Daniel Kahneman’s underappreciated last published paper: Empirical implications for benefit-cost analysis and a chat session discussion with bots

Abstract Nobel Prize winner Daniel Kahneman’s last published paper is an adversarial collaboration in which he and Matthew Killingsworth reconcile conflicting empirical results from their previous research on income and reported happiness, with Barbara Mellers as a facilitator. The empirical results use quantile regression to allow for measured income heterogeneity effects that include notch points in the estimated marginal utilities of income. Our analysis examines Kahneman’s last paper’s conceptual innovations and challenges to assumptions about diminishing marginal utility of income. We review his contributions to emotional well-being measurement and employ a novel AI-simulated dialogue between the late Amos Tversky and Sir Angus Deaton to explore interdisciplinary perspectives on the findings. Our paper demonstrates how Kahneman’s final research undermines recent arguments for incorporating income redistribution simply into benefit-cost analysis, suggesting that such objectives remain better addressed through fiscal policy rather than regulatory interventions. His final published work exemplifies Kahneman’s commitment to empirical precision and theoretical flexibility, even when contradicting his earlier conclusions.