Smoothed Analysis of Social Choice, Revisited

Addresses the fundamental challenge in voting theory that designing voting rules satisfying all desirable properties is impossible, exploring this through “smoothed analysis” that introduces small random perturbations to votes. The paper provides simple sufficient conditions for smoothed-satisfaction or violation of various voting axioms, focusing on the Mallows noise model to present a more nuanced picture of when smoothed analysis can help resolve voting paradoxes versus scenarios offering limited practical benefit.