This paper examines the liberal paradox. When the set of alternatives, X, is a disjoint union of sets \(X_ i\), one for each voter, the liberal paradox occurs if and only if the Pareto set contains no element u of any \(X_ i\) such that voter i prefers u to any other element of \(X_ i.\) When the set of all preference profiles is given the equiprobable distribution, the size of the Pareto set is estimated and it is shown that (1) if the number of voters is fixed but the number of alternatives tends to infinity, the probability of the liberal paradox tends to one; and (2) if the number of alternatives is a fixed ratio of the number of voters and both diverge to infinity, the probability of the liberal paradox tends to zero. It is also shown that for any preference profile and any set of rights (partition of X into a number of sets equal to the number of voters) we can redistribute those rights (permute the sets) in such a way that the liberal paradox does not occur. Also, if preferences are single-peaked and the rights of each person are adjacent to one another in the ordering used to define single-peakedness, the liberal paradox does not occur.
- Liberals and information
- Justice, liberty, unanimity and the axioms of identity
- The Libertarian resolution of the Paretian liberal paradox
- Respect for experts vs. respect for unanimity: the liberal paradox in probabilistic opinion pooling
- The Paretian liberal with intuitionistic fuzzy
- The effect of unconditional preferences on Sen's paradox
- The publications of Ki Hang Kim
- ``The liberal paradox: A generalization by D. Kelsey
- What to make of the Liberal Paradox?
- ``The liberal paradox: A generalization by D. Kelsey
- On the possibility of Paretian liberalism: a comment
- The Libertarian paradox: Some further observations
This page was built for publication: The liberal paradox and the Pareto set
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1065690)
