AnAxiomaticTheoryofProvably-Fair Welfare-CentricMachineLearning

Wedefineacomplementarymetric,termedmalfare, measuring overallsocietal harm, with axiomatic justification via the standard axioms of cardinal welfare, and cast fair ML asmalfare minimizationover the risk values(expected losses) ofeachgroup. Surprisingly,theaxioms ofcardinal welfare (malfare) dictate that this is not equivalent to simply defining utility as negativelossandmaximizing welfare.