Cross-Dynastic Intergenerational Altruism

Does saving behavior reveal socially relevant intertemporal preferences? People concerned about the next generation as such might assign welfare weights on other dynasties. These concerns are captured in a model of saving by decomposing the present generation’s preference for the next into its dynastic and cross-dynastic components. With such preferences, saving for one's descendants benefits present members of other dynasties if they also care cross-dynastically. These preference externalities imply that socially relevant intertemporal preferences cannot be inferred from saving behavior. The external effect of present saving also decreases over time, implying that intertemporal preferences inferred from saving behavior are time-inconsistent. For illustration, let the cross-dynastic component take 10%, 20% or 50% of the value of the dynastic component from Nordhaus' calibration. This leads to efficient utility discount rates of 1.2%, 0.9% or 0.1%, compared to 1.5% from saving behavior. Hence, utility discount rates revealed by saving behavior should be lowered. Such adjustments are of particular importance for long term problems and reduce the efficient temperature increase by 2100 by 0.1ºC 0.2ºC or 0.5ºC, respectively, in a recent update to the DICE model