By Enzo A. Cerletti and Josep Pijoan-Mas
http://d.repec.org/n?u=RePEc:cmf:wpaper:wp2012_1206&r=dge (In case of doubt, working link here)
In this paper we study the transmission of income shocks into nondurable consumption in the presence of durable goods. We use a standard a life-cycle model with two goods to characterize the interaction of durability of goods, durability of shocks, and borrowing constraints as determinants of shock transmission. We show that borrowing constraints lead to a substitution between durable and non-durable goods upon arrival of an unexpected income change. This substitution biases the conventional measures of insurance based on the response of non-durable consumption to income changes. The sign of this bias depends critically on the persistence of the shock. We show that households have less insurance against transitory shocks and more insurance against permanent shocks than commonly measured. We calibrate the model economy to the US in order to measure the size of this bias.
This is a nice illustration on how ommitting durable goods can seriously mess up the estimation of the response of non-durable goods to income shocks. This is an example of why it is so important to think in terms of theory when estimating elasticities, and of general equilibrum in particular.