Strategic Behavior is Bliss: Iterative Voting Improves Social Welfare

Authors: Joshua Kavner, Lirong Xia

NeurIPS 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical Our first main result (Theorem 1) states that, unfortunately, for any fixed m 3 and utility vector u, the ADPo A is Θ(n) for n agents. Therefore, the positive result achieved by Brˆanzei et al. [2013] is not upheld if u differs from plurality utility under the iterative plurality mechanism. To overcome this negative worst-case result, we introduce the notion of expected additive dynamic price of anarchy (EADPo A), which presumes agents truthful preferences to be generated from a probability distribution. Our second main result (Theorem 2) is positive and surprises us: for any fixed m 3 and utility vector u, the EADPo A is Ω(1) when agents preferences are i.i.d. uniformly at random, known as Impartial Culture (IC) in social choice.
Researcher Affiliation Academia Joshua Kavner Department of Computer Science Rensselaer Polytechnic Institute Troy, NY 12180 kavnej@rpi.edu Lirong Xia Department of Computer Science Rensselaer Polytechnic Institute Troy, NY 12180 xialirong@gmail.com
Pseudocode No The paper does not contain any structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide any statement or link regarding the availability of open-source code for the methodology described.
Open Datasets No The paper is theoretical and focuses on mathematical proofs and analyses under specific distributions (e.g., Impartial Culture), rather than using empirical datasets. No specific dataset access information is provided.
Dataset Splits No The paper is theoretical and does not describe training, validation, or test splits for any dataset.
Hardware Specification No The paper is theoretical and does not mention any specific hardware used for experiments.
Software Dependencies No The paper is theoretical and does not mention any specific software dependencies with version numbers.
Experiment Setup No The paper is theoretical and does not provide details about an experimental setup, hyperparameters, or training configurations.