Biased experts, majority rule, and the optimal composition of committee Articles uri icon

publication date

  • May 2021

start page

  • 1

end page

  • 27


  • 127

International Standard Serial Number (ISSN)

  • 0899-8256

Electronic International Standard Serial Number (EISSN)

  • 1090-2473


  • An uninformed principal appoints a committee of experts to vote on a multi-attribute alternative, such as an interdisciplinary project. Each expert evaluates one attribute and is biased toward it (specialty bias). The principal values all attributes equally but has a status quo bias, reflecting the organizational cost of a change. We study whether the principal would compose the committee of more or less specialty-biased experts. We show that her optimal composition is nonmonotonic in the majority rule, with the most biased experts appointed under intermediate rules. We then show that the principal would be less concerned about the committee composition if its members can be uninformed, as they induce the informed to vote less strategically. Surprisingly, although uninformed members lower the quality of the committee's decision, the principal may prefer to have some when its composition is suboptimal, and the majority rule is sufficiently extreme, such as the unanimity.


  • Mathematics
  • Statistics


  • bias; committee composition; majority rule; partisanship