Abstract:
We study risky inter-temporal choice in a large random student sample (n=721) and a large rural sample (n=835) in Malawi. All respondents were exposed to the same 20 Multiple Choice Lists with a rapid elicitation method that facilitated the identification of near-future Certainty Equivalents of future risky prospects placed 6, 12, and 24 months into the future. The probabilities of winning in the risky future prospects varied and facilitated the estimation of probability weighting functions for the risky prospects placed 6 and 12 months into the future. The experiment is used to test whether decision errors can explain or be highly correlated with hyperbolic discounting and non-linear (inverse-S-shaped) probability weighting. We find evidence that decision errors are strongly correlated with hyperbolic discounting but do not find that decision errors are correlated with the strong inverse-S-shaped probability weighting (w(p)) patterns in our two samples. We find stronger S-shaped and more pessimistic w(p) functions for 6-month horizon risky prospects than for 12-month horizon risky prospects in both samples. Both patience and optimism bias contribute to subjects taking higher risks related to more risky distant future prospects. This can lead to the postponement of climate action.