Dynare forums

dear professor Pfeifer;

I have a question about parameter estimation using bayesian,: can we estimate parameters those are not related to the specified shocks, i.e., if I do not specify a wage stickiness shock, nor do I have wage data, can I use bayesian estimation and other variable data to estimate the wage stickiness parameters?

best
shuixing

In general, the answer is yes. To stay with your example, the value of the wage stickiness parameter will typically have implications for the probability distribution of the other observed variables. However, you should always check whether your observed data is informative for the parameters you want to estimate, i.e. you should use the

Code: Select all

identification

command

jpfeifer wrote:In general, the answer is yes. To stay with your example, the value of the wage stickiness parameter will typically have implications for the probability distribution of the other observed variables. However, you should always check whether your observed data is informative for the parameters you want to estimate, i.e. you should use the

Code: Select all

identification

command

I see, thanks, professor Pfeifer.