Risk Homeostasis in an Experimental Context

Abstract

The theory of risk homeostasis has been put forward as a tentative explanation for a jurisdiction’s traffic accident rate per km, per hour of road-user exposure, and per capita, as well as their pattern of interrelations. According to this theory, the accident rate per time unit of expo-sure is the end product of a homeostatic control process in which the level of risk accepted by the road user population—in return for the benefits accruing from their manner and amount of mobility—functions as the regulating variable. Although the real-life data lending support to this theory are many and varied, there is a lack of controlled experimental verification. The latter meets with obstacles of two kinds: inherent shortcomings of laboratory studies for the investigation of human behavior, and complications arising from the nature of the theory itself.

In an attempt to elicit risk-taking behavior under well-controlled conditions, an experiment was designed to explore a number of questions related to the theory in question. Individual subjects on their own played a simple “video game” for real money and were faced with two sources of risk: (a) uncertainty due to limitations in their own psychomotor skill and (b), in case of failure to meet a criterion of performance, the risk of financial loss which was externally controlled (by the experimenter) at two different levels of probability. The findings appear to lend partial support to RHT: the amount of money earned was independent of the subject’s skill; behavioral compensation occurred in relation to the two sources of risk, but not complete homeostasis.