Abstract Dynastic models common in macroeconomics use a single parameter to control the willingness of individuals to substitute consumption both intertemporally, or across periods, and intergenerationally, or across parents and their children. This article defines the concept of elasticity of intergenerational substitution (EGS), and extends a standard dynastic model in order to disentangle the EGS from the EIS, or elasticity of intertemporal substitution. A calibrated version of the model lends strong support to the notion that the EGS is significantly larger than one. In contrast, estimates of the EIS suggests that it is at most one. What disciplines the identification is the need to match empirically plausible fertility rates for the U.S. We illustrate the potential role of the EGS in macroeconomics.

We show that state non-separable preferences à la Epstein–Zin–Weil (EZW) provide a tractable and flexible framework to study the economics of health and longevity. This utility representation: (1) admits a preference for timing of resolution of uncertainty regarding mortality risks; (2) links the marginal valuation of survival to the level of survival; (3) can preserve homotheticity even for low degrees of intertemporal substitution without generating implausible predictions regarding the value of life; and (4) adds needed flexibility to account for the empirical evidence on the value of life. We illustrate the implications of EZW preferences for the economic value of observed differences in life expectancy across countries and over time, and for the value of life over the life cycle.