My recent research explores how new voting methods can be developped on blockchains. In particular I focus on Liquid Democracy a hybrid form between direct and representative democracy. In Liquid Democracy, voters can transfer their voting rights to other voters if they believe they are more knowledgeable. This delegation is supposed to result in more optimal decision-making while still taking into consideration preferences at an individual level.

Liquid Democracy can be used only as a voting method or as a new form of democracy. The second alternative is often associated with delegating votes per subject because people are generally not expert on all subjects. My research shows that although enticing in terms of representation, Liquid Democracy increases the risk of the “tyranny of the majority” because it fails to account for intensity of preferences.

Currently, no proposition for Liquid Democracy as a new form a democracy proposes a framework for bundling or vote trading which would result in voting the outcome favoured by most of the people but not necessarily the most favoured outcome. This work is greatly inspired by The Calculus of Consent by Buchanan and Tullock (1962).

A Public Economics analysis of voting methods is provided and blockchain-based implementations are discussed. As blockchains easily allow to transfer assets (such as voting rights) and are often considered as potential voting platforms, there is a flourishing literature on the subject of blockchain-based Liquid Democracy. The conclusion is that large-scale implementation of Liquid Democracy seems irrelevant today but that Liquid Democracy may be a inclusive voting method for small-scale homogeneous communities. This is consistent with the current blockchain ecosystem who daily explore new governance methods and implement them on DAOs.

I will present this paper at the 2022 Meeting of the European Public Choice Society and this paper is undergoing peer-review in a Scopus journal.

This paper owes a great deal to the reviews of Sofia Cossar and, as always, to dedication of my supervisors P. de Filippi and Bruno Deffains.