Intertemporal Elasticity of Substitution with Leisure Margin

This paper investigates whether leisure time definitions matter in the estimation of the intertemporal elasticity of substitution of consumption (IES) by using a utility specification that allows interaction between consumption and leisure time. We find that the IES estimated using a narrowly defined leisure measure that excludes quasi-leisure activities is larger than that estimated using nonmarket time. The discrepancy is largely driven by the substitution of consumption and several leisure components over the lifecycle. This finding is robust in alternative specifications and holds well for subsamples of higher socioeconomic status. Our results demonstrate the inseparable nature of consumption and time allocation.