This paper first considers the implications of biological evolution for economic preferences. It analyzes why utility functions evolved, considers evidence that utility is both hedonic and adaptive, and suggests why such adaptation might have evolved. Time preference and attitudes to risk are treated—in particular, whether the former is exponential and the latter are selfish. Arguments for another form of interdependence—a concern with status—are treated. The paper then considers the evolution of rationality. One hypothesis examined is that human intelligence and longevity were forged by hunter-gatherer economies; another is that intelligence was spurred by competitive social interactions.

Consider the possible biological origin of the expected utility criterion. On the one hand, if individuals possess a utility function stemming from the rate of production of expected offspring, they can rapidly adapt to arbitrary unknown distributions in a bandit problem. Embedding such a utility function in a simple rule of thumb involving no beliefs about probabilities leads to evolutionary optimality. On the other hand, if any rule whatever yields evolutionary optimality for all distributions, this precise utility function must be implicit, in a revealed preference sense.