Liquid Democracy

[Türkçe]
[Substack]

Trust in democratic institutions has been on the decline globally, particularly in decision-making bodies like parliaments. This trend isn’t a left or right thing – it seems to span all corners of the ideological spectrum. My guess is that if citizens had more direct and visible mechanisms for influencing institutional processes, it might help rebuild public trust. Let’s see how we can do this.

When we think of a typical parliament in today’s world, where elected representatives vote for laws, we are imagining an institution that runs on what is called a representative democracy. Let’s understand the trade-offs of this system:

• Advantages of representative democracy: The population does not have to be actively involved in every decision, which means that they can continue to live their lives and contribute to society in other ways. Also, the representatives form a group of people who are responsible for drafting laws (which is a group that we need), and it is convenient that the laws are proposed and voted on by the same people.
• Disadvantages of representative democracy: It is extremely improbable for a representative’s views to align 100% with the views of each person they represent. So, some views will not be represented fully in the parliament, and it is practically impossible to prevent this. According to a study on Swiss politics, “on average, individual legislators vote in accordance with the majority of their constituents in 66.8% of all decisions”. In a representative democracy, the public faces difficulties in expressing its views in a fine-grained way.

If we imagine representative democracy as one end of a spectrum of democratic systems, then the other end of the spectrum would be direct democracy. In a direct democracy, every decision typically made by a parliament would instead be determined through a nationwide referendum. Let’s see the advantages and disadvantages of direct democracy:

• Advantages of direct democracy: The population has a direct say in every vote. An ideal form of democracy, in some sense.
• Disadvantages of direct democracy: Everybody is kind of expected to participate in every vote. You need to have a deep understanding of many subjects to be well-informed about everything happening in the country. Too much mental load in many cases. Also, if there are no representatives, who is responsible for drafting the laws people vote on? Maybe a community forum where citizens propose laws, but that would be very, very messy.

We observe that both representative democracy and direct democracy have decent advantages, but they also have nontrivial disadvantages. Specifically, we saw that it is kind of difficult in a representative democracy for a citizen to express their views outside elections. Could this fact (at least in part) explain the decline in trust toward democratic decision-making bodies?

Liquid democracy is an interesting middle ground between representative democracy and direct democracy. Sources differ on the exact definition of liquid democracy, and it has many flavors like ice cream. To keep things simple here, let’s use a straightforward system that incorporates key liquid democratic principles. In this system, citizens elect representatives to the parliament as usual, and each citizen has the additional ability to override the portion of their representative’s votes (in the parliament) that corresponds to them. For example, if each representative represents 100k people, if you override, you override 1/100k of your representative’s vote. The decision of the parliament is calculated by separately summing the “yes” and “no” votes from each electoral region. If a representative votes “no” but 20% of the people they represent override this vote, then the contribution of that region to the sums would be 0.8 votes for “no” and 0.2 votes for “yes”. Let’s predict the implications of this system:

• This solves the problem of representatives providing a very coarse-grained representation of their constituents’ views.
• The population does not have to participate in every vote. You only show up when you disagree with your representative in a topic that you feel strongly about.
• There are still representatives, so we don’t have to search for a group that is responsible for drafting laws and proposing the topics of the votes.

What is especially beautiful about liquid democracy is that it is a modular (or add-on) system, i.e., you can integrate liquid democracy into an existing electoral system without substantial changes in the existing electoral system. The flavor of liquid democracy I described above does not care how the representatives were selected. It concerns itself only with what happens in the parliament, after another system selects the representatives. That said, it assumed that there is one unambiguous representative for every citizen, which is true only in countries that use the first-past-the-post method, like the UK. For proportional representation-based systems, some slight adjustments would be needed.

Some other notes about liquid democracy:

• Sure, liquid democracy gives leverage to citizens. However, with great power comes great responsibility. In return, a representative gains the power to say, “if you really cared about this issue, then why did you not override my vote?” Representatives also benefit from liquid democracy.
• Citizens should be able to submit their parliamentary vote both before and after the representatives’ vote. The after is important, as some people might want to first see what their representative does.

Liquid democracy needs online voting

A national parliament conducts hundreds of yes/no votes every single year, as demonstrated by the vote schedule of Estonia’s Riigikogu. If we want every citizen to have the option to participate in each of those votes, then we absolutely need online voting. Not the voting machines that we see in voting booths around the world – I mean voting from your phone/computer. No other system can reach the speed and efficiency needed to enable liquid democracy.

This speed and efficiency (and comfort) are always brought up by the proponents of online voting. You can vote without leaving your home (even your bed), and the country saves X million [insert country’s currency] per election. For a lot of people, these reasons are not really enough. Understandably so. Is the risk of cyberattacks really worth it? How do I verify that my vote was counted? In my experience, people tend to support online voting mostly after learning about liquid democracy. When a new technology arrives, we shouldn’t just focus on how it makes the old system better. We should also think about what new systems it can enable.

If there is one thing you should know about the tech behind online voting, it’s that it is here, and that it is just beautiful (especially the cryptography). We have the know-how to build a secure, reliable, verifiable and flexible online voting system.¹ You don’t have to take my word for it, just check out what Josh Benaloh says ("… end-to-end verifiability is ready to go now"). Some countries have already been using it (e.g. Estonia). There are also effective approaches for integrating online voting into existing systems (paper ballots never have to, or should, disappear completely). Of course, as always, change will take time, change will be bumpy, change will be controversial. But looks like it is a question of when, not if.

I won’t go into how to build good voting systems in this post. That deserves its own blog post.

How far should we go with liquid democracy?

We might be tempted to broaden the scope of liquid democracy to include decisions made by professionals such as judges, central bankers, or military generals. Why not stop with representatives in a parliament? I would argue that the effects would be net negative. A representative is responsible for representing a certain portion of the population. They get that seat in the parliament not because they are experts at population representation (I doubt that’s a thing). They get that seat because the population believes that they are a good approximation of the population’s views. The argument for liquid democracy asserts that if a representative was always meant to be an approximation of the views of the citizens, and we now have the technological means to directly capture those views, it might be worth considering this opportunity.

This logic doesn’t work for a lot of professionals serving a country. A military general earns their position not because they are a good approximation of the decisions citizens would make in a war. The general earns their position because they are an expert in military decision making. Similarly, judges and central bankers earn their positions because they are experts in their respective fields. If citizens override the decisions of the experts, it’s more likely than not to go bad. However, note that I was talking about the decisions by those experts while they are doing their job. In many cases, there is still a vote (usually by representatives in a parliament) to determine whether the expert gets the job (e.g. Supreme Court judges in the US). Such votes are necessary and acceptable.

It is also worth noting that liquid democracy can work only for decisions with clear-cut options. A president or a prime minister makes numerous small and unquantifiable decisions every day, as opposed to the well-defined votes taking place in a parliament.

So, if we ever implement liquid democratic ideas in governance, let’s not take it too far. At its most extreme, every decision made by a representative in a parliament could be fair game, but anything beyond that is likely too much.

What are the next steps?

We should always start small. If a group of people ever decides that liquid democracy is a good idea, they should first run a pilot program in a small, local decision-making body. Try it for some time. Collect feedback. If it works, go one level up. The general election should come much later.

The first system to incorporate liquid democracy should be something as straightforward as the one I described above. Only if that works should we attempt more complex flavors of liquid democracy.