Michel Balinski - Majority judgment:Why it should be used to rank and elect

Intervenant : 
Michel Balinski, CNRS and CREST, Ecole Polytechnique

Information détaillée : 


Michel Balinski, earned his B.A. at Williams College, then completed an M.S. in economics at MIT and a Ph.D. in mathematics at Princeton. He has held academic positions at Princeton University, University of Pennsylvania, City University of New York Graduate Center, Yale University and Stony Brook University. Among other short-term appointments he was a visiting professor of mathematics at Grenoble in 1974-75. Beginning 1982 he was Directeur de Recherche de classe exceptionnelle of the CNRS at the Ecole Polytechnique, Paris. An INFORMS Fellow, he was awarded INFORMS’s Lanchester Prize in 1965, the MAA’s Lester R. Ford Award in 1976 and in 2009, an honorary degree in mathematics from the University of Augsburg in 2004, and INFORMS’s John von Neumann Theory Prize in 2013. He is the founding editor of Mathematical Programming and a past President of the Mathematical Optimization Society. He is the author of Fair Representation: Meeting the Ideal of One Man, One Vote (1982, reissued 2001, with H. P. Young), Le suffrage universel inachevé (2004), and Majority Judgment: Measuring, Ranking and Electing (2011, with R. Laraki), and the author or co-author of about 150 articles. His principal current interest is the design of electoral systems. One of his electoral systems is used in several Swiss cantons.


Résumé : 

Every well-known voting system in use today hides important vices that can deny the will of the electorate including majority vote with only two candidates (the domination paradox), approval voting, all methods that ask voters to compare candidates (i.e., rank-order them), and point-summing methods. The underlying reason: the inability of voters to adequately and honestly express their opinions.
   Majority judgment asks voters to evaluate every candidate in an easily understood common language of ordinal grades such as: Great, Good, Average, Poor, or Terrible. Majorities determine the electorate’s evaluation of each candidate and the ranking between every pair of candidates (necessarily transitive), with the first-placed among them the winner.
   Majority judgment is described together with illustrations of its use (notably, French and U.S. presidential elections).
   It was specifically designed to
   • permit voters to express their opinions,
   • be meaningful in the sense of measurement theory,
   • avoid Condorcet’s paradox (guarantee a transitive order-of-finish),
   • avoid Arrow’s paradox (when the order-of-finish of two candidates depends on the presence/absence of other candidates),
   • combat strategic manipulation and encourage the honest expression of opinions.
    Majority judgment has proven itself in practice. It can and should be used in elections with many voters as well as by juries with few judges (e.g., for figure skaters, gymnasts, wines, films, restaurants, prize winners, . . . ).

Some references :
• M. Balinski and R. Laraki. 2011. Majority Judgment: Measuring, Ranking, and Electing, M.I.T. Press.
• – and –. 2013. “Jugement majoritaire versus vote majoritaire (via les présidentielles 2011-2012).” Revue Française d’Economie XXVII 11-44.
• – and –. 2014. “Judge: Don’t Vote! ”, Operations Research 62 483-511.
• – and –. 2016. “Trump and Clinton victorious: proof that US voting system doesn’t work,” TheConversation, May (available on web).
• M. Balinski. 2016. “How majority voting betrayed voters again in 2016,” TheConversation, December (available on web).
• –. 2018. “Réponse à des critiques du jugement majoritaire,” CREST, working paper series #2018-10 (available on web).