Fairness in group identification

Research output: Contribution to journalArticlepeer-review

Abstract

We study the problem of classifying individuals into groups, using agents’ opinions on who belong to which group as input. Our focus is on the rules that satisfy equal treatment of equals, a minimal fairness property, in addition to independence of irrelevant opinions and non-degeneracy. We show that a rule satisfies the three axioms if and only if it is the liberal rule, a strong one-vote rule, a one-row rule, or a one-column rule. The last three families of rules can be ruled out by simple, intuitive properties. Thus, invoking equal treatment of equals, which is substantially weaker than symmetry, we obtain a characterization of the liberal rule.

Original languageEnglish
Pages (from-to)35-40
Number of pages6
JournalMathematical Social Sciences
Volume94
DOIs
Publication statusPublished - 2018 Jul
Externally publishedYes

Bibliographical note

Funding Information:
I thank an Associate Editor and two anonymous referees for comments that greatly improved the earlier version of this paper. I am also grateful to Biung-Ghi Ju and William Thomson for their helpful discussions. This work was supported by the Ministry of Education of the Republic of Korea and the National Research Foundation of Korea ( NRF-2016S1A3A2924944 ).

Publisher Copyright:
© 2018 Elsevier B.V.

ASJC Scopus subject areas

  • Sociology and Political Science
  • General Social Sciences
  • General Psychology
  • Statistics, Probability and Uncertainty

Fingerprint

Dive into the research topics of 'Fairness in group identification'. Together they form a unique fingerprint.

Cite this