Learning Variability from Experience
- Elizaveta Konovalova, Max Planck Institute for Human Development, Berlin, Germany
- Gael Le Mens, Department of Economics and Business, Universitat Pompeu Fabra, Barcelona, Barcelona, Spain
AbstractLeading theories of risky choice predict that decision makers are sensitive to the variability of payoff distributions. Yet, little is known about how experience affects perceived variability. Existing empirical research on risky choice provides only inconclusive evidence about this issue because choices are not only affected by perceived variability but also perceived value and (unobserved) risk preferences. In re-analyses of experimental data and survey data from two nationally representative panels, we show that perceived variability strongly depends on sample variability. In a new experiment, we also demonstrate that perceived variability systematically depends on sample size, a result consistent with the predictions of a recent theoretical paper by the authors (Konovalova & Le Mens, 2017).