Theorems 1 and 3, pp. 321, 323 — a solution satisfies WPO, An, S. Inv, Cont and Mon iff it is the Kalai–Smorodinsky solution
ProvedThomsonKS.Char.ks_characterizationConsider a variable population of agents. For each nonempty finite group , is the class of division problems: compact, convex, comprehensive subsets of containing a strictly positive vector. A solution selects a point for every and every . Then
where is the Kalai–Smorodinsky solution ( is the largest point of on the segment from the origin to the ideal point , ). The axioms are: weak Pareto-optimality (no point of is strictly better for everyone), anonymity (invariance under relabelling the agents), scale invariance (covariance with positive rescalings of each agent's utility), continuity in the Hausdorff metric, and population monotonicity (when new agents arrive and the problem of the original group is the slice of the enlarged problem , no original agent gains).
The paper states: "Theorems 1 and 3 together constitute the announced characterization of the Kalai–Smorodinsky solution."
Formalization Note Agents are natural numbers and groups are nonempty finite sets of them (for the empty group no solution could satisfy WPO, see the Setting item). A solution is a total function together with the hypothesis that for ; "is the Kalai–Smorodinsky solution" means agreement on every , the only problems on which a solution is defined. Every axiom evaluates only on problems in .
import Mathlib import Definitions.Def_ThomsonKS_Char_Setting
namespace ThomsonKS.Char
theorem ks_characterization (F : Solution) (hF : IsSolution F) :
(WPO F ∧ An F ∧ SInv F ∧ Cont F ∧ Mon F) ↔
∀ P : Finset ℕ, ∀ S ∈ DivProb P, F P S = KS P S := by sorry
end ThomsonKS.Char
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.