Liquid Democracy Documentation Project – Interviews #37-38 – Axel Kistner and Andreas Nitsche, Interaktive Democratie e.V. from Jose Ramos on Vimeo.

This interview is with two of the lead developers of the Liquid Feedback system, Axel Kistner and Andreas Nitsche. Liquid Feedback was the first software system that operationalized a liquid democracy voting system.

This interview starts with technical questions, but then ends with the story of how Liquid Feedback was born, Berlin’s hacker culture and Liquid Feedback’s integration into the German Pirate Party.


Axel and Andreas start by addressing some misconceptions about Liquid Feedback. The first issue is a question I raise with them at the start: if a liquid democracy system is not binding, does it not effect the legitimacy of that system, and a willingness for people to participate in that system?

A binding voting system is one in which a representative or a policy making body are compelled to put the outcomes of the liquid democracy vote into law. The opposite of binding is a reference system, in which the outcome of a liquid democracy process is used only as a ‘reference’ for the actual legal decision makers.

On this point they are very clear. Liquid feedback was not designed to be a decision making platform that compels binding decisions on a state / statutory basis. Rather it is designed to empower political parties, civic organizations or companies to self organize decision making. Parties or organizations using Liquid Feedback can make it binding if they want to, but do not necessarily need to.

Laws in Germany, and in most democracies, are very specific… that representatives are voted in by an anonymous / secret ballot. Therefore, the transparent way in which liquid democracy allocates votes transitively (everyone can see everything in the liquid democracy system) simply does not work at the state / statutory level of elected representatives. Also, once in, a representative can exercise their conscience on any vote. Thus even if it is binding at a party or factional level, representatives of parliament can still vote how they want. This is why Liquid Feedback is for parties to do distributed decision making on policy development, but is not design to replace the way that popular voting happens during state elections, and it cannot bind party representatives to liquid democracy based decisions.

In addressing this problem, the only exiting answer is that those standing for election in a party using Liquid Feedback or similar, are simply making a promise to their party members that, if elected, they will honor the general will of the results of the Liquid Feedback process.

Liquid democracy can also be used as suggestions to a board or a council of elected people – there is transparency in its use and if a board or council decide not to follow a suggestions, at least it is clear the difference between their decisions and the liquid democracy outcomes. A council can also get feedback that they would ordinarily not get, they can pose governmental or administrative issues to the community and get feedback.

Alex and Andreas are both satisfied with this compromise, as they believe political parties are a key institution for expressing political will. At least in German, parties are indeed foundational. Ben de Biel, another interviewee, told me how in Germany, a party gets elections funds per party member. This differs greatly from other countries where parties or candidates need to reach a threshold to qualify for election funds. They also argued that parties are identifiable – there is a clear boundary between inside and outside a party. They feel Liquid Feedback should not be used where there are not clear boundaries, such as a grassroots social movements or the like, as processes and results can be manipulated. Finally, parties or other organizations also do not require anonymity in voting, unlike the mandatory secret ballot for elected representatives of state elections.

Division of Labor and the Iron Law Of Oligarchy

An important aspect in considering the logic and rationale for Liquid Feedback is how the existing representative democracy systems uses a division of labour in politics. They argue we need this division of labour, as decision making does not scale well when we get to large numbers. But while representative democracy does this division of labor in a static fashion, liquid democracy helps to make this division of labor more fluid.

This issue of division of labour is also related to the idea of the “Iron Law Of Oligarchy”. German sociologist Robert Michels developed the theory of the Iron Law of Oligarchy in 1911. The theory put forth the proposition that all political systems and organizations, even when they start out as democratic, ultimately become oligarchic and ruled by a handful of people. He reasoned that, because systems that grow become large and complex, they cannot function as direct democracies, and power must be delegated; power eventually accrues within a small group of people, a “leadership class” that specializes in modes of administration. Even when democratic systems exists, this leadership class will over time dominate, as they can control the flow of information, use the system to reward personal loyalty among members and control the procedures used to make decisions. Michels theory is related to bureaucracy. Because any large system needs a bureaucracy to function effectively, this centralization through bureaucracy leads to an Oligarchy who’s primary motivation is to maintain its power.

Thus, division of labor in a representative democracy is subject to this “Iron Law”, by virtue of the fact that a representatives job is removed from voters for longish periods. They argue that this division of labor via liquid democracy helps to break the Iron Law, by reducing the prevalence of a specific decision making class – everyone is the politician.

I asked them if liquid democracy style transitivity leads to an uncontrollable accumulation of power, as some people in the system can be delegated hundreds of votes from other members in their own system. They responded that it actually does the opposite, it is the key to break the Iron Law of Oligarchy.

They argued that every step in a transitive system allows for interventions. Usually someone who is not in the political sphere is excluded by the existing bureaucracy or political administration. In liquid democracy those people who are normally excluded can come up with an idea and be heard and given delegated votes. They also argued liquid democracy taps the expertise across a system – liquid democracy can even be seen as the implicit search for expertise within an organization – it allows a faction to strengthen within an organization, which they see as healthy.

Birth of Liquid Feedback

They explain how Lewis Carroll, the British author of Alice in Wonderland, was the first to propose the idea for transitive voting in a thesis titled “The Principles of Parliamentary Representation” (1884). However, only with the advent of computer technology was liquid democracy technically possible.

Around the year 2000 several solutions to the problems of noisy minorities (over represented minorities who post over and over again) were being developed by those working in the online discussion forums area. Moderation was time intensive, and time based rank was problematic, thus people were looking for other solutions, and different system were being trialled that allowed types of delegation of powers.

In 2009 in the Berlin hacker community, “rumors” (by which they mean discussions) began about liquid democracy, where people actively talked about a computer software that could do this. People would use the term “Wenn liquid wird kommen” as a reference to an expectation that liquid democracy will arrive soon. People discussed this possibility in groups, but there was no software at the time that actually did this. They then talked about coding it and designing it, and then finally decided to do it. It required 4 month of voluntary full time work for the team to get to a first prototype. There was no money – just work as volunteers. Liquid Feedback was thus born.

The board of directors of the German Pirate Party wanted to implement it party wide, and thus they did a test phase from January 2010 (for 2 months) to the end of February. During the party’s general assembly they created new by-laws in which Liquid Feedback was added into their legal statues, not as a binding system, but as reference one as basis for their decisions.

*This research has been supported and draws upon work done at Leuphana University’s Centre for Digital Cultures as part of the Grundversorgung 2.0 Project. This post is part of the Liquid Democracy Documentation Project and

2 Comments Re-inventing Democracy: “Wenn liquid wird kommen”, an interview with Axel Kistner and Andreas Nitsche

Leave A Comment

Your email address will not be published. Required fields are marked *