Diversity and Expertise in Representative Governance

Representative democracy is a widespread and essential part of democratic governance. Our understanding of it has been largely shaped by the triumph of elections that vest peripheral access to power through episodical polls. How would representative democracy look under different selection rules? In an attempt to reflect on foundational principles on which we could build a renewed case for representative democracy as democratic governance, this thesis explores democratic innovations for selecting representatives and focuses on the interplay between selection mechanisms, epistemic performance, and procedural aspects. The first chapter investigates the optimal number of voters needed to aggregate votes on a binary issue under majority rule. It takes an epistemic view where the issue at hand has a ground truth “correct” outcome and each one of 𝑛 voters votes correctly with a fixed probability, known as their competence. Assuming that the best experts, i.e., those with the highest competence, can be identified to form an epistemic congress, this chapter surprisingly shows that the optimal congress size should be linear in the population size, even with expert decision-making. The next chapters delve into the concept of liquid democracy, a governance mechanism in which citizens can either vote directly or delegate their votes to others, and examine the epistemic and procedural performances of this approach offering insights from both theoretical and empirical perspectives. Taking an epistemic view, the second chapter highlights delegation scenarios where liquid democracy is likely to find the ground truth. It treats delegations as a stochastic process akin to well-known processes on random graphs —such as preferential attachment and multitypes branching process—and relate their dynamics to liquid democracy’s performance. Along the way, it proves new bounds on the size of the largest component in an infinite Pólya urn process which may be of independent interest. The third chapter presents empirical experiments designed to compare liquid democracy with direct democracy, the counter-factual. It validates the theoretical findings of the second chapter, providing evidence that delegation mechanisms align with theoretical expectations. The fourth chapter analyzes delegation dynamics in a real-world setting and explores how liquid democracy functions in scenarios with contentious issues. It reveals insights into the patterns of delegation and the usage of liquid democracy in non-epistemic contexts. The fifth chapter reflects on lottocracy (selection of representatives at random) and proxy democracy (selection based on self-selection and flexible nominations that determine the relative influence of representatives) as two models to select representatives. It investigates the procedural aspects of both selection mechanisms exploring how inclusion, equality and legitimacy would be realized under lottocracy and proxy democracy. The sixth chapter, drawing on computational social choice, formulates a unified framework for comparing selection mechanisms. It devises a model in which different selection mechanisms can be formalized and evaluated axiomatically. It classifies selection mechanisms based on whether they are open-closed, flexible-rigid, and direct-virtual and propose the following five axioms: proportionality, diversity, monotonicity, faithfulness, and effectiveness. Throughout, the thesis intertwines insights from mathematics (social choice theory, applied probability, statistics), political philosophy, and empirical analyses to provide a comprehensive exploration of different facets of representative democracy. The interdisciplinary approach reflects the complexity and richness of democratic governance and calls for continued collaboration across disciplines to tackle its challenges and shape its future.

Name thesis:degree_name

Doctoral

Department dc:contributor.department

Massachusetts Institute of Technology. Institute for Data, Systems, and Society

Grantor dc:publisher

Massachusetts Institute of Technology

Year dc:date.issued

2023

Author dc:creator

• Revel, Manon

Advisor dc:contributor.advisor

• Jadbabaie, Ali

Statement dc:rights

• In Copyright - Educational Use Permitted
• Copyright retained by author(s)

Licence dc:rights.uri

• https://rightsstatements.org/page/InC-EDU/1.0/

Handle dc:identifier.uri

https://hdl.handle.net/1721.1/154374

OAI identifier oai:identifier

oai:dspace.mit.edu:1721.1/154374