Which Decision Theory Describes Life Satisfaction Best? Evidence from Annual Panel Data

We use an annual household panel to test which features of prospect theory can be supported by measures of life satisfaction. We also test whether recalled or expected life satisfaction is anchored at current life satisfaction and adjusted in the direction of the recall or expectation. Using a fixed effects estimator we find that life satisfaction contains features of both classic expected utility and prospect theory. Life satisfaction depends positively on levels of income, good health, and on employment. It also depends positively on income and employment improvements, however the reverse is true for health increases. Life satisfaction is concave in income gains and convex in income losses, and it exhibits loss aversion in income and employment status, but not in health. Moreover, we find that current levels of life satisfaction are better predictors of recalled (expected) life satisfaction than past (future) life satisfaction. The results support viewing life satisfaction as representing a mixture of the classic decision utility of expected utility theory, and the value function of prospect theory. Subjects seem to use an anchoring and adjustment heuristic when answering questions about past and expected life satisfaction.