I am estimating some parameters for germany and france.
one of the parameters not estimated is the net replacement rate which is different in both countries.
Should these be different for both countries while estimation or is it not taken into account.
slide 8 of wouter den haan suggests that these values are ignored during estimation by dynare ?
wouterdenhaan.com/numerical/ … mation.pdf
Therefore, the same replacement rate which is an average of the two should not matter during estimation? Am i correct ?
And how does the use calibration command, in the estimated_params_init block efect this ?
No. What Wouter says is that the fixed values before the model block for estimated parameters are superseded by new values when these parameters are estimated. Thus, these initial parameter values are ignored during estimation. If you don’t specify a particular parameter as estimated, Dynare will take this parameter values as given. Of course, this value will matter during estimation. See Remark 21 (Calibration vs. Estimation) in Pfeifer(2013): “A Guide to Specifying Observation Equations for the Estimation of DSGE Models” sites.google.com/site/pfeiferecon/Pfeifer_2013_Observation_Equations.pdf
The use_calibration-command of the estimated_params_init is just used to specify the starting value for the mode-finding to be at the fixed values before estimation instead of starting at the prior mean.