Unfiltered consumption and the elasticity of intertemporal substitution

초록

This paper estimates the elasticity of intertemporal substitution (EIS) from the aggregate consumption Euler equation using Kroencke’s (2017) unfiltered consumption data. Unlike the published NIPA series used in most studies, unfiltered consumption is designed to recover the latent series obscured by statistical smoothing. Using weak-instrument-robust GMM inference, we show that this measurement choice matters for EIS estimation. With conventional filtered consumption, confidence sets typically include zero and are often uninformative. With unfiltered consumption, confidence sets exclude zero and provide a positive lower bound. The near-zero EIS estimates in the literature may partly reflect a measurement artifact rather than household preferences.

키워드