Efficient decision-makers evaluate relative reward per effort

Abstract

Understanding how humans and animals can make effective decisions would have a profound impact on economics, psychology, ecology, and related fields. Neoclassical economics provides a formalism for optimal decisions, but the apparatus requires a large number of evaluations of the decision options as well as representations and computations that are biologically implausible. This article shows that natural constraints distill the economic optimization into an efficient and biologically plausible decision strategy. In this strategy, decision-makers evaluate the relative reward across their options and allocate their effort proportionally, thus equalizing the reward per effort across their options. Using a combination of analytical and simulation approaches, the article shows that this strategy is efficient, providing optimal or near-optimal gain following a single evaluation of decision options. The strategy is also rational; satisficing and indifferent decision-makers are found to perform relatively poorly. Moreover, the relativistic value functions underlying this efficient strategy provide an account of time discounting, effort discounting, and resource discounting in general.