I’m trying to replicate a modified version of Christiano, Eichenbau and Evans 2005, where, for pure didactic purposes, there is no money. The model must be linearized around the steady state. The baseline version, when shocking public expenditure, works, but in a second version I’m required to add distortionary taxes on consumption, labor and capital, plus a lump sum transfer/tax that evolves followibg an AR(1) process. The distortionary taxes follows a feedback rule reacting both to output and government bonds, as government spending. This model works as well, but it produces impulse response functions that oscillate around the steady state. I know from this foum that this maybe a timing problem, and I’ve also tried to simplify the model but I cannot understand where the issue come from. also typing “model_diagnostics(M_,options_,oo_)” doesn’t give me a clear answer, but simply “No obvious problems with this mod-file were detected”.
With fiscal rules, it’s not necessarily a problem of timing, but of parameterization of the feedback rules. Have you tried different parameter values?
Thank you professor! We’ve tried with a different calibration and results have definitely improved.