Edited by Jan Toporowski and Jo Michell
Chapter 39: Risk
Risk, in economics, has traditionally been taken to be a matter of the variability of a magnitude of wealth or income or of the utility of such a magnitude. To differentiate risk from uncertainty, which also signifies variability, economists have generally followed Frank Knight ( 1985), who defined risk as a situation where the outcomes can be given specific probabilities and uncertainty where specific probabilities cannot be assigned. Knight thus held that risk can be insured against, while uncertainty cannot. A situation where either risk or uncertainty, as so defined, exists, however, cannot be one of certainty, so one might ask whether Knight’s distinction is one of degree or difference. Taking risk to be variability has quite naturally led economists to measure risk by the standard deviation or by the variance of the distribution of the relevant variable. This has been criticized on the grounds that these measures require a utility maximizer to have a quadratic utility function, which imposes some implausible conditions on an individual, such as greater aversion to constant additive risk at high levels of wealth than at low levels of wealth, which contradicts observed behavior, and Karl Borch’s (1969) demonstration of an inconsistency in the rank-orderings generated by the mean-standard deviation approach.
You are not authenticated to view the full text of this chapter or article.