Probabilistic versus non-probabilistic decision making: Savage, Shackle and beyond

This paper discusses the evolution of decision theory after Savage’s Foundations. Two developments are examined. First, it is presented the rationale of Shackle’s proposal to abandon probabilistic decision making. Second, it is discussed the axiomatisation provided by the non-additive probability approach to account for the experimental evidence originated by the Ellsberg Paradox. An attempt is made to establish a connection between Shackle’s non-probabilistic instances and non-additive probabilistic decision making. The main outgrowth of the paper is that the similarities between the non-additive approach and Shackle’s theory are not limited to a number of methodological statements. In fact, it is also the formal measures used in the two contexts for representing individual preferences in uncertain environments that resemble each other