]

Bit late to posting these, but after Elon's lovely attempt to legitimise Russian annexation of the south of Ukraine by quoting the 2012 election map, I decided to give Ukrainian democracy the respect it deserves by making better maps. I intended to do presidential elections as well but the divisions used are different and I don't have the boundaries, plus the 2010 election already has a pretty good map, so I figure it's not really worth it.

bless you Erin, tis a sight for sore eyes

I do have to wonder what the hell was up with that 48th district near Mariupol in 2007

 

[Erinthecute]

I do have to wonder what the hell was up with that 48th district near Mariupol in 2007

It's a really weird anomaly. The SPU surged in Mariupol between 2006 and 2007, going from 18% (already its best area by a wide margin) to 42%, mostly at the expense of the Party of Regions, who fell from 56% to 42%. I have no idea why specifically, but it seems to be the epicentre of a wider trend where the SPU improved in the southeast, albeit marginally, while suffering massive losses elsewhere and falling below the national threshold.

 

[Ares96]

It's a really weird anomaly. The SPU surged in Mariupol between 2006 and 2007, going from 18% (already its best area by a wide margin) to 42%, mostly at the expense of the Party of Regions, who fell from 56% to 42%. I have no idea why specifically, but it seems to be the epicentre of a wider trend where the SPU improved in the southeast, albeit marginally, while suffering massive losses elsewhere and falling below the national threshold.

I smell a UPLB here.

 

[Erinthecute]

I smell a UPLB here.

?

 

[Thande]

?

Unbeatably Popular Local Bloke, a term we use in our coverage of UK local elections to describe a councillor whose majority is completely undented by their party suffering reversals elsewhere (or an independent who always survives). Basically someone with a huge personal vote that's immune to shifts in wider politics.

 

[Alex Richards]

Unbeatably Popular Local Bloke, a term we use in our coverage of UK local elections to describe a councillor whose majority is completely undented by their party suffering reversals elsewhere (or an independent who always survives). Basically someone with a huge personal vote that's immune to shifts in wider politics.

Sometimes capable of getting over 90% of the vote democratically!

 

[Erinthecute]

Unbeatably Popular Local Bloke, a term we use in our coverage of UK local elections to describe a councillor whose majority is completely undented by their party suffering reversals elsewhere (or an independent who always survives). Basically someone with a huge personal vote that's immune to shifts in wider politics.

I see. That would make sense, but bear in mind these were arbitrary statistical divisions, not constituencies, and all members of the Rada were elected from nationwide party lists. The strength of the SPU was most concentrated in district 48, Mariupol north, but it affected the whole city and surrounds. Compare and contrast:

In 2006 their strength came from the central part of the country with an interesting anomaly in Mariupol, but in 2006 they died out almost everywhere while experiencing a huge surge in Mariupol and, to a lesser degree, Odesa oblast. There must have been local factors. I'm completely mystified by Mariupol, but Odesa might be explained by Vasyl Tsushko, a leading SPU member who was head of the oblast government from 2005 until being appointed interior minister at the end of 2006. In May 2007 he suffered a heart attack, possibly as a result of poisoning, and spent the time leading up to the election in hospital in Germany. It seems to have been a rather high-profile thing which could have driven up the party's vote in his home region.

 

[Erinthecute]

Electoral reform has become an increasingly pressing topic in Germany over the last five years. The Bundestag has grown larger in every election since 2002. It was a slow process at first, only a few seats at a time, so it was quite a shock when the number of seats jumped from 631 to 709 in the 2017 election. Last year it grew further to 736. Its standard size is 598 seats. This seems an unusual problem for a country to have. Most legislatures have a fixed size which can only be changed by legislation, or even constitutional amendment in some cases. Germany's unique difficulty arises from a conflict between its electoral system and the constitution.

I'll open with an explainer on the German electoral law and its history, so if you're familiar with the current situation, feel free to scroll down to the bottom.

Mixed-member proportional representation

The Bundestag is elected via mixed-member proportional representation (MMP), a system designed in the post-war era in an attempt to get the best of both first past the post and proportional representation. In the variant implemented for the Bundestag, half the seats are elected via single-member constituencies using FPTP voting, and the remaining half are elected via party lists. Voters have two votes, corresponding to each component of the system. The first vote (Erststimme) is cast for individual candidates for your constituency, and the second vote (Zweitstimme) is cast for a party's state list.

What distinguishes MMP from the parallel voting system used in places like Italy and Russia is the principle of proportionality. MMP is compensatory in nature: the number of constituencies won by each party is taken into account when determining how many PR seats they get, with the goal of ensuring the final makeup of the Bundestag is proportional to the votes cast (specifically, the second votes.) If a party wins many constituencies, they will get few list seats; likewise, if a party wins few constituencies, they will likely get many list seats. An elegant system: it provides the local representation of FPTP without sacrificing the spirit of proportional representation. However, when you take a closer look, cracks start to appear. There are flaws in the system - quite a few, in fact - and they can and do arise in reality. The most prominent, and the one of interest here, is the concept of overhang seats.

Overhang seats

Imagine the following scenario: there is a region with ten seats. Five of them are constituencies, and the remaining five are PR. Party A wins 40% of the second votes, while parties B, C, and D win 20% each. Party A sweeps all five constituencies and thus wins five seats. However, according to the seat allocation, they should only have four, since they only won 40% of the vote. What happens?

For most of the Federal Republic, the answer was nothing. Party A would be allowed to keep the extra seat, which would not be considered during the distribution of PR seats, effectively making the Bundestag a seat larger than it would otherwise have been. These are known as overhang seats (in German Überhangmandate, overhang mandates). This allowed the principle of proportionality to be ever so slightly damaged. Essentially, overhang seats emerge when a party's ability to win pluralities outpaces its popularity as a proportion of the electorate. In the first decades of the Federal Republic, this did not often occur, since the large majority of the electorate backed either the SPD or CDU, and the 50/50 split between FPTP and PR gave a lot of wiggle room during seat distribution. Therefore, overhang seats remained rare: there were no more than half a dozen recorded in each federal election until 1994.

After reunification, however, voting patterns began to diversify. The FDP and Greens continued to grow and the PDS put on an increasingly strong performance in the new states. As a result, there were a record sixteen overhang seats in the 1994 election, and thirteen in 1998. Gerhard Schröder's first government shrank the Bundestag to 598 seats, and as a result of the reduced number of constituencies, there were only five overhangs in 2002. This didn't last, though: the 2005 election saw a fall in the popularity of the major parties and a corresponding rise in overhangs to sixteen. Things got particularly bad in 2009, when the SPD experienced a crushing defeat and minor parties surged, resulting in a new record of 24 overhang seats. An important thing to note is that the calculation of votes and seats is performed separately for each state, preventing overhang seats from being easily compensated by list seats in other states. Essentially, overhang seats could arise independently in different parts of the country, exacerbating the issue.

Enter the Federal Constitutional Court, the surpeme judicial body of Germany. It has broad powers to judge and rule on various issues, including the ability to mandate changes to statute if it finds them incompatible with the Basic Law. The Court has long held that voters have a constitutional right to fair representation, namely in a proportional sense. As a result, non-proportional electoral systems used in some state elections were replaced by variants of MMP or list PR during the first decade of the Federal Republic. The distorting effect of overhang seats on the Bundestag began to raise concern in the 1990s, resulting in a Constitutional Court case in 1995. The ruling was far from decisive: the court tied 4-4 on the constitutionality of overhang seats. As a result, the case was dismissed and nothing changed - at least for the time being.

Leveling seats

A series of rulings in 2008 and 2012 concerning another, far more complicated flaw of the system known as negative vote weight resulted in, essentially, a roundabout condemnation of overhang seats. The Constitutional Court did not rule them unconstitutional per se, but the electoral reform that was eventually implemented focused on ensuring proportionality, and addressed overhang seats in the process. This was a simple fix which had already been adopted across most of the country on the state level: leveling seats. When overhang seats arose, additional PR seats were added to balance them out and maintain proportionality. The electoral law was thus amended to provide for leveling seats at the start of 2013, in time for the September federal election.

The results of the election proved a troubling start for the amended system, however. A landslide CDU victory, divided opposition, and uneven voting patterns meant that the SPD were awarded leveling seats in CDU strongholds and vice versa. Overall, there were just four overhang seats, but they were balanced by 29 leveling seats, boosting the Bundestag to 631 seats. Even worse, applying the new system to the 2009 election results gave an overall size of 648 seats.

A growing Bundestag was an issue for a few reasons. Most pressingly, parliament itself would not be able to cope if it got too big. The institution, and even the building itself, was designed for a legislature of about 600 members. Each of them worked full-time and needed an apartment in Berlin, an office, staff, salary, and all the other logistics that come with the job. An inflated Bundestag could lead to gridlock and serious issues with the political and legislative process. More salient in most people's minds was the cost: more members naturally meant more cost to the taxpayer.

Political developments in the 2013-17 term sparked fears of an explosion in the number of MdBs. The Union and SPD were both deeply unpopular while minor parties were riding high, and with the rise of the AfD, it seemed parliament would be home to six parties (seven if counting the CDU and CSU separately) for the first time since the 50s. In the 2017 election, the Bundestag indeed blew out to 709 seats - comprising a mind-boggling 46 overhangs, all but three of which were won by the CDU or CSU. These were balanced out by 65 leveling seats.

2017-2021: The debate

The growth of the Bundestag was now an urgent problem and there was major pressure, including from then-President of the Bundestag Wolfgang Schäuble, for the government to implement a reform to address it. However, there was and still is no simple fix. The issue lies with the fundamental conflict between the two components of the MMP system, and is exacerbated by the continued diversification of the German political landscape, wherein the major parties win pluralities with smaller and smaller percentages of the vote.

There are several different ideas for electoral reform. Of these, only one suggests sacrificing the principle of proportionality. This position, held by the CDU and CSU, says that leveling seats should be abolished and overhang seats simply allowed to stand, as they were before 2013. Not coincidentally, the Union would be the primary beneficiaries in this scenario - especially the CSU, whose strength in Bavaria leaves them dramatically overrepresented, requiring dozens of leveling seats in that state alone. The most popular school of thought, if mainly for its simplicity, is to simply reduce the number of constituencies. The fewer constituencies there are, the fewer overhang seats, and thus fewer leveling seats. Easy. This is supported by all parties except the Union, though it's not necessarily everyone's most preferred option. It doesn't fundamentally resolve the issue and is still likely to result in a Bundestag with over 600 members, even if it curbs growth to some degree. If only a small number of constituencies are removed, the benefits will be marginal. Likewise, it has proven difficult to get through parliament, since many Union and SPD MdBs are reluctant to mess with the number of constituencies for fear that they will be pushed out by the redistricting.

After much badgering from the Bundestag, media, and experts, a limited reform was passed by the CDU/CSU-SPD government at the end of 2020. This was a compromise between the aforementioned two schools of thought. They agreed to reduce the number of constituencies from 299 to 280, but there wasn't enough time to do it before the next year's election, so it was to be done in the next term instead. A more immediate change was a clause to forgive up to three overhang seats. This allowed a small degree of disproportionality for the sake of keeping growth down - and given that each overhang seat could bring with it three or four leveling seats, it wasn't nothing. Still, this was a clear concession to the Union on the SPD's part, and an underwhelming reform. In the 2021 election, despite the number of overhang seats declining to 34, the number of leveling seats ballooned to 104, and the Bundestag grew to yet another record size: 736 members. The three forgiven overhangs all went to the CSU.

2021-present: Further debate

The defeat of the Union in 2021 allowed the government and Bundestag to approach the size issue from a fresh perspective. Here we can introduce some more radical schools of thought. The next most prominent calls for the elimination of both leveling and overhang seats, which has been proposed by the Greens, Left, and FDP. In this scenario, overhang constituencies would not be awarded, and remain vacant for the duration of the term. It's a messy solution, and the Union loudly opposes the idea of depriving areas of their local representative. Still, the proportional-party aspect of the Bundestag has been predominant for many decades, both in the minds of voters and politicians. It's doubtful many people would lose sleep over it. However, the SPD are also unhappy with the idea. Instead, the new traffic light government has turned to a fourth concept: deeper reform to the fundamentals of the system to fix it at the roots.

Now we get to the point of this post. As laid out in this Vorwärts article from May, the SPD, Greens, and FDP have tentatively agreed to what they describe as an "extension" of MMP. This reform would eliminate both overhang and leveling seats, stabilising the Bundestag at 598 seats. If a party wins overhang seats in a state, the constituencies they won with the lowest percentage of the vote would automatically be forfeited until all overhangs are corrected. This eliminates the need for leveling seats. However, every constituency would still be guaranteed a local representative by awarding the seat to the next most popular candidate instead. To ensure everyone's vote still counts, a "spare vote" (Ersatzstimme) would be introduced - a second preference which voters could put down for another constituency candidate. In overhang seats, the spare votes of the forfeited candidate's voters kick in and are added to the other candidates' tallies when determining the winner.

Conceptually, constituencies are reimagined as a tool to allow voters to influence the personnel within the Bundestag without distorting its partisan composition. The idea is that constituency seats should only be awarded if they're accounted for by the party's overall strength in the state - i.e. their list seats. If they can't, the next party who can is given the win instead. The government's spokesmen for electoral reform argue this makes the system more consistent, alleviating the conflict between the different components of MMP. To reinforce the concept, the "first vote" and "second vote" monikers would be replaced by "personal vote" and "list vote" respectively. As far as I know, the plan to reduce the number of constituencies to 280 has not been changed.

This is an interesting proposal. It's messy at first glance, but it elegantly accomplishes the formiddable task of fixing Bundestag inflation without resorting to ineffective or ugly bandaid fixes or a total overhaul of the system. As someone (albeit a foreigner) who thinks about this a lot, I think it's a good and fair augmentation of the existing system which I would be happy to see implemented. To give an idea of how it would look in practice, I modeled the results of the 2021 election. In this election, the CDU won 12 overhangs (all in Baden-Württemberg), the CSU 11, the SPD ten (across several different states), and the AfD one (in Saxony). With the Ersatzwahl, 14 of them would go to the SPD, 11 to the Greens, 6 to the CDU, and 3 to the AfD. This is a good visualisation of how it works in a diverse electoral environment. On the following map, reallocated constituencies are indicated with asterisks.

I calculated the spare vote results in each constituency using a fairly simplistic model based on party and geography. I imagine there would be a large number of voters choosing not to utilise the spare vote and thus their votes would exhaust, which I accounted for under the "Ex." column. You can see the full table below. The results on the left are after reallocation of votes. This didn't actually affect the outcome in any constituency, but it messed with the margins and order a bit. I used a pretty scattershot model, which is hopefully realistic. Finally, the performance for the winning candidate is given as a percentage of the overall valid votes, including exhausted ballots.

You may notice a peculiar result in Rottal-Inn. The Free Voters actually placed second in this constituency with 17%, and according to the model they would win it - however, since they didn't earn any PR seats, I didn't award it to them; instead it went to the third-placed AfD. I figure this is how the system would work if implemented, though it certainly would be interesting if the Free Voters were able to earn a seat this way.

 

[Alex Richards]

Oh god. It's the SV system for the metromayors awkardly smashed into MMP.