Premium
Dynamic value voting and governance applications
Author(s) -
Flood Merrill M.
Publication year - 1991
Publication title -
behavioral science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.371
H-Index - 45
eISSN - 1099-1743
pISSN - 0005-7940
DOI - 10.1002/bs.3830360104
Subject(s) - voting , probabilistic logic , computer science , corporate governance , set (abstract data type) , value (mathematics) , selection (genetic algorithm) , mathematical economics , weighted voting , group (periodic table) , operations research , series (stratigraphy) , measure (data warehouse) , constant (computer programming) , mathematical optimization , economics , mathematics , artificial intelligence , political science , data mining , management , law , machine learning , organic chemistry , paleontology , chemistry , politics , biology , programming language
We discuss procedures that can be used by a group of voters to choose one from a set of alternatives on each of a series of elections. Our search is for procedures that do well for the group given some measure of group success. We approach this very general problem by considering a game that can be interpreted in terms of various kinds of practical applications, and outline some such applications. Our emphasis is on a dynamic procedure that enables the voters to vary the number of times they are counted on each vote, to reflect the varying importance of the votes from their standpoint, from a constant total allocation of votes for all voters combined. The final selection of an alternative is sometimes made probabilistically, calculated so as to maximize the expected value of the geometric mean of the normalized utility values supplied by the voters. We show that the dynamic procedure and the probabilistic selection have interesting consequences, well worth consideration for some applications. One application is a governance procedure for a university, specifically for Stanford University as an example.