Identifying Technology Shocks in Models with Heterogeneous Inputs

We show that the key identifying assumptions underlying the existing approaches to identifying technology shocks in the data are violated in models with heterogeneous capital and labor. We propose a new method to identifying technology shocks in the data in presence of factor heterogeneity and prove its identification. We find that hours respond positively to positive technology shocks identified using the proposed procedure. The identified technology shocks account for a sizable fraction of the business cycle volatility in macroeconomic aggregates.