User:Closed Limelike Curves/Negative voting weight

Negative voting weight (also inverse success value ) refers to an effect in elections in which voters' votes act against the will of the voters; that is, either votes for a party that mean a loss of parliamentary seats for that party, or votes that are not cast for a party and bring that party more seats. It contradicts the principle of one vote, one value, according to which every vote should count equally, and violates the requirement that the vote must not explicitly go against the will of the voters.

In Germany, the Federal Constitutional Court has ruled that negative vote weights are incompatible with the constitutional guarantee of equal and direct suffrage.^[1]

Mixed Member Proportional Representation

In federal elections up to and including 2009, the first step was to determine how many seats a party was entitled to nationwide based on the distribution of second votes. In the second step, these mandates are distributed among the states depending on the second vote results in the individual federal states. Finally, in each country it is checked individually how many mandates the party has already received through constituency seats. The remaining mandates to which the party is entitled in this federal state are allocated based on the party's respective state list.

This can lead to overhang mandates: If a party has received more constituency seats in a federal state than it is entitled to according to the second vote result, it still retains all constituency seats. Seats exceed the number of seats to which the party is entitled based on the party vote are called overhang seats. Previously, there were no possible [[compensatory mandates for the other parties.

The following scenario can lead to a negative voting weight: Suppose a party receives an overhang mandate in federal state A. If it were to win additional second votes in this federal state, this could mean that the total number of seats to which it is entitled based on the second vote result would remain unchanged, but their distribution between the federal states would change. The additional second votes can result in the party receiving one more mandate in federal state A based on the second vote result, but one less mandate in another federal state B. However, the additional mandate gained in State A would eliminate the overhang mandate, so that the party from State A would not be able to send any additional representatives to the Bundestag. In federal state B, however, it would receive one fewer mandate, provided it does not receive any overhang mandates there. Despite receiving more second votes, the party would have lost one mandate overall.

Federal Constitutional Court ruling 2008

There have been no major efforts by legislators to address this phenomenon. Following a corresponding lawsuit, the Federal Constitutional Court declared the existing electoral law to be unconstitutional:

An electoral system that is designed to or at least typically allows for an increase in votes to lead to a loss of seats, or for a party to achieve more seats overall if it receives fewer votes itself or if a competing proposal receives more votes, leads to arbitrary results and makes the democratic competition for support among the electorate appear absurd.^[2]

The legislature was instructed to amend the electoral law by 30 June 2011 so that this effect would no longer be possible in the future and the principles of electoral law would no longer be violated. On 30 September 2011, well after the deadline set by the Federal Constitutional Court had passed, the Bundestag passed an electoral reform with the votes of the CDU, CSU and FDP, which was intended to largely eliminate the negative voting weight in elections. ^[1] ^[3]

Since the SPD and the Greens felt that the new regulation had not achieved this goal, the opposition filed a constitutional complaint. The Mehr Demokratie also announced a lawsuit against the electoral reform and called on interested parties to participate. By December 12, 2011, almost 3,000 complaints collected in this way had been submitted to the Federal Constitutional Court. The complainants were of the opinion that the effect of the negative voting weight could occur in a constitutionally unacceptable manner even under the amended federal electoral law.^[4]

Federal Constitutional Court ruling 2012

Together with the organ dispute proceedings and the judicial review proceedings, the Federal Constitutional Court heard the constitutional complaint orally on 5 June 2012.^[5] On 25 July 2012, the court confirmed the concerns of the plaintiffs and declared the new regulation passed by the CDU/CSU and FDP unconstitutional.^[6]^[7] The Federal Election Law, which was amended in 2011, "violated the principles of equal and direct suffrage, as well as equal opportunities for the parties [...] because it enables negative vote weights."^[8]

Other occurrences

Some countries that hold referenda require a minimum level of turnout for the referendum to be binding. In such a case, votes against the proposal can lead to the threshold being exceeded and the proposal adopted. To the extent that quorums are considered useful in referendums at all, they are now increasingly being implemented via an approval quorum—i.e. a minimum number of favorable votes—in order to avoid negative vote weights.

Negative vote weights occasionally arise in the context of apportionment by Hamilton's method. Systems with compensatory mandates, which are awarded in some state elections, are particularly susceptible to negative voting weights. Their cause usually lies in the Hare/Niemeyer-specific Alabama paradox.

It also appears in runoff elections and instant runoff voting.

Generalization

In 2012, the Federal Constitutional Court used the term negative voting weight in a more general sense. Accordingly, negative voting weight also exists if the number of seats of a party correlates unexpectedly with the number of votes cast for a competing party. ^[9]

According to the Federal Constitutional Court, a seat allocation procedure that allows a party's nomination to win more seats overall if a competing nomination receives more votes contradicts the spirit and purpose of a democratic election. The votes work against the will of the voters.

General example

Description

In a federal election with 5,980,000 valid votes, a party P1 receives a total of 250,000 second votes, of which 106,000 are in federal state A and 144,000 are in federal state B (A and B are the only federal states). In country A, the party achieved 11 constituency seats through first-past-the-post votes, in country B 6, for a total of 17.

Due to the statutory total number of 598 seats in the German Bundestag according to § 1 BWahlG results in 25 seats for P1 (598 seats × 250,000 votes ÷ 5,980,000 votes = 25 seats), of which 11 seats (ideal claim: 10.60) are for State A and 14 seats (14.40) for State B. In State A, all seats to which the party is entitled are already occupied by the constituency seats. In State B, the party has only won 6 constituency seats; the difference of 8 seats will be filled by candidates from State List B. In the final result, the party receives 25 seats.

Assuming that P1 received 5,000 fewer second votes in country A (101,000 second votes and 5,975,000 votes in total) with an otherwise equal number of votes, the number of 245,000 second votes obtained also results in a claim to 25 seats in this case (ideal claim: 24.52). However, calculated separately by country, there are only 10 seats (10.11) for country A + 14 seats (14.41) for country B, i.e. a total of 24 seats. The difference of one seat would be made up by an additional 15th. Seat for State B equalized, occupied by a candidate from State List B. In addition, the party would receive an overhang mandate in State A because, regardless of the distribution of second votes, 11 candidates received a constituency seat. In the final result, the party receives 11 + 15 = 26 seats.

With 5,000 fewer second votes, P1 would therefore be represented with 26 instead of 25 seats in the Bundestag. There is a disproportion of 5.77 % of the ratio of the number of votes to the number of mandates: while in the case of 25 seats 10,000 second votes were required for each seat, in the other case only 9,423.1 (disproportion: 1 - (245,000 / 26) / (250,000 / 25) = 5.77 %).

(For the sake of simplicity, this example ignores the provision of the Federal Election Law that a party that receives the absolute majority of second votes automatically also receives the majority of seats.)

Change options

Since the negative voting weight described above can occur independently of the seat allocation procedure, the change in the Federal Election Law in March 2008 from the Hare/Niemeyer procedure to the Sainte-Laguë/Schepers procedure is not an improvement. Even compensatory mandates do not solve the problem, since at least one of the affected parties regularly does not receive compensatory mandates.

The negative voting weight can be avoided if the creation of internal overhang mandates is prevented. This could be achieved through four different strategies without fundamentally changing the current federal electoral law:

Internal overhang mandates could be prevented by offsetting direct and list mandates at the federal level, which would mean that other federal states would have fewer list mandates. If a constituency winner from a surplus country is eliminated, a list mandate in the other country would be revived. Such models have existed for some time. In contrast to the currently valid electoral law, majorities based on internal overhang mandates cannot change or even reverse within an electoral period due to the effect described above.
Overhang mandates could be prevented by eliminating surplus constituency seats. Such a rule applied in Bavaria in the state elections from 1954 to 1962 inclusive: if a party received more constituency seats than it was entitled to seats based on the share of the vote, the direct candidates with the smallest number of votes received no seat. Alternatively, the regulation could be designed in such a way that instead of the direct candidates with the smallest number of votes, those with the smallest share of the vote are eliminated.
The abolition of state lists and the introduction of federal lists would eliminate internal overhang mandates.
As in the federal elections of 1949 and 1953, the mandates could be distributed proportionally among the federal states, with each federal state having a fixed number of seats (without overhang mandates). However, this would lead to greater disproportionate effects, especially in favour of parties that are strong where voter turnout is below average; the Left Party would most likely benefit from this. It would then be much more likely than before, for example, that one party would get more seats than another party despite receiving fewer votes. ^[10]

In 1996, the parliamentary group of Alliance 90/The Greens presented a draft law which, among other things, a. implements the accounting model described above under No. 1, in order to exclude a negative voting weight:

This was intended to neutralize internal overhang mandates and thus prevent negative voting weights. The Bundestag rejected the draft with the votes of the CDU/CSU, SPD and FDP parliamentary groups and relied on a reduction in the number of seats to 598 and a new constituency structure. The draft law also provided for the elimination of the possibility of excluding the list alliance, meaning that internal overhang mandates would no longer have been possible.

Legal policy discussion

The phenomenon of negative voting weight has been rather neglected in public discussion. Before the 2002 federal election, the news magazines Der Spiegel and Focus ^[11] reported on the issue of negative voting weight.

In the case of success value distortions, there is often a lack of a clear causal relationship between overhang mandates and negative voting weights, as they often interact cumulatively. In 1996, a reform commission of the Bundestag looked into electoral problems - including overhang mandates - and sought the opinion of experts such as Professors Ernst Gottfried Mahrenholz, Wolfgang Löwer and Markus Heintzen . In this context, the Federal Ministry of the Interior developed, among other things, the so-called Compensation Model I, which provides for an offsetting similar to the above-mentioned modification option 1 .

Many of the mathematicians and legal scholars involved in this field strongly criticize the negative voting weight and consider it a defect that is incompatible with the requirement of equality, freedom and immediacy (transparency) of an election:

In their opinion, the equality of the election is violated because the success value of a vote is lower – namely negative – than the success value if one had not cast a vote.
Freedom of choice is violated because the voter is no longer free in his or her voting decision if his or her vote could harm the party he or she wants to vote for. This could unsettle a voter and prevent them from voting for their party.
Finally, they do not consider the immediacy of the election to be given, since the defect in the necessarily intervening mathematical calculation procedure means that the votes for a party can no longer be counted in its favor, but to its detriment. The will of the voters is no longer directly converted into mandates for a party, but is distorted. A voter may not vote for his party in order to express his support for it.

A negative voting weight can also occur in many proportional representation systems. However, the uncertainty is usually limited to a maximum of one mandate per party. In addition, most other procedures are less vulnerable than the federal electoral law.

Legal action against norms

Abstract review of norms 1995/96

In 1995, the government of Lower Saxony had parts of the Federal Election Law reviewed by the Federal Constitutional Court and explicitly argued about the effect of negative voting weights. In their opinion, overhang mandates and negative voting weights create a performance value distortion in cumulative causality that is so contrary to equality that § 6 and § 7 BWahlG are essentially unconstitutional and void. She referred to the case law of the Court and the character of the Bundestag as a unitary parliament, which is why at least three possible amendments were available that would meet the equality requirements of Art. 38 GG . She underlined the possibility of amendment mentioned above 1 . The outcome of the hearing of the Federal Returning Officer in this matter supported the critical findings in Lower Saxony’s application.

The court confirmed the Federal Election Law by a tie-break decision with the votes of judges Jentsch, Kirchhof, Kruis and Winter. They pointed out that the legislature has freedom of design and that this could lead to inherent distortions in the value of success. However, the standards established in case law to date are not even remotely conclusive, and it is possible that other inequalities may even be permitted.

In the opinion of Judges Graßhof, Hassemer, Limbach and Sommer, the electoral system within the scope of the judicial review application is unconstitutional and violates the principle of equal voting rights. The legislature certainly has some leeway for arrangements deemed necessary, but only within the framework of strict equality of choice: "Necessity alone does not constitute justification." They point out that the above-mentioned bill by Alliance 90/The Greens completely solves the problems caused by overhang mandates. Due to the requirement of judicial self-restraint, the Court does not have to dictate which legislative measures Parliament should take. However, the legislature must take one of these measures in order to comply with the constitution.

Legal challenges 2011

In response to the 2005 election complaint, the German Bundestag passed a reform of the Bundestag electoral law in September 2011 with the votes of the CDU/CSU and FDP. It was the first time in the history of the Federal Republic that the Bundestag electoral law was decided solely by the votes of the governing coalition and not by consensus of all parties. ^[12] (The later electoral reforms of 2020 and 2023 were also passed only with the votes of the respective government factions.) The electoral reform was discussed extremely controversially in advance, since, in the opinion of the opposition, it did not fundamentally eliminate the abuses criticized by the Federal Constitutional Court, but rather continued them and unnecessarily complicated the electoral law. Both the SPD ^[13], the Greens and the Left Party felt that their rights had been violated and filed an action for judicial review .

Election audits due to negative voting weight

Examinations by the German Bundestag

Election objections have been regularly lodged with the German Bundestag on the grounds of negative voting weight, most recently in the 1998 federal elections, 2002, and 2005. The Bundestag always decided – as prepared by the Electoral Committee – to reject the objections because the provisions of the Federal Election Law were complied with and the decision on the unconstitutionality of provisions of the Federal Election Law is left to the Federal Constitutional Court.

Constitutional Court review of the 1998 federal election

In 2001, the Federal Constitutional Court rejected two electoral complaints citing, among other things, the negative voting weight^[14]. However, reasons for this were not given either in these a limine decisions or in the rapporteur’s letters.^[15]

Constitutional Court review of the 2002 federal election

An electoral review complaint regarding the aforementioned objection to the 2002 federal election was pending before the Federal Constitutional Court for an exceptionally long time. In this case, ^[16] in 2004, the then rapporteur of the court, Judge Jentsch, casually described the possible emergence of negative voting weights as still constitutionally acceptable. This view is not supported per curiam in this form. Only after the decision of 3 July 2008 in the electoral review proceedings for the following federal election (2005) did the Federal Constitutional Court decide with reference to the judgment.^[17]

Constitutional Court review of the 2005 federal election

Three further complaints were filed during the electoral review of the 2005 federal election. ^[18] The Federal Constitutional Court heard two of these cases on 16 June. April 2008. At 3. In July 2008, the second senate announced its verdict: According to the Karlsruhe judges, the “negative voting weight” is not compatible with the principle of equality and directness of the election. This was the first time that the Federal Constitutional Court had declared a provision of the Federal Election Act to be unconstitutional in an election review procedure, thus creating the need for new legislation.

Appearance at federal elections

In the history of federal elections, the occurrence of negative voting weight has been proven in the federal elections of 1990, 1994 and 2002 (see web links). There are also other examples:

In 1961, the CDU Schleswig-Holstein would have received one more mandate with 39,671 fewer votes. In the same year, the CDU Saarland would have received one more mandate with 48,902 fewer votes.
In 1983, the SPD Bremen would have received one more mandate with 73,622 fewer votes. Likewise, the Hamburg SPD would have received one more mandate if it had received 73,569 fewer votes.
In 1987, the CDU in Baden-Württemberg would have received one more mandate if it had received 18,705 fewer votes.
The SPD would have received one more mandate in 1990 if it had received 8,000 fewer votes in Bremen. Likewise, if the CDU had received 2,600 fewer votes in Thuringia, it would have received one more mandate.
In 1994, the CDU would have received one more mandate if it had received 19,089 fewer votes in Baden-Württemberg, Mecklenburg-Western Pomerania, Saxony or Saxony-Anhalt or 13,629 fewer votes in Thuringia.
The SPD would have received one more mandate in 1998 if it had received 70,955 fewer votes in Brandenburg, or 21,323 votes in Saxony-Anhalt, or 21,228 votes in Thuringia and at the same time 1,000 fewer votes in Brandenburg.

Furthermore, there are several cases in which a party would have received fewer seats if it had received more votes. This applies to:

the CDU Schleswig-Holstein 1957 (88,833 votes more, then two mandates less),
the CDU Saarland 1961 (10,828 more, then one mandate less),
the SPD Schleswig-Holstein 1980 (7,809 votes more, then one mandate less),
the SPD Bremen 1983 (4,083 votes more, then one mandate less),
the SPD Hamburg 1983 (8,199 votes more, then one mandate less),
the CDU Mecklenburg-Vorpommern in 1990 (13,545 votes more and at the same time in Thuringia 1,000 votes more, then one mandate less),
the CDU Saxony-Anhalt in 1990 (6,314 votes more and at the same time in Thuringia 1,000 votes more, then one mandate less),
the CDU Thuringia 1990 (66,693 votes more, then one mandate less),
the SPD Bremen 1994 (1,042 votes more, then one mandate less),
the SPD Brandenburg 1994 (73,403 votes more, then one mandate less),
the SPD Hamburg 1998 (16,651 votes more, then one mandate less),
the SPD Mecklenburg-Vorpommern in 1998 (6,628 votes more, then one mandate less) and
the SPD Brandenburg 1998 (4,015 votes more, then one mandate less)

In the 2002 federal election, the SPD lost a seat in Brandenburg because it received 50,000 too many second votes; the seat would otherwise have gone to the Bremen SPD state list candidate Cornelia Wiedemeyer.