Risk Attitude

Risk taking is generally considered to be the expression of a personality trait called risk attitude. People are assumed to take a certain amount of risk across a range of situations. Some people are more risk-averse because the uncertainty and potential downside of risky options makes them anxious, while others are more risk seeking because the uncertainty and potential upside of risky options excites them.

Depending on how risk attitude is assessed, the assumption that it is consistent across situations is either true or false. Risk attitude as defined within the expected utility (EU) theory framework can vary greatly across decisions in different content domains and when outcomes are described as gains or losses. Risk taking also depends on how people know about possible decision outcomes (from personal trial-and-error learning vs. being given a statistical description). Risk attitude is more stable across situations when conceptualized within a risk-benefit framework that models risk taking as a trade-off between decision makers' perception of the riskiness and benefit of choice options and assumes that this trade-off is governed by their attitude toward risk, that is, willingness to trade off (perceived) risk for (perceived) benefits: (Preference for option X) = (Perceived benefit of X) – b(Perceived risk of X). Larger values of trade-off parameter b mean that the decision maker is more risk-averse. Take the (hypothetical) example of two cancer treatments, surgery and radiation, which offer different prospects of extending life. Surgery entails a 10% chance of death during the procedure but a 90% chance of extending life by 10 years. It thus has the benefit of extending life on average by 9 years (Expected value (Surgery) = .10(0) + .90(10) = 9) but also carries some risk of immediate death. This unpredictability of outcomes is modeled in finance by the statistical variance of outcomes (which also happens to be 9), but other, less analytic factors also play a role. Radiation has no danger of immediate death but offers shorter possible life extensions, with a 50% chance of living for 2 additional years and a 50% chance of living for 5 additional years. The benefit of extending life is thus lower for radiation, namely, only 3.5 years on average (Expected value (Radiation) = .50 (2) + .50 (5) = 3.5), but its variance of outcomes is also lower, namely 2.25, which factors into perceptions of risk. If expected life extension and the variance in life extensions fully described perceived benefits and risks, respectively, then a person who had a trade-off coefficient b of .5 would have a preference of 4.5 (i.e., 9 – .5(9)) for surgery and one of 2.375 (i.e., 3.5 −.5(2.25)) for radiation, and thus would prefer surgery.

When modeled within a risk-benefit trade-off framework, situational differences in risk taking turn out to result from different perception of risks and benefits across situations (e.g., risks are perceived to be larger or benefits smaller in some situations compared with others) rather than from differences in the trade-off coefficient b, that is, the willingness to take on (perceived) risk to obtain (perceived) benefit, which is fairly stable for a given individual.

...