“Individual and Social Welfare: A Bayesian Perspective”
Abstract: Ragnar Frisch, one of the greatest proponents of scientific methodology in economics, warned that one cannot unreflectively apply the methodology of natural science to economics. I will argue that the unreflective application of logical positivism to welfare economics in the mid-twentieth century did great harm to that discipline. It led Arrow to impose the independence of irrelevant alternatives on any ethical method of preference aggregation, which I join many others in considering an unfortunate idea. It further caused Arrow and most of the profession to adopt an unnecessary and unnatural interpretation of his Impossibility Theorem. Viewing that theorem as a disturbing paradox encouraged many in the profession to consider theoretical work on social welfare and social choice an unpromising dead end. Samuelson’s positivist-inspired criterion for a meaningful concept has left us without a fully developed and widely accepted language for discussing utility information richer than that which is generated by standard revealed preference. The second half of the paper wades into that forbidden territory, examining different ways that measurability, intensity, cardinality and comparability of utility are used in the literature and searching for alternative ways of understanding them.
CONTACTO DEL EVENTO