Berlin, 2015-10-12

What's Wrong With Liquid Democracy?

Die deutsche Version hat einige Paragraphen anders als die englische.

Improving Participation with Liquid Democracy

In the year 2010, with the introduction of Liquid Feedback (LQFB) as a nationwide survey tool for the German Pirates, the tool saw a staggering participation of over ten thousand people. That is roughly a third of the size of the entire party at the time. Usually active participation in political parties is below ten percent, yet Liquid Democracy mobilised over a third! The only reason the number of participants quickly collapsed was because the tool didn't get properly integrated into the party statutes. The results of the surveys were ignored by the activerts that dominated the physical general assemblies, so the Piraten killed the tool that could have made them make a difference. With such a large chunk of population regularly participating in Liquid Feedback, such an electronic parliament could very well be the path for better citizen participation in any democracy, being not only more efficient than paper referendums, but also wiser.

As late as March 2015, researchers of the Mainz and Cologne universities published a paper called Voting Behaviour and Power in Online Democracy: A Study of LiquidFeedback in Germany's Pirate Party describing how democratising a liquid democratic platform can be. How the so-called "super delegates" stabilise the democratic will, protecting effectively against demagogy and manipulation.

What if the Participants aren't Uniform?

You may think the methods of LQFB only work for a uniform slice of population, but that has already been disproven. Back in 2012, when the Piraten hit that 12% high in surveys, a FORSA survey stated that the Piratenpartei was representative of the entire German population, not just a niche.

But even if it was a biased view of one party, the method alone makes it much closer to the needs of the population: All traditional parties are ruled by a small leadership group who can all be corrupted or blackmailed. If a political organisation is ruled by most of its members, then the mere fact that thousands are deciding rather than a chosen few is a guarantee that they will be closer to the general will of the population than any political party ever before in human history.

There can even be extremily divergent views in the participation group, yet the method forces everyone to put ideological barriers aside and focus on pragmatic options within the fact-oriented problem space. The problems started when the Piraten betrayed the LQFB method.

Improving the Transparency of Trust Relationships

It is quite problematic in Representative Democracy that trust towards a representative is an intransitive relationship. Liquid Democracy makes this aspect openly visible, teaching the delegating person to choose a better person to delegate to, whenever the previous one has been acting in a dissatisfactoring way. In particular Liquid Democracy allows to change the trust delegation at any time, without having to wait for the next round of elections and without needing to coordinate with other participants. That alone makes liquid delegations fundamentally different from representative elections.

Secret Vote, an Instrument of Lobbyism

It is generally assumed, that secret vote serves the purpose to protect the parliamentary from being coerced to vote in ways against their will. Research has shown instead, that secret vote is a much more powerful instrument of corruption. If the parliamentary wants to do a favour to a lobbyist they need a secret vote in order to be able to deny any influence. Research at Berkeley University questions the usefulness of secret voting:

We found that the imposition of the secret ballot always increases the scope of vote buying — more people vote insincerely under the secret ballot than under the open ballot. We also found circumstances where, paradoxically, the imposition of the secret ballot makes it easier for interest groups to wield influence. In particular, for close elections where the bulk of the supporters of an interest group's desired policy are lukewarm, it is cheaper for that interest group to buy the election under the secret ballot than under the open ballot. Taken together, this suggests that the common intuition about the effectiveness of the secret ballot as a robust deterrent to electoral corruption needs to be revisited.

With its history of mafia, subversion and corruption it is quite surprising that the people of Sicily would choose Liquid Feedback as a tool to collaboratively develop their election program in full transparency, with profile photos and names and surnames. The motherland of the mafia is sick of its social cancer. It has understood that standing up in masses makes them immune, using transparency as a weapon. When hundreds or thousands take the right decisions together, what is the mafia politician to do? How can the lobbyist wield influence?

As neither corruption, nor blackmailing, nor spread of manipulative false information works very well within Liquid Feedback, Germany's spin doctors needed to damage the acceptance of the instrument itself. This has been quite successful, even though none of the criticism we've seen is founded on facts.

How Important is the Inactive Majority?

Just because in most countries a majority of citizen doesn't actually want to be involved with everyday politics, even a minority of a percent, if representative of its population and equipped with a corruption- and lobby-resistant technology such as Liquid Feedback, can make political choices that are more in the interest of the general population than any few hundred elected parliamentarians possibly could.

The question I asked myself is how to get the grandmother who lives two doors down to use the computer, or how to get the mother of 5 children two floors down to do the same.

Is this really the right question to ask, David Bovill? If, in a choice of suboptimal variations of democracy, you focus on collecting everybody's opinion rather than working towards consensus, you will run into the many problems of direct democracy and once again re-enforce the rule of representation.

In February 2017, a research study was published that found out how aggregated knowledge from a small number of debates outperforms the wisdom of large crowds. So a couple of roundtables of average Brits could have made a wiser choice than the BREXIT referendum. What if these roundtables weren't actually limited in size of participation? That is essentially what a correct use of liquid democracy enables — maximizing on the ability of humans to make better decisions when doing so in a collaborative way rather then summing up everybody's prejudices and indoctrinations.

The method employed in LQFB recognizes fallacious proposals (see below), therefore by the time the debate is through those prejudices and indoctrinations will no longer be up for vote.

Considering the depressing condition of our worldwide representative democracies, it would be a quantum leap into a brighter future if all the ones who care about what is going on, finally get to have a say. And we've seen in our deployment in Italy that, as soon as we agreed to take decisions on the platform, all the pensioners (who traditionally are the spinal column of Italian politics) quickly figured out how to have their say on it.

It was merely a question of not offering any alternative, which is one of the preconditions for successful deployment listed in the "Principles of Liquid Feedback" book. If the extroverts can bypass the platform and impose their will in a physical assembly, the platform will no longer be of interest.

Whereas, whenever somebody has a brilliant idea to contribute (if we are working towards consensus, then it's the ideas, facts and perspectives that count, not the opinions), the German Pirates experienced that a computent person in their social surroundings would help them have it contributed to the platform. No important contribution gets lost.

This idea of text based, deliberative democracy, is fundamentally elitist hacker culture.

No, certainly not. Parliaments aren't hackerspaces. All political will needs to be fixated in written words or the next speaker will turn it around to mean the opposite. Therefore written text is an absolute inevitable necessity for any legal transaction, including taking decisions for large common goods like a city, a nation, a continent or our finite planet. Again, anyone can express a new perspective on a political problem in form of a speech or video, but then, in order to work with it, somebody needs to care sufficiently to put it in written form.

What you describe is a perfect medium for a certain type of human - it is ideal democracy for hacker culture.

Liquid Feedback and similarly advanced platforms reflect the complexity of a parliamentary system. Everyone who tries to do something super-intuitive will first start out with something far too simplistic and then in the process of improving and fixing the problems, end up re-inventing existing tools with their need to learn the concepts behind collective rationality and debate. The idea that this can be simple is a fallacy, and hackers need just about the same amount of time to understand liquid democracy as anybody else.

We should recognise that there is a real danger of hackers / coders becoming the new lawyers, bankers, politicians and high priests of our age.

Yes, but that is because technology has been left in an unregulated condition for several decades now, allowing the market to never put the necessary amount of attention into computer security. But that is a debate we are discussing elsewhere.

We need to recognise that people who see things this way are not "rational" while other people are "irrational". They are just different.

As the 'aggregated knowledge' paper shows, Individual Cartesian Rationality has failed. It has succumbed under the weight of psychological biases. You can put as many opinions on a stack, they can still end up being quite irrational. That's why LQFB uses an approach which is equivalent to people sitting down at a roundtable and discussing things out. Just below in this document we discuss such Method of Collective Rationality.

For those reasons we need physical meetings, we need face-to-face discussions, we need art, and documentary, and anthropological techniques in our politics and our political systems.

It is wonderful to have face-to-face exchanges of ideas, but when it comes down to structuring consensus, the power of actives and extroverts in real-life encounters makes it hard to obtain honest results. Some people will not dare to speak up. Some people will not have found the time to participate. And the time is not enough to reconsider all the facts and details needed to make good decisions. Some people may even have an interest in not mentioning certain aspects. The bigger the group of people that should decide together, the unlikelier the results indeed qualify as democratic. If the group can't even meet in one space at one given time, then the attempt to hold voting across multiple locations will introduce biases as different congregations discussed different perspectives of the topic. In thousands of years of democracy, humans still haven't figured out a way to hold decision-making meetings in a satisfactory way. So, when it comes to structuring the facts needed for decision-making, unless we know anything better, please don't try to do without liquid democracy.

Methods of Collective Rationality

While methods of Direct Democracy can trigger demagogic effects of collective satisfaction which are not based on facts, this is quite different with the Liquid Democracy procedures. Only rationally convincing argumentations can produce the necessary consensus. The Dunning-Kruger effect, for example, by which the incompetent isn't aware of their own incompetence, is dampened in Liquid Feedback, as both the above research papers illustrate. This is the case for several psychological biases which would otherwise come to play in Direct or Representative Democracy. The delegations also serve a purpose of compensating for the deficiencies of individual psychology and rationality.

But that doesn't Fix Individual Rationality

We're not trying the impossible here. There will always be individuals who individually think the collective has made whatever terrible mistake. You can't stop that. You need good social rules for that, to make sure the now underdog activerts don't smash the collective process like it happened with the Piraten.

How can we even Agree what Facts are?

Let's say you are proposing a referendum on leaving the European Union, making all sorts of claims on what improvements it would bring. Let's say I see that you are making claims that do not correspond to factual reality. Since facts are part of a pluralist space themselves, there is no way to logically categorise them as facts, but there is a social way to do so. I can start a quick consensus process to see if a large majority agrees with me that this or that claim you made is factually incorrect. Should the majority agree with me, it would establish a fact as a fact we agree upon — this is something a Validating Council can work with: it can invalidate your proposal until you remove those false claims. This way we ideally arrive at a condition that all the options on the table for the final political choice to be made are free from manipulatory false information.

A distinction of facts and moral beliefs isn't even necessary: if your proposal hits against a wall of moral beliefs, then the consensus can be that your proposal is based on something everyone else fundamentally rejects. Let's take the ideology that is at the foundation of our current economic system as an example. To accept de-facto slavery of the poor is a precondition in order to be able to enjoy the leadership of the wealthy sociopaths and their broken idea of economic growth. Since slavery is fascist, such policy would be ruled out as a reasonable political option. Next thing that proponent could present would be to deny the existence of slavery, possibly promoting some trickle-down ideology. So we return to a fact-checking constellation, considering today's awareness of the failure of trickle-down.

What if someone attacks a proposal on formal grounds?

It is very frequent in all kind of democratic organisations that, should somebody have malevolent reasons to oppose your proposal but can't come up with any convincing arguments, they will try to question the formal validity of the proposal. Was it presented in the right form, in the right place, using the right policies? Or, in a more advanced form, claiming that the proposal wouldn't be legal and so it must not be ratified. Lawyers would typically find themselves bearing such a power on the Assembly. With a good dose of so-called FUD the participants can be manipulated into not voting according to their political convictions, but rather to avoid trouble. Unfortunately, this scenario doesn't just happen in physical conventions, it is also very frequent in digital assemblies. Liquid democracy alone has shown not to be enough to defuse manipulation.

Luckily, this constellation can be avoided by putting an organ like the above-mentioned Validating Council in charge of checking the formal validity of proposals. If anyone thinks a proposal isn't legit, they must not confuse the participants but speak directly and only to the VC for them to check the matter. The Assembly must never judge a proposal on a meta level of formality, but always strictly in regard to its political contents. It must trust the VC that it will weed out a proposal if there is anything formally incorrect about it.

Who is to Bear the Responsibility of Consequences?

For each successful proposal in Liquid Democracy there are authors whose heads can roll. Wherever there is no elected board, there must be a coordinating group to guide the implementation of the collective decisions. In fact, the responsibility usually lies with the taxpayer who has to finance the good and the bad choices of the politicians. By bringing decision-making power back to the collectivity, we are bringing it back to the ones who already bear the financial consequences.

Those cantons of Switzerland that make their finance planning using direct democratic decision-making have been shown to be much more reasonable in the use of financial resources than their representative counterparts. Representative politics motivates its actors to make themselves be seen for re-election, whereby money plays a secondary role. Apparently, direct democracy is a better choice for these kinds of decisions. Liquid Democracy inherits this type of corruption resistance of direct democracy, making it a more reasonable tool also regarding finances and expenses.

And there it was, unexpectedly: the Collective Intelligence

Only few installations of Liquid Feedback had such a large number of participants to show the impressive effects of collective intelligence that the Piraten experienced in those early years. Whenever several smart people would work on the proposals together, the quality would rise to impressive levels. No way that somebody would be able to slip in a non-scientific statement, nor a special interest earmark. No way a demagogic initiative would arrive to the voting phase without getting unmasked first. There were always enough people paying attention. It would take a massive public relations effort to convince an entire parliament, be it physical or virtual, to agree on a false information. Even then, liquid democracy is better suited to dismantle a falsehood than a traditional representative parliament.

Leadership without Elected Leaders

Using Liquid Democracy as a principle for inner organisation of political parties or any other kind of social structure of humans allows to avoid the traditional approach of representative leadership. Both Partito Pirata and Lista Partecipata work with such a method of online decision-making and replaced the traditional board by a series of control organs and special appointees.

Permanent Assemblies are also in use by the Piratenpartei Österreich and the German Glitzerkollektiv. In combination with a traditional board of directors, however.

That's it!

Frequently people tend to expect Liquid Democracy to fulfil some wondersome ideal of political perfection. That kind of rhethorics distracts from the fact that it is a great step forward from the broken status quo of Representation. Liquid democracy is the new least worse form of governance, nothing more and nothing less.


Last Change: 2018-06-18

An earlier version of this document was cited and criticized by Wessel of DiEM25's Democracy Lab. It has subsequently been reorganised and partially rewritten for clarity and in an attempt to answer the questions raised.

My pages.

Go have a chat.
Use Tor.
Use IRC.
Use IRC over Tor.

But, if you already sold your
soul to the surveillance
market, you can . . .  

Tweet this.
Share on Facebook.
Stumble upon this.
Find it delicious.
Digg this.

Follow tweets in
Deutsch, Italiano
and English.

Follow on Facebook in
Deutsch, Italiano
and English.

CC-BY-SA, carlo von lynX