Abstract

This paper assesses SVARs as relevant tools at identifying the aggregate effects of news shocks. When the econometrician and private agents ’ information sets are not aligned, the dynamic re-sponses identified from SVARs are biased. However, the bias vanishes when news shocks account for the bulk of fluctuations in the economy. A simple correlation diagnostic test shows that under this condition, news shocks identified through long–run and short–run restrictions have a correlation close to unity.