Complete markets strikes back: Is the reduced-form measure of consumption insurance reliable?

This paper highlights the identification problem of the reduced-form approach in quantifying the degree of consumption insurance as in Blundell et al. (2008, BPP thereafter). I argue that the reduced-form estimates are difficult to interpret in terms of the degree of consumption insurance. I show that BPP's empirical estimates of partial insurance parameters are consistent with a complete market model, both qualitatively and quantitatively, if the household income growth rate is positively correlated with patience. I also find that this model can simultaneously match the empirical increase of consumption dispersion over the life cycle.