This article was originally published in German language on August 17, 2011 at http://liquidfeedback.org/2011/08/17/5-jahre-liquid-democracy-in-deutschland/ and covers the events up to August 17, 2011 prior to the Berlin city-state elections in that year.
Already in 1884, Lewis Caroll described in his work “The Principles of Parliamentary Representation” the idea of Delegated Voting: the previously known principle of transferring one's own vote on a given matter to another eligible voter (proxy voting) was extended in such way that votes may also be transferred through multiple steps (transitive delegation or delegated voting).
Those voting methods were not really put to practice; certainly a big issue was the necessary logistic efforts, considering the available instruments at those times. Not until more than 100 years later, in regard of the new possibilities of the internet, this idea was rethought, and the idea of Liquid Democracy as a democratic system with direct votes and the ability to delegate one's own vote in a transitive way was devised where delegations may be revoked at any time.
In Germany, the idea of Liquid Democracy is closely linked to the German Pirate Party since their formation in September 2006. On February 16, 2007, the first entry regarding Liquid Democracy was posted in the documentary Wiki of the Pirate Party in which the pursued goals were formulated. Also, the much later emerging conflict between verifiability of electronic ballots on the one hand and the want for data privacy protection on the other hand was already noted in that entry.
Half a year later, Jan Huwald, who was elected political director at the founding congress of the German Pirate Party, answered in an interview with Telepolis to the question concerning the aims of the German Pirate Party, that the party has been created purely grass-roots based and henceforth pursues the concept of Liquid Democracy in order to combine the advantages of representative and direct democracy. According to Huwald, it was planned to try the system inside their own party at first, then offer it other political parties, and finally add it to the manifest of the party.
A broader attention regarding this topic didn't arise until the election campaigning for the federal elections 2009 was all over. Due to the successful “Zensursula”-campaign, the Pirate Party had been able to increase their member count drastically. At the same time however, the party reached a limit regarding grass-roots based organization structures. The prospect emerged that either conventions with thousands of members would need to be held or that the member congress would need to be replaced by an assembly of delegates that couldn't meet the demand for grass-roots participation. In view of the above mentioned considerations, the wish for a Liquid Democracy solution was growing continuously.
At this time, there were rumors about a Liquid Democracy software that would become available soon. An event of the series “Datengarten” by the Chaos Computer Club was greatly anticipated on September 3, 2009 in Berlin, where Daniel Reichert and Martin Häcker of the Liquid Democracy e. V. would present the theoretical concept of the “direct parliamentarianism” (“Direkter Parlamentarismus”), whose implementation, however, was just a dream. The Pirate Party Germany created own workgroups for the issue. The executive board of the Berlin branch of the party entrusted to the local workgroup the preparations of an amendment to the branch's statutes for a binding usage of Liquid Democracy at the latest on the date of Berlin's city-state elections in 2011.
Not only because of that, Jan Behrens, Axel Kistner, Andreas Nitsche, and Björn Swierczek of the Public Software Group engaged in the subject of Liquid Democracy. Together, they designed rules of procedure for proposing motions in combination with delegated voting and a preferential voting system according to the Schulze method. In view of the legitimate concerns regarding secret electronic voting, e. g. as stated in connection with voting machines by the foundation “Wij vertrouwen stemcomputers niet” (“We do not trust voting machines”) and the Chaos Computer Club, it was decided to create a concept that doesn't involve secret voting.
Instead, only voting by roll call (recorded vote) was being planned for. Using a recorded vote for every cast ballot of each voter, the ballot is connected with a distinguishing mark (usually the person's legal name) which effectually allows the correlating of the person casting a particular vote. Thus, the system would be capable to make reliable decisions that could be trusted even though electronic media were being used. The worked out master plan was put to practice beginning September 2009, resulting in the software LiquidFeedback.
Friedrich Lindenberg was, parallel to the development of LiquidFeedback by Behrens, Kistner, Nitsche, and Swierczek, occupied with developing another Liquid Democracy software named “Adhocracy”, which had different rules and followed the concept of a “permanent poll”. Both LiquidFeedback and Adhocracy followed the Liquid Democracy principle of transferring votes using topic-based transitive delegations.
On November 25, 2009, the Liquid Democracy workgroup of the Pirate Party Berlin decided to test a Liquid Democracy software for preparation of their member assembly. For that purpose, it was suggested to the executive board to install a platform based on the prototype of LiquidFeedback and to test it with the members of the Berlin party. The executive board approved it. Funds for a server were provided and administrators commissioned.
While the particular operating terms were worked out, strong concerns to participate in voting by roll call turned out to be prevalent within the party. Thus, there were demands to be able to participate pseudonymously. Even though LiquidFeedback was not developed for pseudonymous usage, it was possible to achieve pseudonymous usage without modification of the software by abstaining from filling the field “Identification” in the user profiles. This field, which is not alterable for the user, serves the task of connecting votes with the person casting those votes. Even if it is technically possible to leave this field blank, such course of action inevitably leads to a loss of verifiability of the voting.
The pseudonymity pursued through this practice can only be achieved to a limited extent, as for accreditation of the party's members (e. g. for disabling accounts of resigned members), the invite codes for a party member in LiquidFeedback would need to be stored in the member database of the party. Therefore, an indirect link between the user account in LiquidFeedback and the person is made. By merging the information of the party's member database and of the database of the LiquidFeedback system, a depseudonymization of a single user or even all users is possible at any time. This can happen in a legitimate way (due to a decision of the executive board or by court order) or in an illegitimate and unnoticed way.
Because of the experimental character and the nonessential verifiability in case of an experimental system, there was an agreement to proceed as previously explained. In order to allow for a (limited) pseudonymous use of the system, the verifiability of the voting was relinquished.
Therefore, beginning on January 4, 2010, LiquidFeedback was tested as an own software installation by the Berlin branch of the Pirate Party and also used to prepare the upcoming member assembly in a non-binding fashion. At this assembly of the members, on February 28, 2010, the section “Liquid Democracy” was added to the statutes. The decisions made in the Liquid Democracy system of the Berlin branch of the Pirate Party would from now on count as recommendations to the organs (i. e. bodies) of the party. The originally aspired bindingness of the decisions wasn't fixed in the statutes, because it was considered to be illegal to formally bind decisions of a party's organ to decisions made in the LiquidFeedback system.
Such bindingness would be achievable through creation of a new party organ, which could make decisions independently and at any time through Liquid Democracy. Within the limits of statutes and manifest, binding decisions could be made, which could empower the party to react, for example, to politics of the day. For such an organ, however, a verifiability of the recorded votes by directly linking the cast votes with the person who cast the vote would be necessary. Besides a person's legal name as an identification token, a self-chosen name may be possible that would be presented by the member at an assembly or similar event, and which is then sufficiently linked to the person.
In the running LiquidFeedback system, an initiative that was proposed in June 2010 requesting to require “real names” in order to use the system was not approved. The counter-proposals[21a] to keep allowing pseudonyms were preferred by the Berlin Pirates.
At the federal party conference in Bingen, on May 16, 2010, it was decided through the enacted proposal Z013[21b] that the federal party would put a LiquidFeedback system into operation. It was decided that the decisions of the system would count as a non-binding view of opinions. When closing the assembly, it was also decided to hold the next party conference as an assembly to decide on the party's manifest and to use LiquidFeedback to prepare the assembly with regards to content. In the aftermath of the conference, under leadership of Christopher Lauer, the responsible member of the executive board, and of Alexander Morlang, who was commissioned with technical implementing, an operational concept – mostly based on the example of the Berlin branch – was prepared. Also here, a pseudonymous usage was arranged for and, thereby, the missing verifiability relinquished.
Despite the provided possibility to use the system pseudonymously, the two months preparation of system operation were characterized by an intensive, partly hysteric debate in regard of data protection and privacy. On the one hand, there was a debate whether “ordinary party members” should be politicians who are bound to transparent decisions, or whether their voting behavior would even be their privacy. On the other hand, it was discussed whether cast votes should be digitally “forgotten” – thus deleted – after a certain time. The latter was proposed amongst others by Frank Rieger, spokesperson of the Chaos Computer Club, in the Klabautercast by Martin Haase. At the same time, Rieger praised the ability to pseudonymously use the system, without even mentioning the loss of verifiability. The previously almost unnoticeable trenches between the different political camps, inside as well as outside the pirate party, became visible more than ever.
So it was that a so-called Clearing House (“Clearingstelle”) was incorporated into the concept for accreditation, which aimed to provide a better protection from depseudonymization of the users. Also, the executive board of the federal party made – for the reason of data protection – further changes to the accreditation concept, amongst them a deletion of all voting data after 4 party conferences. In addition to the fact that verifiability was already lost, by adding the Clearing House and the other changes to the operational concept, checking the accreditation of the members is now impossible without extensive efforts. The process “Teilnehmerkreisprüfung”, provided in the operation rules of the pirate party, has never to date been executed [i.e. as of publication of the original German article on August 2011].
Felix von Leitner summed up the situation in his usual direct manner on his blog: “Es geht nicht.” (“It doesn't work.”)
Nevertheless, the system was launched on August 13, 2010 on lqfb.piratenpartei.de. More than 3000 party members registered and in hundreds of issues the upcoming party conference for changing the manifesto was prepared. Within those 2 months until the party conference took place, the probably most intensive test of Liquid Democracy at all was undertaken. At the end, 28 initiatives that had been successfully finished in LiquidFeedback were enacted at the member's assembly. The precise effects of the usage of Liquid Democracy in the pirate party and the satisfaction of the pirates were extensively analyzed scientifically by Sebastian Jabbusch in his master's thesis.
The enquête commission “Internet und Digitale Gesellschaft” (short: EIDG), summoned on March 4, 2010 by the the German Federal Parliament (Deutscher Bundestag), got amongst other things the assignment to evaluate possibilities of electronic civic participation. Liquid Democracy then also became an issue, as well as the question regarding linking user accounts with persons. On July 16, 2010, Axel Kistner and Andreas Nitsche presented LiquidFeedback to the online working group of the EIDG and also outlined a potential use case. In order to get trustworthy results, an accreditation method was proposed as well. Every interested person should get a letter with a personalized activation code after their registration, and the name of every participating citizen should be displayed in the user profile.
It is notable that shortly thereafter the chairman of the EIDG, Axel E. Fischer (member of the German Bundestag, CDU), picked up the reasoning for linking user accounts with persons in a press release. In the newspaper “Badische Neue Nachrichten”, he argued for the citizen's possibility to participate in political decisions, but he saw a requirement for stating their real names.
Fischer would be right in regard of the verifiability of electronic voting – but unfortunately he pursued only the agenda of his political party with these statements, because at the same time he called for a “Vermummungsverbot im Internet” (ban on wearing masks on the internet) and a “Radiergummi im Netz” (eraser on the net). Surely by now, it got clear that it was impossible due to the heated atmosphere to discuss recorded votes and the linkage of political points of views with persons in a factual way.
Besides LiquidFeedback, also a use case for Adhocracy was presented to the EIDG. The software Adhocracy was meanwhile contributed by Friedrich Lindenberg to the Liquid Democracy e. V. The administration of the German Federal Parliament (Bundestag) let the two solutions be compared. Representatives of both projects answered a number of questions regarding the software and a concrete system operation. The administration of the Federal Parliament created a comparison out of it. Based on this, the EIDG decided on September 30, 2010 that they want to operate their own platform based on Adhocracy. This couldn't be put to practice though because the president (Bundestagspräsident) of the German Federal Parliament, Norbert Lammert (CDU) submitted the plan to the Commission For The Adoption of New Information and Communication Techniques and Communication Media (“Kommission für den Einsatz neuer Informations- und Kommunikationstechniken und Kommunikationsmedien”, short: “IuK”) of the Council of Elders (Ältestenrat). The commission decided then, that the expenses of € 60,000 to € 80,000 estimated by the administration of the parliament will not be incurred; even though it was the explicit mission of the EIDG amongst other things to explore the possibilities of new forms of participation, citizen participation, and usage of new participation forms (“Möglichkeiten für neue Formen der Teilhabe, der Bürgerbeteiligung und Nutzung neuer Partizipationsformen”).
The Chaos Computer Club didn't want to accept that situation. Its general member meeting decided on February 20, 2010 in Hamburg that the association would provide up to € 80,000 for the operation of Adhocracy. This made things hum. The experts in the EIDG pressed for starting up the system in either case – in case of need even outside of the Federal Parliament. Despite an intervention by Axel E. Fischer, this system went online in February 2010 at enquetebeteiligung.de as so-called “18th expert” of the commission (“18. Sachverständiger”). The operation of the system was carried out by the Liquid Democracy e. V. on behalf of the EIDG. Axel E. Fischer, as chairman of the commission, appreciated that after legal consideration and decision by the Council of Elders, there was legal security from now on and the next steps could be made.
Halina Wawzyniak of the party “The Left” (“Die Linke”) is pleased that after many months' effort of some of the commission's specialists, Adhocracy is finally officially employed by the EIDG. However, she insists on going further and that an authoritative citizen participation shall be established.
Wawzyniak notes with good cause that expertise and opinion will only be contributed to such a system if the system's results have a real impact and there is thus an incentive to participate. There will be only a few citizens willing to do time-consuming text work if the results are completely non-binding.
But Wawzyniak misjudges the state of the launched system. An accreditation of the citizens (as initially contemplated) was not arranged for. Thus, the Adhocracy instance comes with an open user registration, like a testing environment. This means it is possible without technical knowledge to practically create as much user accounts as one wants to, and to use these user accounts for voting. Thereby it is easy to give a higher weight to one's own opinion. The basic principles of democracy are glaringly violated here, making binding decisions completely unfeasible.
This problem was also noticed by the experts of the commission. Alvar C. H. Freude, expert member of the commission denominated by the SPD, noted in a comment to the blog of the EIDG that multiple registrations per person can either be technically prevented by no means at all or by a proper accreditation (e. g. through an activation code that is sent out by letter). According to Freude, such an accreditation would be nonessential, since there would only be determinations of tendencies (“Tendenzabfragen”). Furthermore, Freude carried on that an increased voting weight of members could be ruled out by disabling the delegation function.
This approach, however, is not suitable to solve the problem. A direct increase of their own voting weight by delegations can be prohibited indeed, but casting multiple votes for the same issue per person by using different user accounts is still possible. Notwithstanding, it was decided to deactivate delegations. Thereby, after surrendering the verifiability, the core element of Liquid Democracy to pass one's own vote (transitively) to other persons was also given up.
Besides the Pirate Party and the EIDG, also the party “The Left” (“Die Linke”) and the Social Democratic Party of Germany (“Sozialdemokratische Partei Deutschlands”, SPD) had a go at testing Liquid Democracy. Both with the fixed-term test of the party “Die Linke” as well as on the recently started portal of the SPD, an open registration is possible. Therefore, as previously discussed in regard of the EIDG, democratic principles are glaringly violated because it is easy to give one's own opinion a greater weight by registering multiple times.
A serious test is impossible under these conditions. Nevertheless, the party “Die Linke” admitted the most successful initiatives as so-called “request to speak” to their program commission. The SPD goes further and creates their “first online motion” for their next federal party conference.
After an excursion to federal politics, we shall return to the Pirate Party Berlin, as the party wants to be elected member of the city-state's parliament and the district councils of Berlin. Liquid Democracy is also an issue of the election campaign which the pirates want to score with. So writes the newspaper “Berliner Morgenpost” on August 14, 2011 about Andreas Baum, candidate on position 1 of the party's candidate list, that the Pirate Party has structural advantages compared to other parties. Members could promptly present their own positions using the software LiquidFeedback.
Christopher Lauer, also candidate of the Berlin Pirates for the city-state's parliament, announces at the short message service Twitter more clearly that Liquid Democracy only gains who votes for Pirates:
Let's examine what's behind this. The operation mode of the Liquid Democracy system of the Berlin Pirates has hardly changed since beginning of operation with the first prototype of LiquidFeedback on January 4, 2010 with the ability to participate in a (limited) pseudonymous fashion. Verifiable decisions by recorded vote are thus not accomplishable. The original plan to make binding decisions at the time of Berlin's city-state elections 2011 by using Liquid Democracy was not realized.
Even if Baum only talks of members presenting their own positions – thus not even making decisions – then this still doesn't imply that verifiability is nonessential. Because if Baum and other Pirates enter the Berlin city-state parliament (Abgeordnetenhaus von Berlin) on the 18th of September, and if they earnestly want to let their members participate in the political work, they will need a trustworthy representation of opinions.
Some of the candidates of the Pirate Party Berlin for the city-state parliament go further than Baum. They pledged themselves to base their own voting within the parliament on the decisions in the Liquid Democracy system of the party. According to their text, the signees would act in accordance to the opinion of the party's members, act in accordance to the recommendations decided in the Liquid Democracy system of the party, and, if they do not succeed with this offer, give up their mandate in the parliament.
If there would be undetected manipulations of decisions in the Liquid Democracy system of the party due to the absence of verifiability, then these manipulated decisions would cause real political ruling that would have been created in a non-democratic fashion and does not accord with the will of the party members. Therefore, the pledge of the candidates can only be honored in a responsible way if a Liquid Democracy system with verifiable recorded votes was available.
It has been shown that it is possible with Liquid Democracy to develop political positions and to constructively improve them in defiance of a big number of participants with differing interests and to make those positions capable of winning a majority. However, it will stay a dream for the time being to durably gain political influence by pseuydonymously taking part in trustworthy (hence verifiable) electronic voting without a visible link between the opinion and the own person. So it is necessary to make a decision: either bid farewell to the wish to stay incognito when voting by roll call, or bid farewell to the goal of real political influence through binding decisions with Liquid Democracy.
Already Jan Huwald, who was cited in the beginning of this article, advised of the necessity to deal with detail questions when implementing Liquid Democracy, since it was something novel and required its own “democratic culture”. In spite of everything, it shall be a worthwhile goal for being able to let citizens participate at any time and in the long run in political decisions, because the progress of human internetting would render this possible.