From Computational Social Choice to Digital Democracy

Digital Democracy (aka e-democracy or interactive democracy) aims to enhance democratic decisionmaking processes by utilizing digital technology. A common goal of these approaches is to make collective decision-making more engaging, inclusive, and responsive to participants’ opinions. For example, online decision-making platforms often provide much more flexibility and interaction possibilities than traditional democratic systems. It is without doubt that the successful design of digital democracy systems presents a multidisciplinary research challenge. I argue that tools and techniques from computational social choice should be employed to aid the design of online decision-making platforms and other digital democracy systems. 1 The Potential of Digital Democracy Recent years have witnessed an increasingly intense debate around the potential (and the risks) of the usage of digital tools for democratic decision-making [Contucci et al., 2019; Sgueo, 2020; Bernholz et al., 2021]. A common goal of digital democracy approaches is to utilize modern information technology—in particular, the Internet—in order to enable more interactive decision-making processes. Designing a digital platform for collective decision-making requires a huge amount of design decisions regarding, for example, interaction possibilities, elicitation techniques, and preference aggregation mechanisms. However, most existing designs are rather ad hoc in nature and little attention is devoted to a principled comparison and evaluation of methods. The study of collective decision-making lies at the heart of social choice theory [Arrow et al., 2002]. I argue that tools and techniques from social choice theory and, in particular, from computational social choice (COMSOC) [Brandt et al., 2016b], an interdisciplinary research area at the intersection of computer science and economics, should be employed to build and to evaluate digital democracy systems. Putting digital democracy on a solid social-choice-theoretic foundation decreases the risk of employing methods with unintended flaws and has the potential to enable fair and participatory collective decision-making processes even for very large groups. 2 Enabling Democratic Participation at Scale In the following, I provide examples of challenges that are encountered when building digital democracy systems, together with pointers to my own work in COMSOC that is relevant for tackling these challenges.1 What these examples have in common is their attempt to make sense of large amounts of contributions stemming from a large number of participants. 2.1 Liquid Democracy The paradigm of liquid democracy (aka delegative voting) aims to reconcile the idealistic appeal of direct democracy with the practicality of representative democracy by allowing participants to choose whether they want to vote directly on a particular issue or whether they want to delegate their vote to somebody they trust [Blum and Zuber, 2016; Valsangiacomo, 2021]. Delegations are topic-specific (i.e., voters can specify different delegatees for different issues), transitive (i.e., voting power accumulates along delegation paths), and delegation decisions can be changed at any time in order to hold delegatees accountable. Liquid democracy, which is an integral part of the digital democracy platform LiquidFeedback [Behrens et al., 2014], enables participation at scale by giving participants the opportunity to have their say on all issues, but not requiring them to do so. Liquid democracy has been studied theoretically, and applied practically, in various ways in recent years [Ford, 2014; Paulin, 2020]. Many variations and extensions of the basic model have been proposed [Gölz et al., 2018; Colley et al., 2020; Kavitha et al., 2021]. My own work in this area explores ways for making liquid democracy more flexible by allowing voters to delegate different parts of their preference ranking to different delegatees [Brill and Talmon, 2018] or to specify ranked lists of delegatees [Brill et al., 2021a]. 2.2 Preference Elicitation & Aggregation Digital technology also enables novel preference elicitation methods. For example, in so-called pairwise wiki surveys [Salganik and Levy, 2015], participants are repeatedly asked to make pairwise comparisons between alternatives. Each participant can answer arbitrarily many pairwise queries, and I apologize for the focus on my own work, which is due to the nature of the Early Career Spotlight track. Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence (IJCAI-21)