Evolution of Risk Aversion: The 'Having Descendants Forever' Approach

Abstract

Recent models of the evolution of preferences have provided profound new insights into the origins of risk attitudes. In most of these models the evolutionary "objective function" is the maximization of the expected number of offspring, or alternatively, the maximization of the geometric-mean growth rate. We suggest that careful consideration should also be given to the objective of maximizing the probability of Having Descendants Forever (p(HDF)), first suggested by Meginniss (1977). We show that the p(HDF) criterion implies risk aversion. Moreover, it leads to preferences that are very closely approximated by the power CRRA preferences. Thus, constant relative risk aversion can be viewed as an evolutionary-developed heuristic aimed to maximize the probability of having descendant forever.

Keywords: Evolution, preference, having descendants forever, constant relative risk aversion, loss aversion

JEL Classification: D81, D03