Agent Frameworks

New paper finds Borda count beats STAR, plurality in democratic voting models

Phase transitions in optimal voting rules mapped across rugged landscapes

Deep Dive

Joshua Nunley's paper "Democracy on Rugged Landscapes: Phase Transitions in Optimal Voting Rules" (arXiv:2606.02813, submitted to ALIFE 2026) models collective governance as optimization on NK fitness landscapes. In this framework, shared bits represent laws updated by voting, while individual bits remain fixed. A cross-dependency parameter α controls how legislation's effects depend on personal circumstances. The study compares eight standard voting methods—including plurality, Borda count, STAR voting, and cardinal score voting—across landscape ruggedness K from 1 to 20 and α from 0 to 1, with 1,000 runs per configuration.

The results reveal sharp phase transitions in direct democracy: cardinal score voting dominates on smooth landscapes, ordinal scoring with p=0.35 at low-to-moderate ruggedness, Borda count across a wide middle range, and STAR voting at the highest complexity. A two-parameter empirical formula reduces the (K, α) plane to a single complexity axis for visualization. Remarkably, Borda count achieves the highest mean fitness and lowest variance across most of the parameter space. The paper also introduces a representative democracy model with identity weight β and candidate self-interest p_self. Under representation, cardinal score voting dominates most regimes, with plurality emerging as the top method at high β and low-to-moderate p_self.

Key Points
  • Borda count outperforms all other voting methods on both mean fitness and variance across most complexity levels
  • Phase transitions in optimal voting rules occur as landscape ruggedness increases
  • In representative democracy models, cardinal score voting dominates except at high identity weight where plurality wins

Why It Matters

This research provides a rigorous computational framework for choosing voting rules that maximize collective outcomes in complex policy landscapes.