“…Although the resulting model is set-identified and the approach therefore raises serious robustness concern as we discussed above, the common practice is to consider single-prior Bayesian inference in set-identified SVARs. The large body of the empirical literature adopting this approach includes Canova and Nicolo (2002), Faust (1998), Mountford (2005), Rafiq and Mallick (2008), Scholl and Uhlig (2008), Uhlig (2005), and Vargas-Silva (2008) for applications to monetary policy, Dedola and Neri (2007), Fujita (2011), andPeersman andStraub (2009) for applications to business cycle model, Mountford and Uhlig (2009) for applications to fiscal policy, Kilian and Murphy (2012) for applications to oil prices. Alternative approaches that do not suffer from the pitfalls of single-prior Bayesian inference are Moon, Schorfheide, and Granziera (2013) and Gafarov, Meier, and Montiel-Olea (2016a,b), who consider frequentist inference for the identified set and Giacomini and Kitagawa (2015), who propose a robust Bayesian approach.…”