Emotions and risk attitudes

Previous work has shown that preferences are not always stable across time, but surprisingly little is known about the reasons for this instability. I examine whether variation in people's emotions over time predicts changes in risk attitudes. Using a large panel data set, I identify happiness, anger, and fear as significant correlates of within-person changes in risk attitudes. Robustness checks indicate a limited role of alternative explanations. An event study around the death of a parent or child further confirms a large relationship between emotions and risk attitudes.