Economic Quarterly

Two principal conclusions emerge regarding the problem of identification in structural vector autoregression (SVAR) estimation. First, in a manner analogous to instrumental variables estimation with weak instruments, certain identification schemes can lead to parameter estimates whose distributional properties diverge from the traditional ones. The validity of the technique is, therefore, implicitly linked to the identification strategy adopted. Second, the dynamic implications that result from SVAR work can be highly sensitive to the choice of identifying restrictions. It follows that sensitivity analysis should form an integral part of the interpretation of SVAR results.