Identification in structural vector autoregressive models with structural changes

An increasing strand of the literature uses structural changes and different heteroskedasticity regimes found in the data constructively to improve the identification of structural parameters in Structural Vector Autoregressions (SVAR). A standard assumption in this literature is that the reduced form unconditional covariance matrix of the system varies while the structural parameters remain constant. Under this condition it is possible to identify the SVAR without the need to resort to theory-driven restrictions. With macroeconomic data, the hypothesis that the structural parameters are invariant to breaks is untenable. This paper investigates the identification issues that arise in SVARs when structural breaks occurring at known dates affect both the reduced form covariance matrix and the structural parameters. The knowledge that different heteroskedasticity regimes characterize the data is combined with theory-driven restrictions giving rise to new necessary and sufficient local identification rank conditions which generalize the ones which apply for SVARs with constant parameters. This approach opens interesting possibilities for practitioners. An empirical illustration shows the usefulness of the suggested identification strategy by focusing on a small monetary policy SVAR of the U.S. economy. Two heteroskedasticity regimes are found to characterize the data before andafter the 1980s and this information is combined with economic reasoning to identify the effect of monetary policy shocks on output and inflation