Title: Robust inference in structural VARs with long-run restrictions
Authors: Guillaume Chevillon - ESSEC Business School (France) [presenting]
Sophocles Mavroeidis - Oxford University (United Kingdom)
Zhaoguo Zhan - Tsinghua University (China)
Abstract: Long-run restrictions are a very popular method for identifying structural vector autoregressions (SVARs). A prominent example is the debate on the effect of technology shocks on employment, which has been used to test real business cycle theory. The long-run identifying restriction is that non-technology shocks have no permanent effect on productivity. This can be used to identify the technology shock and the impulse responses to it. It is well known that long-run restrictions can be expressed as exclusion restrictions in the SVAR and that they may suffer from weak identification when the degree of persistence of the instruments is high. This introduces additional nuisance parameters and entails nonstandard distributions, so standard weak-instrument-robust methods of inference are inapplicable. We develop a method of inference that is robust to this problem. The method is based on a combination of the Anderson and Rubin test with instruments derived by filtering potentially non-stationary variable to make them near stationary. We find that long-run restrictions yield very weak identification is some cases. On the hours debate, we find that a previous difference specification is very well identified, while another level specification is weakly identified.