Project

European City²: Rethinking Democratic Decision-Making

European City² is a Horizon Europe project researching the physical and mathematical foundations of democratic decision-making, combining agent-based simulation, large language models, Active Inference, and quantum computation to test how different voting rules shape democratic resilience. It draws on four strands typically treated separately: Arrow’s impossibility theorem in social choice, quadratic voting (Weyl), Active Inference and the Free Energy Principle (Friston), and recent results in quantum social choice (Bao, Fabinger, Freedman).

The central hypothesis is that democratic aggregation is incomplete when it considers only individual preferences. Introducing a second input—the collective goal and the reciprocal valuation of citizens who sustain it—addresses the limitation identified by Arrow’s theorem. The project refers to this hypothesis as emergent altruism and tests it experimentally.

Objectives of European City²

European City² aims to:

  • Investigate the limits of Arrow’s impossibility theorem using quantum and classical methods, building on results showing that quantum voting can violate its classical assumptions.
  • Test emergent altruism empirically by comparing one-person-one-vote, quadratic voting, and quantum-inspired aggregation on synthetic populations calibrated against real-world experiments in Denmark.
  • Simulate democratic dynamics at scale using large language models and Active Inference, modelling attraction–repulsion dynamics in voter belief change.
  • Conduct pilots in contrasting civic contexts: the City of Aarhus (high-trust, governmental) and QuantumBasel in Switzerland (industrial, competitive).
  • Translate findings into policy through GDPR-compliant, FAIR-aligned recommendations for municipal and EU institutions.

Expected Outcomes of European City²

The project will deliver: a calibrated simulation platform of up to 10,000 agents based on Active Inference and large language models; a classical, polynomial-time implementation of a quantum-inspired voting rule usable without quantum hardware; empirical tests of five falsifiable predictions on when enriched voting rules outperform majority voting; and policy recommendations for integration into existing democratic processes.

All methods and outputs will remain open and documented under FAIR principles.