Fairness and Efficiency in Fair Division: An Empirical Analysis of Mechanisms for Fair Allocation

Abstract

This thesis presents a comprehensive empirical analysis of two prominent fair division algorithms deployed in real-world platforms: the Maximum Nash Welfare (MNW) algorithm for indivisible goods allocation (Spliddit platform) and the Adjusted Winner (AW) algorithm for household chore division (Kajibuntan platform). We evaluate both algorithms across eight fairness and efficiency metrics including envy-freeness (EF, EF1, EFX), proportionality (PROP), maximin share (MMS), equitability (EQ, EQ1), and Pareto optimality (PO).